33rd Symposium of the Society for Nonlinear Dynamics & Econometrics

Summary of All Sessions
Click here for an index of all participants

Date/Time	Title/Location	Papers
March 30, 2026 9:00-10:00	Plenary I - Ricardo Reis, London School of Economics     Location: Room: 512	0
March 30, 2026 10:30-12:10	A1: Empirical Macro I     Location: Room 528	4
March 30, 2026 10:30-12:10	A2: Empirical Macro - Inflation I     Location: Room 513	4
March 30, 2026 10:30-12:10	A3: Empirical Macro - Oil Markets     Location: Room 514	3
March 30, 2026 10:30-12:10	A4: Empirical Macro II     Location: Room 515	4
March 30, 2026 10:30-12:10	A5: Macro Theory I     Location: Room 516	4
March 30, 2026 10:30-12:10	A6: Time Series - Cointegration     Location: Room 517	3
March 30, 2026 10:30-12:10	A7: Empirical Macro III     Location: Room 526	4
March 30, 2026 10:30-12:10	A8: Time Series I     Location: Room 527	4
March 30, 2026 13:30-15:10	B1: Empirical Macro IV     Location: Room 528	4
March 30, 2026 13:30-15:10	B2: Time Series III     Location: Room 513	4
March 30, 2026 13:30-15:10	B3: Empirical Macro V     Location: Room 514	3
March 30, 2026 13:30-15:10	B4: Macro Theory - Inflation     Location: Room 515	4
March 30, 2026 13:30-15:10	B5: Macro Theory - Monetary Policy I     Location: Room 516	4
March 30, 2026 13:30-15:10	B6: Time Series III     Location: Room 517	4
March 30, 2026 13:30-15:10	B7: Time Series IV     Location: Room 526	4
March 30, 2026 13:30-15:10	B8: Empirical Macro VI     Location: Room 527	3
March 30, 2026 15:40-17:20	C1: Empirical Macro - Inflation II     Location: Room 528	2
March 30, 2026 15:40-17:20	C2: Empirical Macro, Time Series & Forecasting     Location: Room 513	4
March 30, 2026 15:40-17:20	C3: Empirical Macro - Fiscal Policy     Location: Room 514	4
March 30, 2026 15:40-17:20	C4: Macro Theory II     Location: Room 515	4
March 30, 2026 15:40-17:20	C5: Macro Theory - Monetary Policy II     Location: Room 516	4
March 30, 2026 15:40-17:20	C6: Time Series V     Location: Room 517	4
March 30, 2026 15:40-17:20	C7: Empirical Macro - Climate     Location: Room 526	4
March 30, 2026 15:40-17:20	C8: Empirical Macro VII     Location: Room 527	4
March 31, 2026 9:00-10:40	D1: Empirical Macro & Finance     Location: Room 527	4
March 31, 2026 9:00-10:40	D2: Empirical Macro & Macro Theory - Uncertainty     Location: Room 513	4
March 31, 2026 9:00-10:40	D3: Empirical Macro VIII     Location: Room 514	3
March 31, 2026 9:00-10:40	D4: Forecasting I     Location: Room 515	4
March 31, 2026 9:00-10:40	D5: Macro Theory III     Location: Room 516	4
March 31, 2026 9:00-10:40	D6: Empirical Macro - Unemployment     Location: Room 517	3
March 31, 2026 9:00-10:40	D7: Macro Theory IV     Location: Room 526	3
March 31, 2026 11:10-12:10	Plenary II: Martin Ellison, University of Oxford, Nuffield College, CEPR     Location: Room: 512	0
March 31, 2026 13:30-15:10	E1: Empirical Macro - Inflation III     Location: Room 527	4
March 31, 2026 13:30-15:10	E2: Empirical Macro X     Location: Room 513	4
March 31, 2026 13:30-15:10	E3: Empirical Macro XI     Location: Room 514	4
March 31, 2026 13:30-15:10	E4: Finance     Location: Room 515	4
March 31, 2026 13:30-15:10	E5: Forecasting II     Location: Room 516	3
March 31, 2026 13:30-15:10	E6: Empirical Macro & Time Series     Location: Room 517	2
March 31, 2026 13:30-15:10	E7: Empirical Macro IX     Location: Room 526	4

 

40 sessions, 140 papers, and 0 presentations with no associated papers

33rd Symposium of the Society for Nonlinear Dynamics & Econometrics
Detailed List of Sessions

Session: Plenary I - Ricardo Reis, London School of Economics March 30, 2026 9:00 to 10:00 Location: Room: 512
Session: A1: Empirical Macro I March 30, 2026 10:30 to 12:10 Location: Room 528
Session Chair: Tatevik Sekhposyan, Texas A&M University
The Credit Channel of Economic Policy Uncertainty Abstract Please see submission of extended abstract
Presented by: Anastasia Allayioti, European Central Bank
High-Frequency Identification of Uncertainty Shocks: Reconciling Proxy-SVAR and Local Projection Methods Abstract The identification and propagation of uncertainty shocks remain central questions in macroeconomics, particularly given recent episodes of elevated uncertainty from pandemic disruptions, geopolitical tensions, and monetary policy normalization. While both proxy-SVAR and local projection methods have become standard tools for shock identification, practitioners face a puzzling reality: these methods often yield substantially different impulse responses even when using identical identification strategies. This paper provides the first systematic investigation of why these discrepancies arise and when each method should be preferred. We develop a novel high-frequency identification strategy using minute-level VIX movements in narrow windows around FOMC announcements and other policy events, exploiting variation within [−30, +90] minute windows around events spanning two decades to provide cleaner exogenous variation than existing daily or monthly measures. Using this identification, we document economically significant and systematic differences between proxy-SVAR and local projection estimates of uncertainty effects, with important implications for optimal policy responses to uncertainty spikes. We trace these discrepancies to three mechanisms: (i) dynamic misspecification from VAR lag truncation, (ii) state-dependent propagation that linear SVARs cannot capture, and (iii) differential weighting of observations when instrument strength varies over time. Through extensive Monte Carlo analysis and empirical investigation, we identify which mechanisms dominate under different conditions, providing practitioners with concrete guidance on method selection based on their specific empirical context.
Presented by: Alessia Paccagnini, University College Dublin
Measuring Growth Uncertainty with Random Forest Abstract We propose two general machine learning based conditional density estimators that are flexible and remain effective in high-dimensional settings. First, the quantile density forest estimates conditional quantiles without quantile crossing and also yields full conditional density estimates in a single forest. Second, the Kullback-Leibler density forest estimates a joint density and then recovers the conditional density by marginalization. Both methods effectively predict conditional densities with high-dimensional covariates and outperform the standard local kernel density estimator, particularly in capturing tail behavior. Density predictions for future U.S. GDP growth and inflation show that both methods deliver superior forecast accuracy in measuring tail risk in high dimensions compared to the existing linear quantile regression and skewed t-distribution approach. These advances provide policymakers with robust, nonparametric tools for assessing macroeconomic tail risks.
Presented by: Joon Park, Indiana University
Uncertainty Shocks and Macroeconomic Effects: Insights from Volatility Term Structures Abstract This paper introduces a new approach to measuring uncertainty shocks by exploiting the term structure of VIX futures. We document that VIX futures display notable within-month dynamics, including frequent episodes of inversion. Harnessing these dynamics allows us to assess the causal effects of uncertainty shocks, which can be either expansionary or contractionary depending on how the term structure rotates. Furthermore, we show that correlations between existing uncertainty measures—such as the Economic Policy Uncertainty index—and VIX futures vary across horizons: at times stronger with short-term futures and at other times with longer maturities. This variation provides evidence consistent with the potential decoupling of uncertainty indices, a phenomenon frequently emphasized in the literature.
Presented by: Tatevik Sekhposyan, Texas A&M University
Session: A2: Empirical Macro - Inflation I March 30, 2026 10:30 to 12:10 Location: Room 513
Session Chair: Christian Conrad, Heidelberg University
Financial Market Reactions to U.S. Fiscal News: Evidence from a Novel Daily News Database Abstract We construct a novel daily-frequency fiscal news database for the United States from a historical archive of news feeds that closely represent financial market participants’ real-time information set. Leveraging the archive’s granular coverage, we recover a fiscal policy surprise series to study how fiscal news transmits to the macro-financial environment. We find that expansionary fiscal news significantly increases long-term Treasury yields and term premia. These movements coincide with upward revisions in market-based inflation expectations and risk premia at both short and long horizons. By contrast, fiscal-uncertainty news induces a risk-off response, lowering yields and raising market volatility, underscoring the importance of distinguishing the information content of fiscal developments. We further assess the role of systematic monetary policy in shaping fiscal transmission to financial markets. Employing Lucas critique–robust empirical counterfactuals, we show that the monetary policy reaction to fiscal news shapes the responses of stock prices, the U.S. dollar and near-term inflation expectations. Finally, we document substantial spillovers of U.S. fiscal policy to international sovereign bond and equity markets.
Presented by: Gökhan Ider, DIW Berlin, Berlin School of Economics
Who’s on FIRE? Household characteristics and the formation of inflation expectations Abstract We study how consumers form and revise inflation expectations at the aggregate and at the individual level using a panel of Dutch households surveyed monthly since 2019. We apply a Bayesian framework which nests Full Information Rational Expectations (FIRE) alongside other forecasting heuristics. At the aggregate level, consumers systematically overreact to current inflation, mirroring patterns found among professionals. At the individual level, only 2.5 percent of households behave as FIRE-consistent forecasters—mostly wealthy, educated men. The majority rely on adaptive expectations with upward bias or on fundamentalist anchors, while a sizable share cannot be captured by simple heuristics. These results highlight substantial heterogeneity in forecasting behavior and suggest that FIRE is the exception, not the rule.
Presented by: Richhild Moessner, BIS
Firms' Inflation Expectations in a Monetary Union Abstract Using data from the euro area SAFE, a novel survey of firms’ inflation expectations including a randomized controlled trial (RCT), we show that firms’ inflation expectations exhibit significant heterogeneity, challenging the predictions of full-information rational expectations models. At the same time, we document that firms update beliefs rationally but under incomplete information, with geographic location playing a dominant role in shaping expectations. Firms extrapolate from regional and national inflation to form euro area inflation expectations. A basic "Lucas island" model calibrated to euro area data replicates key empirical moments and highlights the structural “pass-through” from national to aggregate expectations. Our findings underscore challenges in anchoring inflation expectations in a heterogeneous monetary union.
Presented by: Timo Reinelt, Federal Reserve Bank of San Francisco
Inflation Forecast Targeting Revisited Abstract Under inflation forecast targeting, central banks such as the ECB adjust policy to keep expected inflation on target. We evaluate the ECB’s inflation forecasts: they are unbiased and efficient but contain little information at forecast horizons beyond three quarters. This suggests a quick impact of monetary policy on inflation, which calls for short target horizons. This is in line with standard theory. In a New Keynesian model with transmission lags, inflation forecast targeting is indeed effective in stabilizing inflation—provided there is no forward-looking behavior—though the information content of forecasts is unrealistically high. In the presence of forward-looking behavior, the information content declines because monetary policy becomes more effective in meeting the target, but inflation is best stabilized by targeting current inflation.
Presented by: Christian Conrad, Heidelberg University
Session: A3: Empirical Macro - Oil Markets March 30, 2026 10:30 to 12:10 Location: Room 514
Session Chair: Hilde Bjørnland, BI Norwegian Business School
A Global Oil Market Model with Shipping Costs Abstract This paper investigates the role of shipping costs in global crude oil and refined petroleum markets. For this purpose a Global VAR (GVAR) model is estimated jointly for the oil and refined petroleum markets; this includes the Baltic Dirty Tanker Index (BDTI) and the Baltic Clean Tanker Index (BCTI) as measures of the cost of shipping crude oil and refined petroleum commodities, respectively. The results suggest that shocks to the cost of shipping petroleum commodities have a particularly severe negative impact on real economic activity and on refined petroleum consumption in most regions. Shocks to the price of crude oil and refined petroleum instead have inflationary effects, especially in countries that are net importers of those commodities. Further, it appears that the relationship between commodity prices and their respective shipping costs has broken down since the beginning of the Covid-19 pandemic. Specifically, a counterfactual analysis shows that the pandemic moved the prices of crude oil and refined petroleum and their costs of shipping in opposite directions. A second counterfactual scenario concerning the impact of Russian oil sanctions shows that there is a high probability that they increased shipping costs.
Presented by: Guglielmo Maria Caporale, Brunel University of London
Estimating the Macroeconomic Effects of Oil Supply News Abstract A common approach for estimating the macroeconomic effects of oil supply news employs SVAR-IV models identified using changes in oil futures prices around OPEC quota announcements as an instrument. We show that the reduced-form oil price innovations, structural shocks, and the instrumental variable in such estimations are all Granger-caused by financial variables, indicating informational deficiencies in the VAR model and contamination of the instrument. To resolve these issues, we incorporate financial indicators into the econometrician’s information set, yielding significantly different results: the shocks contribute less to oil price variation, are less inflationary, and induce a sharper short-term output contraction. The revised results also exhibit greater stability over time and the disappearance of puzzling responses. Notably, we find that oil supply news accounts for a substantial share of S&P 500 volatility. Finally, we identify similar informational deficiencies in other prominent oil-market SVAR models, suggesting this problem is pervasive in oil-market research. Our findings highlight the critical role of financial variables in accurately analyzing the causes and consequences of oil-market shocks.
Presented by: Lorenzo Mori, Bank of Italy
The Anatomy of Asymmetric Oil Supply Shocks Abstract This paper studies how oil supply shocks propagate across the conditional distribution of macroeconomic outcomes. We specify a novel structural Quantile Vector Autoregression (QVAR) that combines external instruments with sign restrictions. The model identifies oil supply news shocks and traces their effects at the median and in the left and right tails of the distribution. We find that the same identified shock, normalized to a 10 % oil price increase, produces qualitatively different macroeconomic responses depending on where in the distribution one conditions: at the 10th percentile, U.S. industrial production expands mildly, while at the 90th percentile it contracts persistently. Impulse responses are very similar across identification strategies: different high frequency instruments, whether complemented or not with sign restrictions, appear to recover the same underlying oil supply disturbance at each quantile. Historical decompositions confirm that this distributional heterogeneity concentrates in well-known oil market episodes, where the quantile dimension reveals tail dynamics that mean-based models cannot capture.
Presented by: Hilde Bjørnland, BI Norwegian Business School
Session: A4: Empirical Macro II March 30, 2026 10:30 to 12:10 Location: Room 515
Session Chair: Ana Beatriz Galvao, Bloomberg Economics, U of Warwick
On the Economic Implications of Political Polarization Abstract Political polarization has two components: (i) intragroup homogeneity (in-group polarization) which manifests as rising agreement and/or ideological alignment within a group of voters and (ii) intergroup heterogeneity (out-group polarization) which manifests as increasing hostility between opposing groups of voters. We propose novel time series measures of political polarization using monthly unit record data from a population-level survey. We then regress key economic expectations on our measures, allowing for the partisan bias and demographic and macroeconomic controls. Results point to potentially important business cycle implications of political polarization. Rising in-group polarization leads to rising consumer optimism about the economic outlook. Rising out-group polarization has the opposite impact, leading to increased pessimism, in particular in times of high polarization. Replicating our regression results for the U.S. is not possible due to a lack of sufficiently long time series. However, U.S. polarization measures show a large rise in in-group polarization for Democratic voters, entirely driven by wide-spread pessimism among Democrats, and a dramatic jump in out-group polarization for all voters since the beginning of 2025.
Presented by: Edda Claus, Wilfrid Laurier University
Monetary policy transmission in the presence of non-linearities Abstract This paper studies the transmission of monetary policy in the euro area using a structural vector autoregression (SVAR) model with endogenous regime switching. The motivation stems from two recent developments: extended periods in which the effective lower bound (ELB) was binding, and an inflation surge that pushed inflation above 10% in some member states. These conditions, combined with the presence of large and volatile shocks, raise questions about the adequacy of linear models for interpreting macroeconomic dynamics. The analysis focuses on two nonlinearities: the ELB constraint and a potential kink in the Phillips curve. The model, estimated on monthly euro area data from 2004 to 2024, reveals that both nonlinearities materially affect the transmission of shocks. During periods of binding ELB, monetary policy is less effective in stabilizing output and inflation. When inflation is high, the slope of the Phillips curve increases, altering the trade-off between output and price stability. The results suggest that these non-linearities played a key role during the recent inflation surge.
Presented by: Alessandro Guarnieri, University of Oxford
Collateral Policy Surprises Abstract Central bank collateral policy specifies which assets banks can pledge as collateral to obtain central bank funding. Despite its importance, little is known about its effect on banks and financial markets. This paper is the first to propose a high-frequency approach to identify collateral policy surprises, using bank stock price changes around Eurosystem collateral policy announcements. Expansionary collateral policy surprises are associated with positive bank excess returns, a decline in common volatility measures, and a decrease in bank CDS spreads. They also compress core-periphery spreads, even when events are unrelated to the collateral treatment of government bonds. These findings indicate that collateral policy influences sovereign bond markets through an uneven transmission channel distinct from both asset purchases and conventional monetary policy.
Presented by: Pia Huettl, DIW Berlin
Extracting Long-Term Market Expectations from Government Bond Yields Abstract Standard affine term structure models imply limited time variation in long-horizon risk-free forward rates, as expectations converge to their unconditional mean. We propose a model with a market-based time-varying endpoint, using OIS forward rates as an unspanned factor to drive short-rate expectations at long horizons. The model filters relevant variation from swap forwards without imposing assumptions on their risk content. Applied to US Treasuries, German Bunds, and UK Gilts, the model produces economically meaningful long-horizon risk-free rates and 30-year term premia, avoiding the prolonged periods of implausibly negative premia implied by conventional models.
Presented by: Ana Beatriz Galvao, Bloomberg Economics, U of Warwick
Session: A5: Macro Theory I March 30, 2026 10:30 to 12:10 Location: Room 516
Session Chair: Fabian Seyrich, Frankfurt School
Prudential Fiscal Stimulus Abstract When government fiscal interventions are predictable, private incentives are distorted. If firms anticipate fiscal stimulus in a crisis, then they might take on excessive risk today. How can fiscal interventions be designed to mitigate such moral hazard problems? And would such interventions be time consistent? We show that fiscal stimulus programmes can be designed to induce precautionary behaviour ex ante from firms, and we label these programmes prudential fiscal stimulus. We demonstrate the theory with a wage subsidy stimulus policy. We show that countercyclical wage subsidies can be welfare improving even in the absence of aggregate demand or labour market externalities. Prudential fiscal stimulus is time inconsistent, but the presence of aggregate demand externalities can bring discretionary policies closer to optimal policies.
Presented by: Alfred Duncan, University of Kent
Corporate Debt Maturity and Business Cycle Fluctuations Abstract Long-term debt is the main source of firm-financing in the U.S. We show that accounting for debt maturity is crucial for understanding business cycle dynamics. We develop a macroeconomic model with defaultable long-term debt and equity adjustment costs. With long-term debt, firms have an incentive to increase leverage in order to dilute the value of outstanding debt. When equity issuance is costly, this incentive helps firms raise more debt through a debt dilution channel and mitigates the decline in net worth through a balance sheet channel, dampening the decline in investment in response to a negative financial shock. Using firm-level data, we estimate equity issuance costs and incorporate our findings into an estimated medium-scale DSGE model. Accounting for debt maturity and the cost of equity financing implies that credit supply shocks emerge as a new driver of business cycle fluctuations.
Presented by: Francesco Ferrante, Federal Reserve Board
Monetary–Fiscal Interactions and the Liquidity Channel of Debt Sustainability Abstract This paper examines how the steady-state debt-to-GDP ratio shapes the transmission of fiscal and monetary policy shocks in a tractable heterogeneous-agent New Keynesian model. When households value government debt for its liquidity services for self-insurance, higher debt levels amplify the adverse fiscal consequences of expansionary government spending shocks. With debt already high, fiscal expansions require the central bank to maintain higher real interest rates for longer to sustain liquid-asset demand and clear bond markets, raising debt servicing costs and reducing fiscal space. By contrast, the transmission of monetary expansions is largely insensitive to steady-state debt levels. The results highlight the crucial role of the liquidity premium and self-insurance motive in shaping the interaction between initial public indebtedness and debt sustainability.
Presented by: Pascal Meichtry, Banque de France
Being and Consciousness: Fiscal Attitudes according to HANK Abstract Attitudes toward fiscal policy differ: fiscal conservatism and fiscal liberalism vary in their willingness to tolerate budget deficits. We challenge the view that such attitudes reflect national preferences. Instead, we offer an economic explanation based on a two-country Heterogeneous Agent New Keynesian model, bringing its implicit political economy dimension to the forefront. We compute the welfare implications of alternative fiscal policies at the household level to assess the conditions under which a policy commands majority support. Whether the majority supports fiscal conservatism or liberalism depends on a country’s debt level, its wealth distribution, and the nature of the economic shock.
Presented by: Fabian Seyrich, Frankfurt School
Session: A6: Time Series - Cointegration March 30, 2026 10:30 to 12:10 Location: Room 517
Session Chair: Tomás del Barrio Castro, University of the Balearic Islands
VARs and Local Projections Equivalence for Impulse Responses: Unit Roots and Multiple Instruments Abstract We show that the equivalence in population between impulse responses in Vector Autoregressions (VARs) and Local Projections (LPs) can be extended to cointegrated unit roots with unrestricted lag structure. We also prove that structural estimation with multiple instruments for multiple endogenous regressors is equivalent to a recursively block-identified Structural VAR, where the block of instruments is ordered first. Simulations and two applications illustrate our results.
Presented by: Alessio Volpicella, University of Pavia
Cointegrated Models for matrix Valued Time-Series Abstract Traditional econometric analyses represent observations as vectors despite the inherent complexity of empirical data structures. When data are organized along dual classification dimensions, a matrix representation provides a more natural and interpretable framework. Building on recent advances in matrix autoregressive (MAR) modeling, this study introduces a novel error correction representation tailored for matrix-structured data. Through comparative analysis with existing methodologies, we demonstrate two critical advancements. First, the proposed model preserves the interpretative foundations of conventional cointegration analysis, with coefficients that explicitly capture dynamics rooted in adjustment toward steady-state positions. Second, in contrast to previous formulations, our error correction framework allows for an equivalent matrix autoregressive representation, preserving the fundamental structure of the data in both specifications. This ensures that the matrix representation reflects an intrinsic characteristic of the data.
Presented by: Massimiliano Caporin, University of Padova
The Effect of Aggregation on Seasonal Cointegration in Mixed Frequency data Abstract Economic time series often show a strong persistency as well as seasonal variations that are appropri ately modelled using seasonal unit root models in addition to deterministic components. In many cases di¤erent variables within a vector time series are driven by identical common trends and cycles leading to cointegration. This paper investigates the consequences for the properties of vector processes when some components are aggregated in time. This may involve moving from a fully observed system that is seasonally cointegrated at a frequency !k = 2 k=S with k = 1;:::;(S 1)=2 where S is the number of seasons per year, to a system with time series sampled at high sampling rate (HSR) observed for S seasons per year and time series with low sampling rate (LSR) observed SA seasons per year, such that SA = S=Q and Q is an integer. The (partial) aggregation has implications on the unit root and cointegration properties: Aggregation potentially shifts the frequency of the unit roots. This may lead to an aliasing e¤ect wherein common cycles to di¤erent unit roots become aligned and cannot be separated any more, in turn impacting cointegrating relations. This paper uses the triangular systems representations in the bivariate case as well as the state space framework (in a general setting) to investigate the e¤ect of aggregation on the unit root properties of multivariate time series. The main results indicate under which assumptions and in which situations the analysis of the integration and cointegration properties of time series with mixed sampling rate relates to the same properties of the underyling data generating process. The results also discuss full aggregation of all components. These results lead to the proposal of an e¤ective econometric strategy for detecting cointegration at the various sampling rates, as is demonstrated in a simulation exercise. Finally an empirical application with monthly data of arrivals and departures of the Mallorca Airport, also illustrate the ndings collected in the present work.
Presented by: Tomás del Barrio Castro, University of the Balearic Islands
Session: A7: Empirical Macro III March 30, 2026 10:30 to 12:10 Location: Room 526
Session Chair: Rodrigo Sekkel, Bank of Canada
Learning from Surprises: Monetary Policy Shocks and Endogenous Gain Learning Abstract This paper examines how monetary policy surprises affect forecast revisions and proposes a behavioral New Keynesian model with a shock-dependent learning mechanism. Using individual data from the Survey of Professional Forecasters and high-frequency shocks, I find that easing surprises lead to upward revisions in inflation forecasts, while tightening surprises primarily lead to downward revisions in growth forecasts. Building on these findings, I extend a New Keynesian model with a shock-dependent learning gain that adjusts based on recent forecast errors and the monetary policy shock. Bayesian estimation shows that both tightening and easing surprises significantly raise the learning gain, with comparable magnitudes. The estimated gain rises during volatile periods such as the late 1970s–early 1980s and the COVID-19 pandemic, but remains low throughout the Great Moderation. The model’s impulse responses reveal a state-dependent transmission. Tightening has stronger and more persistent effects in high-volatility environments. Easing shocks, in contrast, vary across regimes with their response magnitudes ordered differently. Moreover, the learning gain can act as a stabilizer during credible policy shifts, yet can amplify shocks amid heightened uncertainty, suggesting a dual role in policy transmission.
Presented by: Jongho Kim, University of California, Irvine
Chinese monetary policy and text analytics: connecting words and deeds Abstract We propose a novel approach to estimating the People’s Bank of China (PBOC) monetary policy rule. Given China’s complex monetary policy framework, the PBOC’s actions are not well captured by standard monetary policy rules that can be used for its Western counterparts. This calls for new approach able to capture the complexities underlying the PBOC’s observed behavior. This paper examines whether information provided in PBOC’s official statements can help find their policy rule in practice. We use Latent Semantic Analysis to extract content information embedded in PBOC Monetary Policy Reports and examine whether these are significant in explaining Chinese monetary policy actions.
Presented by: Barbara Sadaba, Universidad Diego Portales
When the Fed Reveals Its Hand: The SEP and Monetary Policy Surprises Abstract Recent advances in high-frequency identification of monetary policy shocks have revealed that traditional measures may be contaminated by information and news effects. We contribute to this literature by arguing that the intermittent release of the Summary of Economic Projections (SEP), an important innovation in Fed communications, provides another source of contamination. We develop a theoretical framework showing how SEP releases provide markets with information about the Fed’s internal forecasts, but surprises about the SEP amplify monetary policy surprises. We show that empirical monetary policy surprises are significantly larger during FOMC meetings with SEPs. To identify the effect, we construct a novel SEP surprise measure using a unique Bloomberg survey that captures market participants’ expectations about Federal Reserve views. We then show that SEP surprises explain up to half the variation in monetary policy surprises during SEP meetings. This accounts for essentially all of the differences between SEP and non-SEP meetings, demonstrating that central bank information effects are empirically important and measurable in specific communication contexts.
Presented by: Andrew Martinez, American University
Money Talks: How Domestic and Foreign Monetary Policy Communications Move Financial Markets Abstract We provide new evidence on how domestic and foreign monetary policy communications, beyond rate announcements, affect the financial markets of open economies. Using a high-frequency dataset covering Bank of Canada (BoC) and Federal Reserve (Fed) policy rate decisions, speeches, press conferences, and minutes releases from 1997 to 2023, we document their heterogeneous impact on Canadian markets. While the BoC is the main driver of short-term interest rates, Fed communications exert sizable effects even at the front end of the yield curve—around half as large as those of the BoC—and dominate movements in long-term yields and equity markets. Since BoC communications have little effect on U.S. interest rates, Canadian announcements have a greater impact on the CAD/USD exchange rate as they generate larger changes in the cross-country interest rate differential. Overall, our findings highlight how domestic and foreign communications transmit to different markets, deepening our understanding of cross-border monetary policy linkages.
Presented by: Rodrigo Sekkel, Bank of Canada
Session: A8: Time Series I March 30, 2026 10:30 to 12:10 Location: Room 527
Session Chair: Paulo Rodrigues, Universidade Nova de Lisboa
Stochastic Volatility-in-mean VARs with Time-Varying Skewness Abstract This paper introduces a Bayesian vector autoregression (BVAR) with stochastic volatility-in-mean and time-varying skewness. Unlike previous approaches, the proposed model allows both volatility and skewness to directly affect macroeconomic variables. We provide a Gibbs sampling algorithm for posterior inference and apply the model to quarterly data for the US and the UK. Empirical results show that skewness shocks have economically significant effects on output, inflation and spreads, often exceeding the impact of volatility shocks. In a pseudo-real-time forecasting exercise, the proposed model outperforms existing alternatives in many cases. Moreover, the model produces sharper measures of tail risk, revealing that standard stochastic volatility models tend to overstate uncertainty. These findings highlight the importance of incorporating time-varying skewness for capturing macro-financial risks and improving forecast performance.
Presented by: Ana Skoblar, European Central Bank
An Extended Score-Driven Dynamic Factor Model: Recovering Composite Indicators from the Pandemic Abstract We propose an extended score-driven (ESD) dynamic factor model (DFM) that accommodates non-Gaussian innovations, nonlinear factor dynamics, and time-varying volatility. The model nests in a special case the state-of-the-art state-space DFM and score-driven DFM, closing the gap between these two classes of models. Empirically, our ESD-DFM proves useful for working with COVID-19–era observations, which have posed substantial challenges for macroeconomic modeling. For instance, while the Federal Reserve Bank of Philadelphia suspended publication of its leading index due to pandemic-related anomalies, our model remains robust to such extreme observations and enables reliable computation of the index. We further apply the ESD-DFM to The Conference Board’s Coincident and Leading Economic Indices (CEI and LEI). The model substantially outperforms both Kalman filter–based and standard score-driven DFMs. Notably, when indices are constructed from the estimated factors, the unprecedented divergence between the CEI and LEI observed during the post-pandemic period disappears: although the reconstructed LEI declines in 2022, it resumes an upward trajectory from late 2023 through mid-2025.
Presented by: Mariia Artemova, Erasmus University Rotterdam
Leverage at different time scales: A stochastic volatility perspective Abstract This paper presents an improved approach to modeling leverage dynamics by incorporating a heterogeneous autoregressive structure within a stochastic volatility framework. Our modification not only captures the propagation of leverage across multiple time horizons but also improves the efficiency of the estimation, simplifies the model, and improves its interpretability. Through simulation studies, we demonstrate that the model effectively replicates the leverage effect and its propagation over time. Additionally, we employ data cloning for parameter estimation, which enhances accuracy and computational efficiency for finite samples. An empirical analysis of several financial return series confirms robust in- and out-of-sample performance of the model and underscores its practical relevance for volatility forecasting and risk management. These findings position the model as a valuable tool for understanding leverage dynamics and their impact on financial stability and asset pricing.
Presented by: Helena Veiga, University Carlos III Madrid
Structural Breaks in Conditional Tail Risk Abstract This paper develops a structural break testing procedure for covariate-dependent conditional tail indices. The tail index, specified as $\alpha(X_t,\beta)=exp(X_t'\beta)$, governs the heaviness of the upper tail of the conditional distribution and determines the probability of extreme outcomes. A structural break corresponds to a regime shift in the relationship between covariates and tail behaviour. Inference is based on tail-selected observations obtained through a Pareto-motivated exceedance rule and transformed into tail space using the Nicolau–Rodrigues log–log mapping, which yields an approximately linear representation for extremes. These features give rise to a triangular array with irregular spacing, regime-specific thresholds, and nonlinear transformations, implying that classical sup-$F$ critical values (Andrews, 1993) are invalid. The paper derives the asymptotic properties of the break-point estimator and the tail-index estimators under this nonstandard environment, establishes consistency, characterises variance inflation due to break-date uncertainty, and develops the limiting behaviour of the supremum-$F$ statistic. Because the limit distribution is non-pivotal, simulation-based critical values are required. Monte Carlo experiments show high power against breaks in tail behaviour and illustrate distinct finite-sample features, including systematic bias in the estimated break location induced by regime-dependent exceedance frequencies. The methodology provides a theoretically grounded and practically relevant tool for detecting regime shifts in the tails of conditional distributions, with applications to financial returns, insurance losses, climate extremes, and macroeconomic risk.
Presented by: Paulo Rodrigues, Universidade Nova de Lisboa
Session: B1: Empirical Macro IV March 30, 2026 13:30 to 15:10 Location: Room 528
Session Chair: Francesco Furlanetto, Norges Bank
Rare Disasters and Consumption Smoothing: Testing a State-Dependent Savers-Spenders Framework Abstract Rare macroeconomic disasters---such as pandemics, wars, and depressions---show sharp declines in real GDP and consumption, often alongside failing or impaired financial institutions and markets. These conditions may weaken households' ability to smooth consumption, but the extent remains unclear. We investigate this using a savers-spenders framework, where the share of rule-of-thumb spenders is larger during crises. This reduces the impact of expected income growth on the average propensity to consume. As such, during disasters, given changes in the consumption-income ratio reflect larger revisions in those expectations. Consequently, the model implies that the consumption-income ratio should be a stronger predictor of future income growth in disaster periods. Using state-dependent panel predictive regressions on annual data from 17 developed economies since 1870, we confirm this prediction. Our estimates suggest that at least one-quarter of households switch to hand-to-mouth behavior during rare disasters, implying a significant reduction of consumption smoothing.
Presented by: Lorenzo Pozzi, Erasmus University Rotterdam
A Millennium of Business Cycles: Evidence from the UK Abstract We study macroeconomic fluctuations in the United Kingdom over seven centuries (1271–2022) using a time-varying VAR with stochastic volatility. We identify the main business cycle shock each year as the innovation explaining the largest share of short-term output variance. Before 1900, shocks exhibit a stagflationary, supply-driven pattern, while post-1900 shocks are demand-driven, raising both output and inflation. Output volatility declines over time, peaking in the seventeenth century. Monetisation had large real effects in the sixteenth and seventeenth centuries, shifting to inflationary impacts thereafter. Our results highlight how business cycle dynamics evolve with institutional, monetary, and structural transformations.
Presented by: Leonardo Ferreira, Central Bank of Brazil
Technology Spillovers from the Final Frontier: A Long‐Run View of U.S. Space Innovation Abstract Space activities generate significant economic benefits. This paper quantifies these effects by modeling both short-run business-cycle and long-run growth effects driven by space sector activities. We develop a model in which technologies evolve through both a dedicated R&D sector and spillovers from space-sector innovations. Using U.S. data from the 1960s to the present day, we analyze patent grants to distinguish between space and core sector technologies. By leveraging the network of patent citations, we also examine the evolving dependence between space and core technologies over time. Our findings highlight the positive impact of the aerospace sector on technological innovation and economic growth, particularly during the 1960s and 1970s.
Presented by: Aldo Paolillo, Free University of Bozen-Bolzano
Inequality, Amplification, and Business Cycles Abstract We quantify the connection between inequality and business cycles using SVAR models, estimated with aggregate and cross-sectional data. We rely on detailed micro data on income and consumption from Norway. We find that inequality does not substantially amplify cyclical fluctuations. The primary source of this limited amplification is cyclical precautionary saving behavior. Savers reduce only to a very limited extent their consumption to insure themselves against the idiosyncratic risk of large income drops, which rises in recessions. We compute counterfactuals in which we switch off inequality fluctuations and we show that results are affected only to a limited extent.
Presented by: Francesco Furlanetto, Norges Bank
Session: B2: Time Series III March 30, 2026 13:30 to 15:10 Location: Room 513
Session Chair: Renee Fry-McKibbin, Australian National University
Estimation of Non-Gaussian SVAR Using Tensor Singular Value Decomposition Abstract This paper introduces a tensor singular value decomposition (TSVD) approach for estimating non-Gaussian Structural Vector Autoregressive (SVAR) models. The proposed methodology applies to both complete and partial identification of structural shocks. The estimation procedure relies on third- and/or fourth-order cumulants. We establish the asymptotic distribution of the estimator and conduct a simulation study to evaluate its finite-sample performance. The results demonstrate that the estimator is highly competitive in small samples compared to alternative methods under complete identification. In cases of partial identification, the estimator also exhibits very good performance in small samples. To illustrate the practical relevance of the procedure under partial identification, two empirical applications are presented.
Presented by: Dalibor Stevanovic, Université du Québec à Montréal
Local Projections with Free-Knot Splines Abstract This paper introduces a nonparametric estimator for functions defined over discrete, naturally ordered supports and applies it to impulse response function estimation within the local projections (LP) framework. The estimator hypothesizes that the function admits a lower-dimensional representation by projecting it onto a basis of piecewise continuous functions. To account for uncertainty about the appropriate dimensionality in estimation and inference, the approach employs Jackknife model averaging across specifications with different dimensions, yielding the Averaged Projected Local Projections (APLP) estimator. Extensive simulations show that APLP has lower variance than the standard LP estimator while introducing little bias, and achieves similar coverage with shorter confidence intervals; the median APLP interval is about 20% shorter. Two applications demonstrate that APLP improves interpretability by smoothing estimates across horizons and tightening confidence bands.
Presented by: Eva Janssens, University of Michigan
Tensor Factor Model with CP Structure Abstract This paper studies the statistical inference in high-dimensional tensor factor models relying on the Canonical Polyadic (CP) decomposition. We show that the factors and their loadings in our models, which are identified up to a trivial permutation of indices and scale changes, can be estimated by a simple alternating least squares (ALS) algorithm. The identified components of our estimators for the factors and their loadings are indeed consistent and asymptotically normal under fairly general conditions. In addition, we develop a test and obtain its asymptotics for determining the number of factors based on the ratio of generalized singular values.
Presented by: Yoosoon Chang, Indiana University
Trump Tariffs and Global Market Interdependence Abstract This paper explores the effect of tariff policy uncertainty on components of U.S. equity markets and on spillovers between the U.S. and foreign equity and FX markets, commodity markets, cryptocurrencies and the VIX. Shocks to tariff policy uncertainty generate large but short-lived and quickly reversed reactions in U.S. equity markets, with effects small relative to fundamentals. Higher order comonent change tests between markets show tariff policy uncertainty transmitting to global equities particularly in Asia, and to FX markets in Canada, Mexico, and Korea. The major alternative assets affected include the GSCI, gold, silver, and oil, with smaller effects on bitcoin. A striking result is a reduction in the VIX. Market correlations generally remain stable but spillovers through risk measured through cokurtosis increase. Other key resource markets remain resilient.
Presented by: Renee Fry-McKibbin, Australian National University
Session: B3: Empirical Macro V March 30, 2026 13:30 to 15:10 Location: Room 514
Session Chair: Alessandro Franconi, Banque de France
From Tariffs to Shelves: Tariff Effects on U.S. Retail Prices Abstract The speed and extent to which tariffs build pressure on consumer prices remains an ongoing and important debate. This paper examines tariff-related retail price changes using a detailed item-level retail spending dataset combined with information on the countries where products are produced. Our headline index closely matches the official PCE deflator for food and beverages and correlates strongly with other inflation measures. We then disaggregate our price index to measure inflation for products based on their country of origin. We find that (i) tariff effects have been greatest for goods produced in China, (ii) tariff pass-through to consumers between April and September 2025 has been approximately 20% for goods imported from China and minimal for goods imported from other countries, (iii) the current extent of tariff pass-through translates into an 8\% year-over-year increase in consumer prices for Chinese goods, and (iv) lower pass-through from tariffs to retail prices suggests the peak of tariff pass-through has not yet materialized and price pressures are likely to continue building gradually rather than showing up as a one-time price spike.
Presented by: Sinem Hacioglu Hoke, Federal Reserve Board
Heterogeneous beliefs and the central bank reaction function Abstract We estimate the perceived monetary policy rule for a broad set of countries by leveraging on the heterogeneity in the responses of Consensus Economics survey participants. We assume that survey respondents perceive central banks to act following a stylised Taylor Rule, but disagree on future inflation and output gaps, as well as the coefficients in the Taylor Rule. Combining survey respondents' expectations on interest rate with the macroeconomic outlook, we estimate not only the policy rule perceived by the market, but also their disagreement on the Taylor Rule. We rely on a Bayesian state-space model, which enables us to document the time variation in the coefficients and its heterogeneity across countries. Finally, we attribute the disagreement on policy rates into four different sources: disagreement about the macroeconomic outlook, disagreement about the monetary policy rule, disagreement about monetary policy inertia, and the interaction between macro and rule disagreement.
Presented by: Marco Lombardi, Bank for International Settlements
Import Tariffs and the Systematic Response of Monetary Policy Abstract We estimate the macroeconomic effects of U.S. import tariff shocks, using several distinct tariff measures and identification approaches. We find that tariff shocks reduce output but increase consumer prices. Monetary policy partially accommodates this shock with a transitory policy easing. To quantify the dependence on systematic monetary policy, we construct counterfactuals using identified monetary policy shocks. This avoids specifying a full structural model, making the results robust against model misspecification. When monetary policy strictly stabilizes inflation, the output contraction is 32% larger at the trough. In contrast, strict output stabilization implies a sizable sacrifice of price stability, with the peak inflation effect doubling.
Presented by: Alessandro Franconi, Banque de France
Session: B4: Macro Theory - Inflation March 30, 2026 13:30 to 15:10 Location: Room 515
Session Chair: Laura Gáti, ECB
Shaping Inflation Inattention through Inequality Abstract This paper investigates how consumption inequality shapes inflation inattention. We develop a Two-Agent New Keynesian (TANK) model with noisy information, where Ricardian households endogenously adjust their attention to inflation based on observable consumption disparities with HtM households. Empirical evidence from a Structural VAR for the U.S. supports this mechanism as rising inequality is associated with a significant decline in inflation estimation errors. Theoretically, incorporating this behavioral feedback improves the responsiveness and stability of inflation expectations, particularly under monetary and cost-push shocks. Moreover, this behavioral mechanism shapes the impact of optimal monetary policy, enhancing the central bank’s ability to stabilize inflation, albeit at the expense of a more pronounced decline in economic activity.
Presented by: Carolina Serpieri, University La Sapienza
Interest rates and inflation in New Keynesian models: the role of expectations Abstract Do higher interest rates lead to higher or lower inflation in New Keynesian models? This paper shows that the answer to this question revolves around the role played by expectations and is related to the effect that exogenous shocks have on the endogenous variables. When expectations do not reverse such relationship, as posited in the structural model, exogenous increases in interest rates lead to lower inflation, and positive demand and cost-push shocks increase inflation. Such equilibria exist and correspond to the forward solution if a version of the Taylor principle is satisfied, which depends on the persistence of the shocks.
Presented by: Michele Berardi, University of Manchester
Households’ Inflation Expectations and Consumption in Macroeconomic Models: The Role of the Real Income Channel Abstract In the standard New Keynesian (NK) framework, an increase in households' inflation expectations raises consumption. This conventional result rests on strong general equilibrium effects and the assumption that households do not perceive expected real income losses when inflation expectations rise. In this paper, I disentangle the underlying economic mechanisms and show that the consumption response can easily turn negative once empirically relevant features are taken into account. I decompose the total effect into an intertemporal substitution channel and a negative real income channel, under the empirically supported assumption that inflation expectations do not fully pass through to nominal wage expectations. In the Representative Agent NK model, consumption still increases because households also receive profits, which offset expected real wage decline. By contrast, in a stylized Heterogeneous Agent NK model, the total effect may turn negative if the profit channel is dampened and the disconnect between inflation and nominal wage expectations is sufficiently strong.
Presented by: Frantisek Masek, Czech National Bank
Away From FIRE Abstract Departures from full information rational expectations (FIRE) pose a challenge for macroeconomics: how to reconcile sluggish expectations dynamics with robustness to policy counterfactuals? This paper proposes two features for models away from FIRE to resolve this dilemma. The first is that the expectations process should correspond to the process that generates the data in the actual model at hand, not to what the DGP would be under rational expectations. Second, the agents need to use information optimally given the information structure and their priors. A New Keynesian model with these features is shown to match not only the sluggishness of expectations, but also their pattern of delayed overshooting that models away from FIRE have struggled to match. Furthermore, the model features hump-shaped responses of inflation to shocks, resolving a long-standing issue in macroeconomics around producing hump shapes without recourse to exogenous modeling assumptions lacking microfoundations.
Presented by: Laura Gáti, ECB
Session: B5: Macro Theory - Monetary Policy I March 30, 2026 13:30 to 15:10 Location: Room 516
Session Chair: Martin Arazi, Bank of England
The Enduring Effects of Unconventional Monetary Policy Abstract This paper investigates unconventional monetary policy transmission linked to credit-financed endogenous growth. Using a dynamic general equilibrium model where financial conditions and growth are interwoven, we study the aggregate aftermath to quantitative easing (QE), forward guidance (FG) and negative interest rate policy (NIRP). All expansionary interventions operate through the credit channel, influencing banks’ conditions and fostering economic growth. In calibrated scenarios, FG and NIRP emerge as optimal options for sustained productivity increases. These policies influence TFP and output, easing the ZLB constraint. While QE boosts TFP persistently, its quantitative impact is relatively lower, with a more short-lived effect on output.
Presented by: Giulio Tarquini, Sapienza University of Rome
The Role of Shadow Banks in Fiscal Transmission Abstract U.S. data shows that domestic financial intermediaries, especially shadow banks, are major holders of Treasuries. Do the Treasury holdings of these intermediaries provide a new transmission channel of expansionary fiscal policy? Empirical evidence based on well-defined Treasury supply shocks documents differential responses by commercial and shadow banks. Using a quantitative dynamic general equilibrium model including two types of financial intermediaries issuing money-like claims and providing capital, I find that an increased Treasury supply reallocates productive capital from commercial to shadow banks. This reallocation is caused by shadow banks reducing their Treasury holdings to absorb claims in productive capital sold by commercial banks. An expanding supply of government debt increases the convenience yield on intermediary debt, also incentivizing leverage expansion. This mechanism is stronger for shadow banks, allowing them to expand. In addition, it also erodes the net worth of intermediaries, with a more persistent impact on commercial banks. The liquidity provision falls immediately, and lending follows in the medium run. Almost half of the welfare loss is attributed to disruption in the financial sector.
Presented by: Fabian Wassmann, Nova School of Business and Economics
Lighting the Shadows: Central Bank Policy Transmission through Banks and Non-Banks Abstract Non-bank financial intermediaries (NBFIs) account for an increasingly large share of credit provision and play a central role in quantitative easing (QE) and tightening (QT). We develop a DSGE model with banks, NBFIs and a central bank interacting through money, bond and reserve markets to study how monetary transmission changes as the size of NBFIs expands. We estimate the model on UK data to evaluate key parameters in the financial sector. Conventional interest-rate policy remains effective for output and inflation, although the composition of credit shifts from bank lending toward bond issuance. In contrast, QE/QT transmission is strongly affected: a larger NBFI sector amplifies investment responses via the corporate bond market but dampens consumption effects. The results underscore that monetary policymakers retain rate control regardless of financial structure, but that the effectiveness of balance-sheet policies depends crucially on the composition and behaviour of the non-bank sector.
Presented by: Jacob Stevens, Bank of England
Different Unconventional Monetary Policies, Different Stories? A HANK Perspective Abstract We develop a two-asset HANK model to examine the effects of various unconven tional monetary policies, including Quantitative Easing, Operation Twist, and Liquid ity Facilities. Our analysis focuses on the macro-financial implications of these policies, as well as their impact on wealth and income inequality through shifts in portfolio com position between liquid and illiquid assets. The model features a financial sector with mutual funds and banks that invest in short- and long-term bonds, reserves, and credit to firms. We find that Operation Twist has weaker effects compared to standard QE interventions, although all strategies are qualitatively similar in their impact on output, inflation, and the composition of liquid and illiquid assets.
Presented by: Martin Arazi, Bank of England
Session: B6: Time Series III March 30, 2026 13:30 to 15:10 Location: Room 517
Session Chair: Elmar Mertens, European Central Bank
Origins and Nature of Macroeconomic Instability in Vector Autoregressions Abstract For a general class of dynamic and stochastic macroeconomic models, we show that (i) non-linearity in economic dynamics is a necessary and sufficient condition for time-varying parameters (TVPs) in the VARMA process followed by observables, and (ii) all parameters’ time-variation is driven by the same, typically few sources of stochasticity: the shocks in the macroeconomic model. Motivated by these results, we model a set of macroeconomic and financial variables as a TVP-VAR with a factor-structure in TVPs. This reveals that most instabilities are driven by a few factors, which comove strongly with measures of macroeconomic uncertainty and the contribution of finance to real economic activity. Furthermore, relative to the TVP-VAR with TVPs evolving as independent random walks, our model delivers an improved forecasting performance.
Presented by: Marko Mlikota, Geneva Graduate Institute
Moderate Time Varying Parameter VARs Abstract This article proposes a new parametric approximation for time-varying parameter models particularly suited for slowly evolving dynamics. Specifically, our ``moderate'' time-variation approach rewrites the transition equation of a coefficient in terms of B-splines and a low-dimension dynamics, reducing the number of parameters to a scale factor equal to the number of spline knots. Using both simulated and real macroeconomic datasets of varying dimensions, we apply the proposed techniques to Vector Autoregressive models. We show that our methodology not only delivers interpretations consistent with benchmark models, but also improves forecast accuracy and computational times.
Presented by: Alessandro Celani, Örebro University
Let the Tree decide: FABART A Non-Parametric Factor Model Abstract This article proposes a novel framework that integrates Bayesian Additive Regression Trees (BART) into a Factor-Augmented Vector Autoregressive (FAVAR) model to fore- cast macro-financial variables and examine asymmetries in the transmission of oil price shocks. By employing nonparametric techniques for dimension reduction, the model captures complex, nonlinear relationships between observables and latent factors that are often missed by linear approaches. A simulation experiment comparing FABART to linear alternatives and a Monte Carlo experiment demonstrate that the framework accurately recovers the relationship between latent factors and observables in the pres- ence of nonlinearities, while remaining consistent under linear data-generating pro- cesses. The empirical application shows that FABART substantially improves forecast accuracy for industrial production relative to linear benchmarks, particularly during periods of heightened volatility and economic stress. In addition, the model reveals pronounced sign asymmetries in the transmission of oil supply news shocks to the U.S. economy, with positive shocks generating stronger and more persistent contractions in real activity and inflation than the expansions triggered by negative shocks. A similar pattern emerges at the U.S. federal state level, where negative shocks lead to mod- est declines in employment compared to the substantially larger contractions observed after positive shocks.
Presented by: Sofia Velasco, Banco de España
Entropic Tilting of Forecasts to SPF Histograms: Analytics & Applications Abstract This paper develops a new, direct approach to entropic tilting of model-based predictive distributions to match histogram forecasts provided in the U.S. Survey of Professional Forecasters (SPF). We focus on tilting to histogram probabilities directly, rather than to moments of fitted distributions. We reformulate the single-histogram tilting problem and derive a novel analytic characterization for the multiple-histogram case, with iterative solutions via Iterative Proportional Fitting. Application to quarterly real-time forecasts of major macroeconomic aggregates from a Bayesian vector autoregression with time-varying volatility shows that tilting to SPF histograms significantly improves on the model's baseline forecasts, particularly during periods around the Great Recession and the COVID-19 pandemic.
Presented by: Elmar Mertens, European Central Bank
Session: B7: Time Series IV March 30, 2026 13:30 to 15:10 Location: Room 526
Session Chair: Tibor Pal, University of Salerno
Reservoir-driven parameters Abstract Inspired by the reservoir computing paradigm in machine learning, we introduce a new class of observation-driven time series models in which the time-varying parameters are reconstructed from a high-dimensional basis of randomly generated states, called reservoirs. This design can be viewed as a recurrent neural network where only the output layer is trained, while the internal weights are randomly sampled and remain fixed. Compared to existing observation-driven models, the proposed method offers several key advantages: a) it can approximate in L^p sense any causal time-invariant filter; b) the invertibility condition is feasible and invariant across model specifications; c) the gradient and Hessian of the likelihood function are available in closed form, making the method computationally efficient and well-suited for multivariate applications. The inferential theory of the quasi-maximum likelihood estimator is established by deriving conditions for consistency and asymptotic normality. Using simulated and empirical data, we show that the method outperforms alternative observation-driven specifications in recovering highly nonlinear parameter dynamics, including jumps, structural breaks, and chaotic behavior.
Presented by: Giuseppe Buccheri, University of Verona
Noise-cancelling location models Abstract We propose a new class of observation-driven location models with a filter that discards new information when deemed irrelevant. The forcing variable in the filtering equation is scaled by a smooth indicator function that is zero when the one-step prediction error is small in absolute magnitude. We establish consistency and asymptotic normality of the maximum likelihood estimator of static model parameters, and demonstrate the usefulness of the new noise-cancelling filter in a US unemployment rate forecasting study. Results suggest that the new filter can improve the accuracy of forecasts compared to standard observation-driven counterparts.
Presented by: Jannik Steenbergen, Aarhus University
Estimation and inference in models with multiple behavioural equilibria Abstract We develop estimation and inference tools for a stylized macroeconomic model with potentially multiple behavioural equilibria in which agents form expectations using a constant-gain learning rule. Under fairly general assumptions, we prove geometric ergodicity of the underlying process. We propose a non-linear least squares estimator for the structural parameters and establish strong consistency and asymptotic normality. We discuss inference for structural parameters and propose uniform confidence bands for the equilibria. At points where equilibria merge, we characterise mixed convergence rates and associated Bernoulli limiting distribution, and we outline procedures that remain valid under weak identification of the learning gain. Monte Carlo simulations verify the the findings in finite samples.
Presented by: Alexander Mayer, Erasmus University Rotterdam
Multiple Quantile Dynamics of the US Phillips Curve with Time-Varying Parameters Abstract This paper studies the dynamics of the conditional distribution of the US inflation motivated by a hybrid New Keynesian Phillips Curve framework. Inflation is modeled using a dynamic multiple-quantile specification that jointly characterizes a predefined set of conditional quantiles. To accommodate time-varying and quantile-specific parameters, we introduce a smoothed dynamic multiple quantile model that allows parameters to vary over time and across quantiles. The empirical analysis covering US quarterly data from 1961Q1 to 2025Q2 reveals state-dependent nonlinearities in the Phillips curve slope and in the relative dependence on inflation inertia and expectations weight. The results indicate an asymmetric, hump-shaped Phillips curve slope across the conditional inflation distribution with pronounced time variation. During the 1970s, the estimated increase in the weight on inflation expectations is consistent with unanchored expectations that raised inflation and contributed to the high-inflation regime of the Great Inflation. By contrast, the collapse of the output gap–inflation trade-off in the lower tails with strong inflation persistence following the Global Financial Crisis and during the 2020 COVID-19 pandemic provides a plausible explanation for the missing disinflation during these episodes.
Presented by: Tibor Pal, University of Salerno
Session: B8: Empirical Macro VI March 30, 2026 13:30 to 15:10 Location: Room 527
Session Chair: Danilo Cascaldi-Garcia, Federal Reserve Board
Monetary Transmission under Stress: Evidence from the Euro Area Abstract This paper studies how systemic financial stress shapes the transmission of monetary policy in the euro area. Using the Composite Indicator of Systemic Stress (CISS) as a measure of financial fragility, I estimate state-dependent local projections to compare the effects of monetary policy shocks in calm and stressed periods. The results reveal a clear disconnect between financial and real transmission. On the financial side, yields and bank lending rates react more strongly under stress—an amplification of price responses. On the real side, industrial production, investment, and consumption respond less, showing that financial stress weakens the pass-through to activity. Systemic stress thus shifts monetary transmission toward financial markets and away from the real economy. I show that this asymmetry extends to the underlying drivers of financial reactions: decomposing yields into expectations and term-premium components assesses whether the stronger financial responses under stress reflect changes in policy expectations or shifts in risk premia. State-dependence also arises along the shock’s sign, with contractionary monetary policy shocks displaying markedly stronger asymmetries than expansionary ones, consistent with nonlinear transmission through credit and risk-taking channels. Differences in fiscal space, banking resilience, and financial structures further shape how euro area economies absorb monetary shocks under stress, highlighting that financial fragility amplifies price effects and contributes to heterogeneity in transmission across countries.
Presented by: Jonas Hölz, BI Norwegian Business School
Banks' Funding Structure and Pass-Through in the Euro Area Abstract This paper investigates the interest rate pass-through of monetary policy as for the euro area by focusing on the role of banks’ funding structure. We estimate the interest rate pass-through for household deposits and loans to non-financial corporations by using bank-level balance sheet data. In doing so, we interact the response of lending and deposit rate with characteristics of the funding structure, and show that banks with more stable funding structure tend to be less responsive to policy changes.
Presented by: Juan Figueres, European Central Bank
Quantifying Deregulation and Its Economic Effects: A Large Language Model Approach Abstract We construct a text-based index of deregulation that identifies all major deregulatory episodes spanning 1930-2024. Our index captures announced and anticipated deregulatory actions in real-time at the moment it becomes salient to economic agents, making it ideal for studying dynamic responses. Using this new measure, we find that deregulation is pro-cyclical and is associated with increases on productivity, investment, hours worked, and GDP, with effects peaking around two years after the shock. These effects operate through channels including productivity improvement, increased business and consumer confidence, and stock market appreciation, with mild deflationary pressures resulting in muted monetary policy responses. In line with the idea that deregulation increases entrepreneurial experimentation, we find deregulation amplifies both tails of the GDP and investment distributions and increases the VIX in the medium term, representing a trade-off between near-term stimulus and higher economic uncertainty.
Presented by: Danilo Cascaldi-Garcia, Federal Reserve Board
Session: C1: Empirical Macro - Inflation II March 30, 2026 15:40 to 17:20 Location: Room 528
Session Chair: Nobuhiro Abe, Bank of Japan
Geopolitical Risk and Inflation: The Role of Energy Markets Abstract Geopolitical shocks are not all alike -- different classes of geopolitical shocks can have different macroeconomic implications, particularly on inflation. This paper exploits the comovement between the Geopolitical Risk Index (GPR) developed by Caldara and Iacoviello (2022) and oil prices across major geopolitical events to disentangle two types of geopolitical shocks within a structural VAR model for the US economy. The VAR estimates suggest that geopolitical shocks associated with disruptions in energy markets are on average inflationary and contractionary. In contrast, geopolitical shocks associated with macroeconomic developments that are unrelated to energy markets are on average deflationary and contractionary. To validate this interpretation, the paper exploits the heterogeneity across sectoral output and prices of the US economy to show that a sector’s response to a geopolitical shock depends on its energy intensity. Sectors characterized by higher energy intensity are subject to larger output losses and price increases in response to geopolitical energy shocks.
Presented by: Marco Pinchetti, Banque de France
What Drives Distribution of Price Changes? Abstract In this paper, we investigate the potential factors influencing both the price fluctuation rates and the distribution of individual items in Japan's Consumer Price Index (CPI), as well as their relationship with underlying inflation, using a time series model. Specifically, we decompose the price changes of individual items into macroeconomic factors and item-specific factors using principal component analysis. Additionally, we further decompose the macroeconomic factors into four types of shocks using a factor-augmented vector autoregressive (FAVAR) model. Our findings indicate that macroeconomic factors have a more persistent impact on individual price changes compared to item-specific idiosyncratic factors. Among the macroeconomic factors, growth expectation shocks are found to have a lasting impact on a wide range of items. These shocks may have served as a primary factor contributing to the low inflation experienced since the late 1990s. Furthermore, we find that, in addition to supply shocks, growth expectation shocks contributed to upward pressure on prices in 2024. Our results suggest that examining the factors driving the distribution of price changes is a useful approach for understanding the movements of underlying inflation.
Presented by: Nobuhiro Abe, Bank of Japan
Session: C2: Empirical Macro, Time Series & Forecasting March 30, 2026 15:40 to 17:20 Location: Room 513
Session Chair: Robinson Kruse-Becher, University of Hagen
Forecasting the Labor Market: How Useful are Detailed Online Job Postings? Abstract We construct a structured dataset of Canadian online job postings from Indeed, matched to industry and occupation classifications, to assess their usefulness for real-time labor market forecasting. Building on prior work that showed potential of online job postings for nowcasting vacancies, we extend the analysis to examine cross-sectoral predictive relationships and forecasts of other labor market indicators such as employment and unemployment. We also explore whether predictive patterns align with sectoral input-output linkages. Our results shed light on the value of online postings as a timely source of labor market information.
Presented by: Tatjana Dahlhaus, Bank of Canada
Model selection confidence sets for ARMAX models with applications to energy demand data Abstract This paper introduces the Model Selection Confidence Set (MSCS) methodology for univariate time series models involving autoregressive and moving average components. Rather than relying on a single model selected by an arbitrary criterion, the MSCS identifies a set of models that are statistically indistinguishable from the true data-generating process at a given confidence level. The size and composition of this set reveal crucial information about model selection uncertainty: in noisy data scenarios, many models may be included, whereas in more informative cases, the set is smaller and more insightful. By examining Lower Boundary Models within the MSCS framework, we quantify the importance of individual predictors, offering deeper insights into time series dynamics. Motivated by the inherent uncertainty in model selection for electricity demand analysis, we apply the MSCS methodology to study its key drivers and alternative plausible explanations.
Presented by: Piersilvio De Bortoli,
Households’ Macroeconomic Beliefs: The Role of Education Abstract We investigate how education shapes agents' macroeconomic beliefs by surveying households on their perceptions and forecasts of inflation, unemployment, mortgage rates, and stock prices. Our findings unveil significant differences between highly educated and less educated households. Highly educated respondents form beliefs consistent with the existence of a monetary policy trade-off between inflation and unemployment, whereas less-educated households adopt a "supply-side" perspective. When exposed to vignette-based scenarios simulating monetary policy shocks, highly educated individuals adjust their beliefs and consumption-saving decisions in line with intertemporal substitution and textbook economic models. In contrast, less-educated respondents often retain pre-existing beliefs or revise them using non-standard mental models. Moreover, highly educated households primarily rely on formal education and newspapers for economic information, while less-educated households are more influenced by social media. These findings point to the need to model education-related heterogeneity and communicate policy targets and decisions in a simplified manner to reach different socio-economic groups.
Presented by: Eleonora Granziera, Norges Bank
EU ETS Market Expectations and Rational Bubbles Abstract Serious concerns about a price bubble in the European Union Emissions Trading System (EU ETS) emerged during its third trading period, as allowance prices rose sharply and several studies attributed the surge to rational bubbles. We reassess this claim using an expectations-based test that exploits futures prices and thus avoids specifying a fundamental value. Importantly, we show that neglecting risk premia within this framework can generate spurious evidence of bubbles. We therefore develop a testing approach that remains valid in the presence of a dynamic risk premium and is robust when the underlying fundamental is unitroot or mildly explosive. Using weekly spot and futures data from 2013 to 2023, we find that the explosive price dynamics are not attributable to rational bubbles. Instead, the evidence is consistent with shifting expectations of future allowance scarcity rather than speculative bubbles.
Presented by: Robinson Kruse-Becher, University of Hagen
Session: C3: Empirical Macro - Fiscal Policy March 30, 2026 15:40 to 17:20 Location: Room 514
Session Chair: Marco Lorusso, University of Perugia
Fiscal stimuli: Monetary versus Fiscal Financing Abstract We investigate the use of money supply issued by the central bank to support expansionary fiscal interventions. We develop and estimate a New Keynesian model using US data for the sample 1960Q1 - 2019Q4. We conduct a quantitative counterfactual analysis to assess the effects of a fiscal stimulus that does not result in an increase in public debt, as it is financed by money supply. Our impulse response analysis indicates that both increases in government spending and transfers that are monetary financed have positive effects on private consumption, investment and output. However, the expansionary impact of monetary-financed fiscal shocks comes at a cost: an increase in inflation. Our sub-sample analysis indicates that monetary-financed fiscal stimuli would have had a greater positive impact on the economy during the Great Moderation. Lastly, we find that as the debt burden increases, the positive effects of a monetary-financed fiscal stimulus diminish.
Presented by: Claudia Udroiu, Newcastle University
Fiscal Multipliers and Political Fragmentation Abstract This paper provides novel empirical evidence on how political fragmentation shapes the fiscal transmission mechanism. Using data from 16 OECD countries (1978--2019) and narrative accounts to identify exogenous fiscal interventions, we show that when political fragmentation is high, the fiscal GDP multiplier is significantly lower. The multiplier is above unity and relatively stable over time when fragmentation is low, but generally well below unity when fragmentation is high. We show that interventions are comparable across states and argue that a conditional confidence channel helps explain our findings: only in low-fragmentation periods do fiscal interventions boost household and business confidence, translating into stronger consumption and investment responses.
Presented by: Marvin Noeller, RWI - Leibniz Institute for Economic Research
Estimating macroeconomic Effects of Government Spending Shocks Abstract Building on the identification strategy proposed by Auerbach and Gorodnichenko (2012), this paper introduces a novel approach to identifying fiscal policy shocks. We exploit potential asymmetries in the information sets of professional forecasters and the central bank to construct an instrument that is orthogonal to both sources of information. We develop and test several alternative specifications of this instrument, comparing their empirical performance to identify the most reliable one for isolating exogenous fiscal innovations. Furthermore, we investigate the role of nonlinearities in the propagation of such shocks, assessing how fiscal disturbances transmit through the economy in a potentially asymmetric manner. Finally, we conduct an extensive set of robustness checks to validate our findings and assess the stability of the estimated fiscal multipliers.
Presented by: Giacomo Porcellotti, Università di Torino
Fiscal Policy and the Role of Immigration Abstract This paper proposes a macroeconomic framework to evaluate how immigration affects fiscal financing and public debt in Canada for the sample period 1997:Q1-2023:Q3. We provide a SVAR identification scheme that allows us to evaluate the impact of immigration shocks on several fiscal variables of interest. This analysis shows that an increase in immigration improves public finances due to the fall in unemployment benefits and government transfers. Then, we develop a general equilibrium model that analyses immigration in the presence of search and matching frictions as well as a rich fiscal sector. Our model is able to replicate the main empirical results.
Presented by: Marco Lorusso, University of Perugia
Session: C4: Macro Theory II March 30, 2026 15:40 to 17:20 Location: Room 515
Session Chair: Hernan Seoane, Universidad Carlos III de Madrid
The Carbon-Adjusted Fiscal Multiplier Abstract Public spending affects not only output but also the environment, by influencing greenhouse gas emissions. To this end, we extend the concept of the fiscal multiplier by introducing the carbon adjustment: the dollar value of climate damages incurred per dollar of public spending. To quantify the carbon adjustment, we augment a multi-sector New Keynesian model with an environmental block and sectoral heterogeneity in carbon intensity. We find a negative carbon adjustment for government consumption, ranging between -7 and -19 cents. Higher values for both the social cost of carbon and sectors’ carbon intensity lead to even more negative adjustments.
Presented by: Omar Rachedi, Esade Business School
Macroprudential Policy with Firm Heterogeneity Abstract I study optimal macroprudential policy when its effects on investment and productivity are taken into account. To do so, I introduce a tractable way of modeling misallocation that generates a link between investment and productivity and can be easily taken to the data. Because macroprudential policies affect investment, they lead to productivity losses. I show that, when the policymaker is constrained in their available instruments, this generates a policy trade-off between financial stability and productivity growth. I derive a sufficient statistics formula for the second-best policy, including its productivity costs. I leverage the tractability of my model to get a range of estimates for the latter using rich firm-level microdata for several European countries. The trade-off is quantitatively relevant: For baseline crisis probabilities, productivity losses switch optimal policy from a capital control to a foreign borrowing subsidy.
Presented by: Emilio Zaratiegui, Bank of England
Climate Minsky Moments and Endogenous Financial Crises Abstract How does a shift in climate policy affect financial stability? We develop a quantitative macroeconomic model with carbon taxes and endogenous financial crises to study so-called “Climate Minsky Moments”. By reducing asset returns, an accelerated transition to net zero initially elevates the crisis probability substantially. However, carbon taxes enhance long-run financial stability by diminishing the relative size of the financial sector. Quantitatively, the net financial stability effect is only negative for higher social discount rates. Even then, the welfare effects of “Climate Minsky Moments” are second-order relative to the real costs and benefits of an accelerated transition.
Presented by: Matthias Kaldorf, Deutsche Bundesbank
An ergodic theory of sovereign default Abstract We present the conditions under which the dynamics of a sovereign default model of private external debt are stationary, ergodic and globally stable. As our results are constructive, the model can be used for the accurate computation of global long run stylized facts. We show that default can be used to derive a stable unconditional distribution (i.e., a stable stochastic steady state), one for each possible event, which in turn allows us to characterize globally positive probability paths. We show that the stable and the ergodic distribution are actually the same object. We found that there are 3 type of paths: non-sustainable and sustainable; among this last category, trajectories can be either stable or unstable. In the absence of default, nonsustainable and unstable paths generate explosive trajectories for debt. By deriving the notion of stable state space, we show that the government can use the default of private external debt as a stabilization policy.
Presented by: Hernan Seoane, Universidad Carlos III de Madrid
Session: C5: Macro Theory - Monetary Policy II March 30, 2026 15:40 to 17:20 Location: Room 516
Session Chair: Alexander Meyer-Gohde, Goethe University Frankfurt
On looking through sectoral shocks: The role of (de-)anchored inflation expectations Abstract The optimal monetary policy response to sectoral shocks is usually to look through them and focus on stabilising core rather than headline inflation. However, while being a crucial concern for monetary policymakers in practice, the potential de-anchoring of inflation expectations is typically not taken into account. We re-evaluate the monetary policy recommendation of looking through sectoral shocks based on a multisector New Keynesian model that is extended to allow for endogenous long-run inflation expectations. While a potential de-anchoring of household inflation expectations results in a more aggressive interest rate policy following an adverse sectoral shock, actual inflation and output dynamics are not very different under the optimal monetary policy compared to a benchmark with perfectly anchored inflation expectations. The limited ability to control price-setting of firms in different sectors via interest rate policy prompts the policymaker to refrain from fine-tuning actual sectoral price movements as well as firms' long-run expectations of sectoral prices and inflation.
Presented by: Joost Roettger, Deutsche Bundesbank
A Multi-Sector Production Network Model for the Euro Area Abstract In this paper, we introduce a dynamic New Keynesian multi-sector model of the euro area that incorporates heterogeneous price and wage rigidities, sector-specific factor shares, and frictions in labour and capital reallocation. We analyse the transmission of sectoral and aggregate shocks and examine how monetary policy affects both aggregate and sectoral dynamics. The framework captures diverse sectoral inflation responses and offers insights into the heterogeneous transmission of monetary policy. It is well-suited for analysing the broader implications of production networks and sector-specific heterogeneities in the transmission of shocks across the economy, as well as for evaluating optimal monetary policy in a structurally complex environment. While sectoral interlinkages, price-setting heterogeneity and reallocation costs do not materially alter the response of output and consumption to a monetary policy shock, they significantly alter the sensitivity of inflation to interest rate changes on the nominal side. Preliminary results when optimizing the parameters of the central bank reaction function suggest a greater emphasis on output and inflation stabilization than the same rule would prescribe under the baseline calibration. We also show that allowing for sectoral information in the central bank reaction function may prove superior in terms of minimizing losses compared to relying on aggregate data on output and inflation alone under certain conditions. Including services inflation instead of headline inflation yields an additional reduction in losses, underscoring its potential for providing a cleaner signal on underlying inflation trends.
Presented by: Tobias Müller, European Central Bank
Relative Price Shocks and Optimal Monetary Policy in a Time Use Model of Consumption Abstract Recent research on multi-sector New Keynesian models shows that relative price changes can act as aggregate supply shocks, creating a trade-off between stabilizing inflation and output. This issue has gained renewed attention since the COVID-19 pandemic exposed global supply chain vulnerabilities. The literature suggests that maintaining price stability may come at the cost of large output losses, especially when sectors differ in price stickiness or production linkages. This paper revisits the question of optimal monetary policy under sector-specific shocks within a Beckerian framework, where consumption involves both time and goods. By integrating household time allocation into a standard New Keynesian model calibrated with U.S. data from ATUS, PCE, and CEX, the study examines how time reallocation mitigates the inflation-output trade-off. The analysis evaluates alternative monetary policy rules—from full price to full output stabilization—and explores how time-use substitutability affects optimal policy design and the construction of cost-of-living and price indices.
Presented by: Stefano Gnocchi, Bank of Canada
Informational Inertia and the Taylor Principle Abstract Determinacy bounds provide the limits on monetary policy's reaction function that rule out self fulfilling equilibria. In standard sticky price analyses, these bounds are generically nonlinear functions of model parameters and all the coefficients in this reaction function. We examine a collection of alternative models with nominal rigidities driven by different informational inertia but all having in common a vertical long Phillips curve. These models share the same determinacy bounds, independent of model specific parameters and dependent only on monetary policy's reaction to inflation. This reaction must be more than one for one - that is, the celebrated Taylor principle is shown to be necessary and not just sufficient if the long run Phillips curve is vertical - no amount of output targeting can substitute for this concern of the monetary authority for inflation.
Presented by: Alexander Meyer-Gohde, Goethe University Frankfurt
Session: C6: Time Series V March 30, 2026 15:40 to 17:20 Location: Room 517
Session Chair: Francisco Libano-Monteiro, London School of Economics
Modeling European electricity market integration during turbulent times Abstract This paper introduces a novel Bayesian reverse unrestricted mixed-frequency model applied to a panel of nine European electricity markets. Our model analyzes the impact of daily fossil fuel prices and hourly renewable energy generation on hourly electricity prices, employing a hierarchical structure to capture cross-country interdependencies and idiosyncratic factors. The inclusion of random effects demonstrates that electricity market integration both mitigates and amplifies shocks. Our results highlight that while renewable energy sources consistently reduce electricity prices across all countries, gas prices remain a dominant driver of cross-country electricity price disparities and instability. This finding underscores the critical importance of energy diversification, above all on renewable energy sources, and coordinated fossil fuel supply strategies for bolstering European energy security.
Presented by: Luca Rossini, University of Milan
Long Run in Exchange Rates, UIP and Durbin Regressions Abstract There is clear evidence from many previous studies that UIP does not hold with monthly data. However following Baillie et al (2025) Econometrics Journal, we apply Durbin regression approach to and finds that UIP is valid over long horizons. This evidence is based on heavily traded 30 day forward markets and avoids dealing with long term bonds and HAC robust standard errors. The evidence confirms the validity of the long-run international Fisher equation.
Presented by: Richard Baillie, Michigan State University and King's College London
The Impact of the Minimum Wage on Inequality: Cross-Sectional vs Lifetime Perspectives Abstract Permanent increases in the minimum wage compress the lower tail of the wage distribution, reducing cross-sectional current labor income inequality. While the effects on snapshot measures of inequality are well established, less is known about how the minimum wage shapes lifetime inequality, which better captures individuals' disparities in resources. Using rich administrative employer-employee matched data from Portugal and exploiting a minimum wage reform consisting of a large and permanent increase in the national real minimum wage between 2007 and 2010, I estimate individual-level effects during the reform and their subsequent effects in the years that follow the reform (8 years after the start of the reform) for workers across the 2006 wage distribution. Although the effects remain persistent for a few years, they relatively quickly dissipate once the real minimum wage stops rising, with workers converging back toward what their lifecycle wage and earnings trajectories in the absence of the real minimum wage increase. The transitory effects of this permanent minimum wage increase arise from two key mechanisms: (1) substantial labor income mobility causes low-wage workers to grow out of minimum wage exposure over time, and (2) the reform generates minimal impact on determinants of future wages---no effects on hours or employment, and only modest firm reallocation toward more productive firms. As a result, the policy meaningfully reduces current income inequality but has more muted effects on lifetime inequality.
Presented by: Francisco Libano-Monteiro, London School of Economics
Testing for Changes in Earth Climate Cycles Abstract Orbital earth variables such as eccentricity, obliquity and precession have been know to influence the long-term earth climate and cause for instance the cycle of ice ages during the current climate period. It is though debated if these so-called Milankovitch cycles are a rather new or since long persisting effect in earth climate history. As the Milankovitch cycles can be identified as poles in the spectral density of earth climate variables, the aim of this paper is to consider testing for changes in the spectral density matrix of a multivariate cyclical long-memory process to identify possible changes in the cyclicality of the earth climate. To do so the spectral density is locally estimated by a lag-window spectral estimator before spectral average statistics are considered. For change-point testing a CUSUM-type test statistic based on these statistics is proposed and its limiting distribution is derived where possible. Its finite sample properties are investigated in an extensive Monte Carlo study. The change-point test is finally applied to data of earth climate variables for the past 67 million years. It turns out that the influence of the Milankovitch cycles on the earth climate is only recently a driving factor beginning with the icing of the antartic continent about 13 million years ago.
Presented by: Philipp Sibbertsen, Leibniz Universitaet Hannover
Session: C7: Empirical Macro - Climate March 30, 2026 15:40 to 17:20 Location: Room 526
Session Chair: Paolo Gelain, Federal Reserve Bank of Cleveland
Temperature and the U.S. Economy: From Demand to Supply-Side Effects? Abstract We examine how the macroeconomic effects of temperature shocks have evolved in the United States since 1947. Using a time-varying parameter vector autoregression with stochas- tic volatility estimated on monthly data, we document a structural shift in the propagation of temperature shocks. Before the 1980s, higher temperatures induced demand-like dynam- ics—output and prices rose together. Since the 1980s, responses have become supply-like: real activity declines persistently while prices rise on impact and turn negative thereafter. A sectoral decomposition confirms shifts in agriculture, manufacturing, and services, with the services sector the primary driver of recent GDP dynamics. Our results reveal that food, services, and energy prices drive most of the aggregate price adjustments, while core prices remain muted. Temperature shocks now explain a rising share of medium-run output and price variation, and greater ex-ante temperature uncertainty depresses equity valuations on impact. Overall, temperature shocks have become increasingly contractionary and inflationary in nature.
Presented by: Clemente Pinilla-Torremocha, Bank of England, and European Research University
Temperature Fluctuations and Economic Activity: Evidence from Weekly US Data Abstract This paper examines how temperature fluctuations affect state-level economic activity in the United States from 1987 to 2024. Using weekly temperature and economic conditions data, we estimate impulse responses of economic activity to temperature fluctuations with panel local projections. This high frequency data allows us to identify both short and longer run effects while capturing variation across seasons and time. We find that higher temperatures are associated with small and short lived increases in economic activity, but these gains dissipate and turn negative over longer horizons. The results also show substantial heterogeneity across seasons and time. Early in the sample, higher temperatures were neutral to mildly positive at medium to long horizons. In recent decades they have consistently reduced economic activity across all seasons. These findings point to a structural break in the temperature–economic activity relationship which would be difficult to detect using lower frequency data with shorter samples.
Presented by: Chadwick Curtis, University of Richmond
Climate and Macroeconomic Volatility Abstract This paper examines how temperature shocks influence both the level and volatility of key macroeconomic variables across a broad panel of countries. We develop a novel global-to-local identification strategy that uses exogenous variation from the El Nino-Southern Oscillation (ENSO) to isolate temperature anomalies at the country level. Our results show that a 1°C rise in temperature anomalies leads to persistent inflationary pressures and a decline in economic activity, with effects particularly pronounced in developing economies – where inflation can increase by up to 1.2%. Incorporating a stochastic volatility-in-mean framework, we find that these shocks also elevate macroeconomic uncertainty by raising the volatility of GDP growth and consumer price inflation. Together, these findings highlight the importance of modeling climate shocks not only through their direct effects on prices and output, but also through their second-moment consequences, which may warrant policy attention.
Presented by: Marco Tibullo, Queen Mary University of London
Global Temperature and the Global Financial Cycle Abstract This paper investigates how global temperature shocks affect the global financial cycle and US macro-financial conditions. Using a proxy-VAR that combines global and US data with exogenous temperature innovations identified from National Oceanic and Atmospheric Administration records, we show that unexpected increases in global temperature lead to a persistent contraction in world output and a synchronized tightening of the global financial cycle. US industrial production and inflation decline in parallel, indicating that climate variability can propagate through international financial linkages. The results identify global temperature shocks as a new, climate-driven source of global financial fluctuations.
Presented by: Paolo Gelain, Federal Reserve Bank of Cleveland
Session: C8: Empirical Macro VII March 30, 2026 15:40 to 17:20 Location: Room 527
Session Chair: Junior Maih, Norges Bank
Identifying Business Cycle Turning Points in Real Time with Quantile Markov Switching Models Abstract The global pandemic created extreme swings in macroeconomic data that complicate statistical analysis. Simply dummying out these observations is not a good option, as this would eliminate macroeconomic objects of interest, such as recessions. The existing literature has focused on model augmentation in the form of time-varying volatility (e.g. stochastic volatility, ARCH, GARCH, etc.) Such approaches have been taken to workhorse models in empirical macroeconomics such as VARs and dynamic factor models. This paper evaluates an alternative approach for markov switching models, which are a popular class of models designed to identify recessions in economic data. Specifically, I investigate the performance of markov switching models that are specified in terms of conditional quantiles of recessions rather than conditional means. I show that these quantile DFMS models are capable of capturing recessions, both retrospectively and in real time, while not being strongly influenced by extreme recessions and expansions, such as those observed during and following the March-April 2020 recession.
Presented by: Jeremy Piger, University of Oregon
Nonlinearities and Heterogeneity in Firms Response to Aggregate Fluctuations: What Can We Learn From Machine Learning? Abstract Firms respond heterogeneously to aggregate fluctuations, yet standard linear models impose restrictive assumptions on firm sensitivities. Applying the Generalized Random Forest to U.S. firm-level data, we document strong nonlinearities in how firm characteristics shape responses to macroeconomic shocks. We show that nonlinearities significantly lower aggregate responses, leading linear models to overestimate the economy’s sensitivity to shocks by up to 1.7 percentage points. We also find that larger firms, which carry disproportionate economic weight, exhibit lower sensitivities, leading to a median reduction in aggregate economic sensitivity of 52%. Our results highlight the importance of accounting for nonlinearities and firm heterogeneity when analyzing macroeconomic fluctuations and the transmission of aggregate shocks.
Presented by: Simone Pesce, Central Bank of Ireland
Alternative Trend-Cycle Decomposition Methods: Real-Time Performance and Forecasting Abstract This paper conducts a critical evaluation of three prominent methodologies for estimating trends and cycles in key Euro Area macroeconomic indicators over the past three decades, including the period affected by the Covid-19 pandemic. The methods under consideration are the Hodrick and Prescott, Christiano and Fitzgerald approximation, and Harvey and Trimbur model. To address the presence of extreme observations during the pandemic, the analysis incorporates an additive outlier correction. A dedicated case study recreates the trend and cycle estimates over 30 rolling windows from January 2020 to May 2022. Results reveal similar GDP and employment estimates across all three filters for the euro area. However, the Harvey and Trimbur model exhibits distinct behavior in the case of the Industrial Production Index (IPI), yielding higher trend estimates during periods of expansion and lower ones during contraction phases, such as the global financial crisis and the COVID-19 shock Additionally, the paper introduces a novel approach aimed at jointly forecasting the trend, cycle, and aggregate series, providing a potential new criterion for evaluating decomposition methods. Despite these differences, all three approaches demonstrate comparable forecasting accuracy, suggesting no clear superiority among them.
Presented by: Francesco Ravazzolo, Libera Università di Bolzano
Structural Outlook-at-Risk (SOAR): A Model-Based Approach to Macroeconomic Tail Risks Abstract We introduce a structural, simulation-based framework for assessing macroeconomic risks—termed “SOAR”—using regime-switching dynamic stochastic general equilibrium (DSGE) models. Unlike reduced-form or univariate methods, our approach delivers joint risk assessments across multiple variables. By simulating predictive densities, we can construct fan charts and quantile paths that quantify how structural features shape tail risks. Scenario analysis leverages two complementary techniques: (i) entropic tilting, which adjusts simulation weights to explore policy behaviors or paths in expectation, and (ii) structural conditional forecasting, which imposes exact paths to assess policy impacts. We apply the framework to a medium-scale DSGE model estimated via Bayesian methods on U.S. data (1963Q1-2019Q4). The model features switching Phillips curve slopes, shifting monetary policy stances, time-varying shock volatilities, and an occasionally binding zero lower bound (ZLB), allowing us to capture complex, interconnected risk dynamics. Preliminary explorations suggest potential for identifying historical periods of elevated risk, hinting that ZLB regimes and high volatility may amplify tail risks, while counterfactual policies could have mitigated them. This approach provides a flexible tool for structural stress testing and risk-based policy evaluation, with further refinements underway to enhance its empirical scope.
Presented by: Junior Maih, Norges Bank
Session: D1: Empirical Macro & Finance March 31, 2026 9:00 to 10:40 Location: Room 527
Session Chair: Andreea Rotarescu, Wake Forest University
Capital, Intangibles, and Financial Frictions Abstract How do financial frictions shape firms’ investment in intangible capital? We show that financial constraints distort firms’ input choices, leading to systematic underinvestment in intangible assets. Exploiting an investment subsidy in Portugal that lowered the cost of both physical and intangible capital while keeping their relative price unchanged, we find that treated firms reduced their capital-to-intangible ratio by 12 percent, with larger effects for small, financially constrained firms. The distribution of treatment effects declines sharply across percentiles - firms with higher initial capital-to-intangible ratios adjust the most - revealing the signature of a binding financial wedge. Going beyond average effects, we recover the entire cross-sectional distribution of wedges between the marginal rate of technical substitution and the price ratio. The recovered distribution corresponds to the least distorted economy consistent with the data, showing that only a small share of firms are unconstrained while roughly one-quarter face wedges exceeding twenty percent - providing a direct quantitative map of financial distortions in production.
Presented by: Joao Monteiro, Einaudi Institute for Economics and Fina
Defining Current and Expected Financial Constraints using AI: Reinterpreting the Cash Flow Sensitivity of Cash Abstract We propose a new approach to identify firm-level financial constraints by applying artificial intelligence to text of 10-K filings by U.S. public firms from 1993 to 2021. Leveraging transformer-based natural language processing, our model captures contex- tual and semantic nuances often missed by traditional text classification techniques, enabling more accurate detection of financial constraints. A key contribution is to dif- ferentiate between constraints that affect firms presently and those anticipated in the future. These two types of constraints are associated with distinctly different financial profiles: while firms expecting future constraints tend to accumulate cash preemptively, currently constrained firms exhibit reduced liquidity and higher leverage. We show that only firms anticipating financial constraints exhibit significant cash flow sensitivity of cash, whereas currently constrained and unconstrained firms do not. This calls for a narrower interpretation of this widely used cash-based constraints measure, as it may conflate distinct firm types – unconstrained and currently constrained – and fail to capture all financially constrained firms. Our findings underscore the critical role of constraint timing in shaping corporate financial behavior.
Presented by: Max Schroeder, Durham University
The Credit Channel of Inflation Abstract This paper shows how inflation affects the real economy through bank balance sheets: unexpected inflation erodes banks' net worth by devaluing long-term assets relative to short-term liabilities, constraining lending and depressing real activity. Using a century of data for 18 advanced economies, I document that inflation surprises reduce aggregate loans and shrink bank balance sheets, even in absence of contractionary monetary policy. To shed light on the mechanism, I exploit rich micro-level data on bank-firm relationships, and show that inflation-exposed banks cut credit supply when inflation increases, and their borrower firms scale back investment due to credit rationing. As such, inflationary pressures on banks can counteract the traditional debt-inflation benefits.
Presented by: Lorenzo Ranaldi, University of Bonn
Zombie Prevalence and Bank Health: Exploring Feedback Effects Abstract This paper investigates feedback effects between bank health and zombie firms---financially distressed firms receiving subsidized credit. The literature focuses on how banks create zombies, overlooking zombies' impact on bank health. Using Spanish firm-bank data (2005-2014), we document a vicious cycle: lower bank capital ratios are associated with higher zombie activity in served industries, while higher zombie prevalence is associated with reduced bank capital. We link this to a previously unexplored mechanism where banks respond appropriately to observable financial distress through higher provisioning, but overlook risks from relationship borrowers receiving subsidized rates. Our findings suggest that this feedback stems not from financial distress alone, but from the combination of distress with interest rate subsidies.
Presented by: Andreea Rotarescu, Wake Forest University
Session: D2: Empirical Macro & Macro Theory - Uncertainty March 31, 2026 9:00 to 10:40 Location: Room 513
Session Chair: Francesco Saverio Gaudio, Sapienza University of Rome
Output gap and estimate uncertainty Abstract This paper explores the dynamics and determinants of estimate uncertainty of output gaps. We acknowledge the pivotal role of output gaps in guiding policymakers and focus on estimate uncertainty as an additional quantifiable variable. We define estimate uncertainty as a relative level of dispersion between various estimates of the output gap. To estimate output gaps and quantify the uncertainty, we employ a diverse set of linear univariate filters, SVAR models, and official output gap estimates from the European Commission, IMF, and OECD. The analysis covers the period from 2002:Q1 to 2021:Q4 for 25 EU countries and the United Kingdom. We employ both linear and nonlinear panel fixed effect estimators with endogenous threshold values. Our results indicate that estimate uncertainty in output gaps tends to increase during economic slowdowns, while no clear correlation between GDP growth and uncertainty is observed during economic expansions. We also provide evidence that uncertainty is persistent and that the level of persistence is higher during booms. In-sample and out-of-sample forecasts are conducted to analyze estimate uncertainty predictive power in forecasting recessions.
Presented by: Vladimir Arčabić, University of Zagreb, Faculty of Economics and Business
Uncertainty through the Production Network: Sectoral Origins and Macroeconomic Implications Abstract We study how uncertainty propagates through production networks. Using option-implied volatility data for U.S. firms, we construct a granular, forward-looking measure of industry-level uncertainty. We then estimate its effects within industries, across supply chains, and at the aggregate level. Uncertainty in upstream sectors (e.g., chemicals, steel) acts like a supply shock, raising prices and reducing employment across the network. In contrast, uncertainty in downstream sectors (e.g., automotive, insurance)resembles a demand shock, lowering prices and employment. Aggregate inflation dynamics depend on the origin of uncertainty. A multi-sector model with time-varying sectoral uncertainty shows that network propagation explains these results.
Presented by: Matteo Cacciatore, HEC Montreal
Macroeconomic Effects of Commodity Price Uncertainty Abstract This paper builds a novel index of commodity price uncertainty and investigates the macroeconomic con- sequences of related shocks. While a rich literature has documented the aggregate effects of macroeconomic, financial, and policy uncertainty shocks on real and financial activities (Bloom, 2009; Ludvigson et al., 2021; Baker et al., 2021), much less attention has been paid to uncertainty originating in commodity markets. Recent evidence by Ponomareva et al. (2024) indicates that unexpected changes in metal and energy price volatility are contractionary for the world economy and tend to coincide with declines in commodity prices. Yet, existing measures of commodity price uncertainty focus on specific dimensions of the phenomenon, motivating the development of complementary approaches. Most studies rely on simple price volatility or disagreement among forecasters as proxies for uncertainty, although disagreement primarily captures het- erogeneity in point expectations rather than the uncertainty surrounding them (Abel et al., 2016; Rich and Tracy, 2021; Boero et al., 2008). Moreover, the existing literature typically treats uncertainty as a symmetric phenomenon, overlooking the possibility that uncertainty arising from unexpected price increases may differ in both magnitude and macroeconomic consequences from uncertainty associated with unanticipated price declines. To address these points, we construct new indices of commodity price uncertainty derived from the distribution of forecast errors collected in the Consensus Economics Survey of Professional Forecasters over 1995–2025, following the methodology of Rossi and Sekhposyan (2015). For each of nine major energy and non-precious metal commodities — crude oil, coal, copper, aluminum, nickel, lead, zinc, tin, and iron ore — we compute the empirical cumulative densities of realized forecast errors and transform them into measures of total, upward, and downward uncertainty. These indices capture the time-varying probability that actual outcomes fall in the tails of the forecast-error distribution, thereby isolating forecast uncertainty without conflating it with volatility or disagreement. The commodity-specific indices display economically meaningful dynamics, featuring marked spikes dur- ing major geopolitical and macroeconomic events such as the dot-com bubble, the Iraq invasion, the Arab Spring, the global financial crisis, the COVID-19 pandemic, and the 2022 Russian invasion of Ukraine. Through principal-component analysis, we extract a common factor from each uncertainty dimension, yield- ing three global indices that represent total, upward, and downward commodity-price uncertainty. The two directional components exhibit distinct behavior: upward uncertainty tends to rise during commodity booms and inflationary phases, when realized prices systematically exceed expectations, whereas downward uncertainty peaks during recessions, when realized prices fall short of forecasts. We incorporate these indices into a structural VAR spanning 2000Q1–2025Q2, alongside world industrial production Baumeister and Hamilton (2019), the MSCI global stock index, and the macroeconomic and financial uncertainty factors of Ludvigson et al. (2021). The impulse-response analysis reveals that shocks to total commodity uncertainty depress global activity and financial markets, consistent with the contractionary patterns documented in earlier work. However, decomposing total uncertainty uncovers a crucial asymmetry: the adverse effects are entirely driven by upward uncertainty, whereas downward uncertainty exerts negligible influence on output, prices, or financial conditions. In other words, uncertainty originating from unexpected price increases — rather than about price declines — is what tightens global conditions and dampens economic activity. Once both components are accounted for, the previously reported negative relationship between uncertainty and commodity prices largely disappears, suggesting that earlier findings may have conflated uncertainty shocks with concurrent demand or supply fluctuations affecting commodity prices. These results indicate that the macroeconomic consequences of commodity prices uncertainty are highly nonlinear. Economic agents appear more sensitive to uncertainty surrounding potential cost-push pressures than to uncertainty about price relief, consistent with asymmetric risk-management behavior among firms and policymakers.
Presented by: Lorenzo Tonni, University of Milan
The Variety Effect in Times of Uncertainty Abstract This paper shows that the variety effect, captured through monopolistic competition and increasing returns to specialization, acts as an amplification mechanism in the transmission of macroeconomic uncertainty shocks. The variety effect generates endogenous fluctuations in aggregate productivity through changes in the range of available products, due to firm entry and exit. The estimated intensity of the variety effect implies sizable productivity fluctuations, which shape households’ saving decisions and strengthen the risk channels through which uncertainty shocks propagate to the economy. When the equity risk-premium channel dominates, the variety effect implies deeper uncertainty-driven recessions, as discouraged investment in the stock market and thus the fall in product creation leads to a drop in endogenous productivity.
Presented by: Francesco Saverio Gaudio, Sapienza University of Rome
Session: D3: Empirical Macro VIII March 31, 2026 9:00 to 10:40 Location: Room 514
Session Chair: Luiggi Donayre, University of Minnesota - Duluth
Retail Inventories and Inflation Dynamics: The Price Margin Channel Abstract Using industry-level panel data and plausibly exogenous variation in supply conditions, we estimate the elasticity of retail price margins with respect to inventories along the retailer’s optimal pricing curve. We find that this elasticity is negative and statistically significant, implying that lower finished-good inventories lead to higher price margins. We assess the implications of this channel for inflation dynamics within a New Keynesian Phillips curve (NKPC) framework that links inventories to retailers’ markup behavior. Incorporating the inventory-sales ratio into the NKPC markedly improves the model’s empirical fit and helps account for two notable recent inflation episodes: the missing disinflation of 2009–2011 and the COVID-era surge.
Presented by: Hyunseung Oh, Federal Reserve Board
Pre-Grant Patents and Innovation Diffusion Abstract Are pre-grant patents effective forward-looking signals of global innovation diffusion? To address this question, I construct an import-share–weighted index of pre-grant patent filings using a novel panel of patent and trade data (1980–2019) covering 21 economies (17 OECD and 4 BRICS). A Bartik shift–share instrument based on partner-country filings isolates exogenous trade-driven shocks, with a comparable index of granted patents used for contrast. Panel local projections reveal that trade-driven pre-grant shocks, though slower to manifest, produce larger long-run TFP gains than granted-patent shocks and elicit stronger, more immediate equity-market responses. At the sector level, pre-grant shocks boost manufacturing TFP and R&D capital within one year and sustain growth thereafter; these effects are amplified in economies with more R\&D-intensive manufacturing sectors. At the industry level, I leverage the quality-adjusted cited patent index of LaBelle et al. (2023) and its interaction with value-added intensity to explore heterogeneous diffusion patterns. Results indicate that citation-driven industries such as chemicals experience gradual but persistent gains; machinery shows stronger trade-responsive dynamics; and textiles fall in between, reflecting a balanced transmission channel. Robustness checks confirm that countries with higher value-added industries and stronger R&D-intensive manufacturing sectors exhibit greater transferability of early-stage knowledge into future productivity gains, highlighting the structural complementarity between trade and innovation networks.
Presented by: Jinghui Yu, Lancaster University
The Shape of State-level Recoveries Abstract We develop a novel, transparent, and flexible algorithm to classify U.S. state-level recessions as having temporary (U-shaped) or more permanent (L-shaped) effects, effectively relaxing the assumptions that recessions are all alike across, but also within states. Our approach classifies recessions based on a relative growth performance during the recovery period, providing a crucial alternative to model-based methods that often suffer from weak identification due to the volatility of state-level data. Crucially, our approach yields a probabilistic classification and quantifies the strength of the recovery for each state and for each state-level recession. Using the Philadelphia Fed's state coincident index for the 1979-2025 period, we document substantial heterogeneity. We find that for most states, recessions are best characterized as a weighted average of L and U shapes, a systematic feature that masks underlying regional dynamics. Specifically, we show that the relationship between industry composition and recession depth is closely related to the recession shape. States with heavy manufacturing exhibit a higher likelihood of L-shaped recoveries, while growth in service-oriented states is associated with more frequent U-shaped downturns. We also show that synchronization with the national cycle is shape-dependent: the high comovement of Heartland states is largely driven by their tendency toward U-shaped recoveries, whereas the noted divergence of resource-based states stems from their exposure to idiosyncratic L-type events. Our findings are robust to different measures of above-average growth, the length of the recovery, and different recession classifications. These results refine existing results in the regional business cycle literature by adding a crucial recovery-shape dimension, highlighting the importance of heterogeneity for effective, targeted countercyclical policy.
Presented by: Luiggi Donayre, University of Minnesota - Duluth
Session: D4: Forecasting I March 31, 2026 9:00 to 10:40 Location: Room 515
Session Chair: Giulia Mantoan, Bank of England
Revisiting EWMA in High-Frequency Portfolio Optimization: A Comparative Assessment Abstract This paper compares the statistical and economic performance of state-of-the-art highfrequency based multivariate volatility models with a simpler, widely used alternative—the Exponentially Weighted Moving Average (EWMA) filter. Using over two decades of 100 U.S. stock returns (2002–2023), we assess model performance through a Global Minimum Variance portfolio optimization exercise across various forecast horizons. We find that the EWMA model consistently outperforms more complex HF-based volatility models, delivering significant utility gains when including transaction costs, due in part to its lower turnover. Even in the absence of transaction costs, the EWMA filter cannot be beaten in most cases.
Presented by: Anne Opschoor, Vrije Universiteit Amsterdam
The income side of growth-at-risk Abstract We revisit the role financial conditions play in U.S. growth-at-risk and its composition. However, in contrast to the now vast literature on growth-at-risk, we do so through the lens of the income, rather than product side of growth. While Gross Domestic Income and Gross Domestic Product are intended to measure the same quantity, measurement issues frequently lead to substantially different pictures of the economy. In the context of growth’s expected shortfall, we use in-sample and out-of-sample evidence to show that these differences also imply distinct perspectives on the magnitude of downside risk to the economy. The different approach to measurement also provides a different perspective on growth-at-risk. On the product side, Amburgey and McCracken (2025) show that the vast majority of growth-at-risk is investment-at-risk. Tighter financial conditions largely affect growth via firms decision to invest. On the income side, tighter financial conditions manifest as a decline in firm profits but also household compensation especially in the near term.
Presented by: Michael McCracken, federal reserve bank of saint louis
Forecasting Structural Change Models Using Band Spectral Regression Abstract We examine out-of-sample forecasting performance in structural change models where model coefficients are subject to structural breaks. Specifically, we compare two approaches, namely, the break-detection based approach of Yamamoto and Perron (2013) and the band spectral regression based approach of Wada (2022). The former involves estimating and testing for break dates following Bai and Perron (1998) in the frequency domain, and generating the out-of-sample forecast accordingly. In contrast, the latter does not explicitly account for structural breaks. Following Yamamoto and Perron (2013), we also allow for low-frequency contamination of unknown form, such as level shifts or changes in the slope of linear trend. Our findings show that the break-detection approach performs better when (i) there is no low-frequency contamination, (ii) the low frequency contamination is well approximated by the trend components that are removed before the break-detection is applied, plus no low frequency truncation is applied; and (iii) the regressor does not contain a unit root. In all other cases, the band spectral regression method delivers superior forecasts. Two empirical applications---one examining the link between oil prices and production, and the other focusing on exchange rate forecasting---confirm that the band spectral regression based approach provides more reliable forecasts in practice.
Presented by: Tatsuma Wada, Keio University
Forecasting Macroeconomic Risks in the UK Abstract We construct statistical measures of UK macroeconomic risks, building quantile-regression density forecasts for inflation and GDP growth. These estimates account for the UK's position as a small-open economy and capture time variation in tail risks, owing to variation in economic and financial conditions, in addition to changes in central forecasts. To highlight how they can provide valuable signals about the balance of risks, we use a battery of tests to compare our predictive distributions to those captured in the Bank of England's fan charts. Our fitted densities for growth and inflation are reasonably calibrated, outperforming the fans in terms of sharpness and accuracy in the tails. Although our estimates for inflation perform similarly to the fans over the last two decades overall, they better capture economic narratives when inflation deviates from target. These tools can contribute to a broader suite for quantifying macroeconomic risks in the UK, with regular evaluation of density forecasts necessary to ensure that the toolkit remains fit for purpose as the constellation of shocks hitting the UK economy evolve.
Presented by: Giulia Mantoan, Bank of England
Session: D5: Macro Theory III March 31, 2026 9:00 to 10:40 Location: Room 516
Session Chair: Ricardo Marto, Federal Reserve Bank of St. Louis
Transport and Structural Change in Latin America 1900–2005 Abstract This paper quantifies the long-term effects of sequential improvements to transport infrastructure, including railways and roads, on trade and structural changes in Latin America from 1900 to 2005. To this end, we calculate the transport index and elasticity of transport to each country. Then we develop a general equilibrium model with four sectors—the Natural Resource Sector (NRS), Man ufacturing Sector (MS), Service Sector (SS), and Transport Sector (TS)—where transport infrastructure are essential for enabling exports and shaping labor al location in the country. The model reproduces the observed expansion of export capacity and labor reallocation associated with network improvements. Counter factual exercises that set at early 20th-century levels result in significantly lower trade volumes and a decline in labor in NRS, MS, and SS. This underscores the critical role of transportation on sectoral composition.
Presented by: Ticona Huanca Wilma, Universitat de Barcelona
The Effects of Uncertainty on Firms’ Pricing Behavior and Activity Abstract This paper examines the causal effects of demand uncertainty on firms’ pricing behavior and economic activity using managers’ subjective expectations. Employing an instrumental variables approach that exploits differential industry exposure to exogenous uncertainty sources, we find that increased uncertainty causes firms to reduce prices through lower markups and decrease activity by reducing capacity utilization. To rationalize these findings, we develop a macroeconomic model where firms face capacity constraints and must commit to prices and capacity before demand uncertainty resolves. In response to increased uncertainty, the model’s putty-clay production technology generates a mechanism where capacity constraints truncate upside gains while firms bear the full downside losses, inducing firms to lower prices preemptively to minimize expected losses from excess capacity. Our calibrated model shows that a one standard deviation demand uncertainty shock reduces output by approximately 0.5 percent, with producer and consumer price inflation declining by roughly one-half and one-tenth of a percentage point, respectively. Absent markup reduction, the recessionary dynamics would be substantially more severe, as lower prices and markups dampen uncertainty’s negative effects. These findings demonstrate that idiosyncratic demand uncertainty generates disinflationary pressures through a distinct transmission mechanism—one com- plementing the inflationary effects of aggregate cost uncertainty emphasized in prior work—establishing demand uncertainty as an economically significant driver of busi- ness cycle fluctuations.
Presented by: Giuseppe Fiori, Board of Governors of the Federal Reserv
Privilege Lost? The Rise and Fall of a Dominant Global Currency Abstract Economic size matters for three key aspects of a dominant currency: safety, liquidity, and insurance. First, larger countries tend to be more diversified, increasing their debt capacity and safety. Secondly, the larger the share of a country in the supply of safe assets, the more liquid and attractive its bonds are. Finally, the larger a safe country’s share in the world economy, the more its bonds appreciate in downturns, providing insurance for investors. We argue that these three aspects reinforce each other and show how they erode if the dominant country grows less than the rest of the world.
Presented by: Nuno Coimbra, Banque de France
Shipping to America Abstract We study the evolution of containerized shipping to the United States and its implications for trade disruptions. Using high-frequency satellite data covering all U.S.-bound container vessels between 2017 and 2025, we document new facts on capacity allocation across the major routes and construct measures of disruptions from port congestion, canal bottlenecks, and rerouting associated with security threats. To interpret these patterns and quantify their implications, we develop a multi-route general equilibrium model in which shipping firms choose sailing speeds, load factors, and capacity allocations. Calibrated to micro data, the model replicates observed trends. The model is then used to quantify the macroeconomic impact of our underlying estimates of U.S.-bound shipping disruptions.
Presented by: Ricardo Marto, Federal Reserve Bank of St. Louis
Session: D6: Empirical Macro - Unemployment March 31, 2026 9:00 to 10:40 Location: Room 517
Session Chair: Irina Panovska, University of Texas at Dallas
Who benefits from running the economy hot? New evidence from Norway Abstract Labor market fluctuations reflect changes in both unemployment and labor force participation. Understanding how participation responds to the business cycle is essential for assessing progress toward the high and stable employment mandates of central banks. Building on the flow-based framework of Hobijn and Şahin (2021, Proceedings of the 2021 Jackson Hole Symposium), this paper studies the participation cycle in Norway — the cyclical behavior of transitions across employment, unemployment, and non-participation — and its response to business cycle shocks. Using comprehensive administrative data covering the entire Norwegian population, we construct monthly transition rates over the past two decades and analyze their conditional behavior in a VAR framework with a main business cycle shock. We document the cyclicality of each transition rate across demographic groups, sectors, and age categories, and evaluate how economic fluctuations shift employment opportunities across marginal and core workers. In particular, we investigate whether the central bank is able to push marginalized workers into the labor force when running the economy hot.
Presented by: Michele Castegini, European University Institute
Population aging through the lens of DSGE-OLG-NK model: implications for unemployment and monetary policy in the Euro area Abstract We study how the evolving age structure of the working population affects unemployment and monetary policy. Using a large-scale New Keynesian overlapping generations (OLG) model calibrated to the Euro area, we show that demographic change has reduced the unemployment rate by about two percentage points. Beyond this long-run effect, we also examine how aging shapes broader macroeconomic dynamics and the conduct of monetary policy. We also find that an aging workforce leads to greater macroeconomic stability. In parallel, the monetary policy sacrifice ratio—the amount of real volatility required to reduce inflation volatility—declines in an aging society. This result is driven by a feedback loop between compositional and behavioral effects: older workers respond less to monetary policy shocks, stabilizing labor demand and allowing policy to focus more on inflation control. Our findings suggest that continued population aging in Europe may keep unemployment low and support a more hawkish monetary policy stance.
Presented by: Joanna Tyrowicz, University of Augsburg, FAME|GRAPE, and IZA
Mexico’s Labor Market through the Lens of the Beveridge Curve Abstract This paper presents the first Beveridge Curves for Mexico and its regions, constructed from a newly compiled dataset on job vacancies. Focusing on the post-COVID-19 period, the evidence reveals a sharp rise in labor market slack at the onset of the pandemic, followed by a gradual tightening from 2021 onward, with northern regions experiencing a faster recovery. By 2023, vacancy rates had declined without a corresponding increase in unemployment, suggesting a rebalancing of labor demand across regions. Building on the vacancy–unemployment relationship that defines the Beveridge Curve, the paper also develops the first measure of labor market slack for Mexico: the vacancy to unemployment (V/U) ratio, which is then used to analyze the relationship between labor market conditions and inflation. In addition, incorporating vacancy data into a Structural Bayesian VAR allows for the identification of structural labor market shocks and the examination of business cycle dynamics.The results indicate that Beveridge Curve variables are primarily driven by labor market shocks: unemployment dynamics are explained mainly by labor supply and matching efficiency shocks, while wage bargaining shocks account for most of the variation in vacancies at the national level, with notable heterogeneity across regions.
Presented by: Irina Panovska, University of Texas at Dallas
Session: D7: Macro Theory IV March 31, 2026 9:00 to 10:40 Location: Room 526
Session Chair: Daniela Hauser, Bank of Canada
The Building Blocks of Inflation: The role of monetary policy and the gap between goods and services Abstract I build and estimate a three-region structural macroeconomic model with a goods, services, housing and oil sector. The model also has meaningful household portfolio decisions over foreign and domestic bond holdings and financial intermediation to investigate the efficacy of large-scale asset purchases (LSAP) by a central bank. The model is built and estimated to ensure that the potential causes that have been pointed to as a reason for the global inflation seen in the COVID recovery economy are accounted for. Examining the dynamics of the model, LSAPs conducted in an economy with relative high demand for goods rather than services will lead to a bigger expansionary and inflationary impact than in an economy where demand for services is relatively higher than goods. I also find that LSAPs are more expansionary and inflationary when the service sector is incurring supply shocks. These findings help us understand why LSAPs conducted in the global financial crisis had such a different impact than those conducted in the COVID economy.
Presented by: Sacha Gelfer, Bentley University
QE Without REs---The Fed’s Triple Mandate and Macro-Stability Abstract It is well established that under rational expectations and the zero lower bound (ZLB), Taylor-type rules permit two steady states: one at target inflation and another at deflation when the nominal interest rate approaches zero (Benhabib et al., 2001). Yet, the U.S. experience during the early phase of the 2008 Global Financial Crisis (GFC) deviated from this pattern. At the ZLB, inflation experienced a brief deflationary episode followed by a rebound towards the Fed’s target. The Fiscal Theory of the Price Level explains the absence of a deflationary spiral (Cochrane, 2022), but fails to account for the heightened volatility. This paper proposes a potential explanation for the U.S. experience through the lens of adaptive learning and unconventional policy in a New-Keynesian framework. The findings are threefold: the economy remained in an indeterminate state in which sunspot equilibria were learnable; explicit Fed intervention in the far end of the yield curve through quantitative easing promoted expectational stability; and wealth effects mattered greatly for the convergence of expectations.
Presented by: Tengyuan Liang, University of California, Irvine
Trade Policy Shocks and Monetary Policy in Partially Dollarized Economies Abstract A period of deglobalization, marked by rising protectionism, trade fragmentation, and renewed geopolitical tensions, is reshaping global economic relations and generating inflationary pressures worldwide. This paper studies how tariff shocks propagate through partially dollarized, emerging economies and how dollarization alters their macroeconomic and monetary policy implications. Empirically, using the U.S. tariff shocks identified by Schmitt-Grohe and Uribe (2025) in a panel VAR for advanced and emerging economies, the analysis shows that emerging markets—especially those with high dollarization—face larger and more persistent depreciations, stronger inflationary responses, and deeper output contractions, consistent with higher exchange rate pass-through and tighter financial constraints. Theoretically, the paper extends the New Keynesian small-open-economy framework of Monacelli (2025) to incorporate price and financial dollarization, calibrated to the Peruvian economy. In this setting, even moderate exchange rate movements have amplified effects on prices and output, giving rise to potentially sizable inflation–output trade-offs for both export and import tariff shocks—whereas in a standard small open economy, export tariffs behave as pure demand shocks. Dollarization thereby magnifies the macroeconomic costs of protectionist trade policies and reinforces the case for partial exchange-rate stabilization under a managed-float regime.
Presented by: Daniela Hauser, Bank of Canada
Session: Plenary II: Martin Ellison, University of Oxford, Nuffield College, CEPR March 31, 2026 11:10 to 12:10 Location: Room: 512
Session: E1: Empirical Macro - Inflation III March 31, 2026 13:30 to 15:10 Location: Room 527
Session Chair: James Morley, University of Sydney
The Role of Global Inflation in Estimation of the US Output Gap in the post Bretton Woods Era: Evidence from Multivariate Unobserved Components Models Abstract We study the role of global inflation in estimation of the US output gap using multivariate unobserved components models. We use the empirical evidence presented in Ciccarelli and Mojon (2010) as our main motivation to augment the US inflation equations with global inflation. We further allow past global and domestic inflation to influence the output gap dynamics. The estimation results from the state-space model show a negative association of past inflation with the output gap. The role of global inflation is large in the negative association with the output gap. The reduction in output was substantial during the 1970s reaching 3.4 percent below trend during 1974-75, and present in the recent data as well. The estimates also reveal a stronger positive association of global inflation with US inflation after allowing for inflation to influence the output gap dynamics. Finally, we do not find evidence of inflation influencing the stochastic trend of output. The results are robust to multiple specifications, extensions, and sub-samples. Additional decomposition suggests that global inflation influences the productivity gap, and not the hours gap. We further develop a dynamic factor augmented unobserved components model that allows for feedback from US output gap to global inflation. The results show that a common factor of global inflation and domestic inflation as the primary contributor of the negative association of inflation on output gap. The analysis highlights the strong informational value of global inflation for US monetary policy for both inflation and output.
Presented by: Arabinda Basistha, West Virginia University
Blockwise Boosted Inflation: non-linear determinants of inflation using machine learning Abstract We propose the Blockwise Boosted Inflation Model (BBIM), a boosted tree framework that decomposes inflation dynamics into predictive components aligned with an open-economy hybrid Phillips curve. Demand and supply contributions are identified by imposing monotonicity constraints, ensuring theory-consistent links between inflation and key indicators. Applied to monthly UK CPI inflation, the model shows that the recent surge has been driven mainly by global supply shocks transmitted through supply chains. We also uncover an L-shaped Phillips curve relationship between inflation and labour market tightness, with tight labour markets amplifying recent inflationary pressures. By contrast, earlier episodes saw non-linearities more strongly tied to broader slack, particularly during recessions. The model further accounts for trend shifts informed by inflation expectations. Short-term household expectations have recently displayed persistent non-linear effects, temporarily raising trend inflation and prolonging inflationary pressures, while longer-term expectations remain anchored. Out-of-sample, the BBIM delivers competitive forecasting performance relative to linear benchmarks and unstructured machine learning methods. Our approach provides a flexible yet interpretable framework that combines economic structure with machine learning for policy-relevant analysis of inflation dynamics.
Presented by: Galina Potjagailo, Bank of England
Unpacking global inflation Abstract We study the international synchronization of inflation through the disaggregation of inflation into goods and services inflation. This split is motivated by the goods and services sectors being mostly naturally aligned with the tradeable and non-tradeable components of inflation, respectively. To do so, we construct a dataset of goods and services inflation from 42 countries, spanning the period 1971Q1–2023Q4. Three main findings emerge. First, the sectoral split reveals multiple sources of global inflation synchronization, in contrast to the single global inflation factor identified by extant work. These sources of inflation synchronization differ both in their persistence and in the way they affect goods and services inflation. Second, while global factors dominate goods inflation dynamics, they also play substantial -- albeit smaller -- roles in services inflation. Third, the influence of global factors has evolved over the sample period. Subsample analysis suggests that their contribution to services inflation has risen in recent decades, whereas the inclusion of the post–COVID-19 period notably increases their contribution to goods inflation.
Presented by: Benjamin Wong, Monash University
Is inflation driven by aggregate or sectoral output gaps? Abstract We examine whether inflation is driven by aggregate or sectoral output gaps. The aggregate output gap may not fully capture inflationary pressures because it can obscure sectoral shocks and heterogeneity in propagation to prices. We find that aggregating sectoral output gaps by weights estimated from real-time regressions produces a better fit of the Phillips curve than using the aggregate output gap. We confirm the sectorally-aggregated output gap based on these weights has significant explanatory power for inflation beyond the aggregate output gap and find it performs better in forecasting inflation, although the aggregate output gap retains some distinct relevant information.
Presented by: James Morley, University of Sydney
Session: E2: Empirical Macro X March 31, 2026 13:30 to 15:10 Location: Room 513
Session Chair: Mattia Alfero, University of Rome "Tor Vergata"
Developing a House Price at Risk framework for the UK Abstract This paper develops a house price-at-risk model for the UK that allows us to track different parts of the distribution around house price growth, understand their drivers, and decompose past developments. We do so at the national level, and across nine English regions, Wales, Scotland, and Northern Ireland. We utilise a comprehensive list of possible variables and indicators that contribute to house price dynamics and develop measures of supply misalignment that capture regional differences. Our main findings are that since the 1970s, the most important factors for median house price growth have been: GDP growth, stock market growth, mortgage interest rate change, transaction growth, as well as exuberance, and business confidence. For the tail of the distribution, we find the key drivers instead to have been transaction growth, mortgage rate change, credit to GDP gap, financial stress, and exuberance. The contributions of these variables in explaining historical house price falls have been different over the decades. At the regional level, we find heterogeneous results for the impact of changes in mortgage rates, with supply-inelastic regions having more reactive house price growth. Finally, we find that an increase in the housing supply in most regions is related to future alleviation of price pressures in regional markets although there are some notable exceptions.
Presented by: Tihana Škrinjarić, Bank of England
Debt-at-Risk Abstract This paper proposes a novel framework for analyzing the risks surrounding the public debt outlook, the "Debt-at-Risk." It employs a quantile panel regression framework to assess how current macro-financial and political conditions impact the entire spectrum of possible future debt outcomes. Many of these factors—including financial conditions and economic variables such as initial debt and GDP growth—predict both the expected level and the uncertainty of future debt, implying pronounced variations in risks, especially in the upper tail of the distribution. By combining the roles of these factors, we find that in a severely adverse scenario—the 95th percentile of the future debt distribution, or debt-at-risk—global public debt could be approximately 20 percentage points higher than currently projected. The magnitudes and sources of debt risks vary over time and across countries, with high initial debt amplifying the effects of economic and financial conditions on debt-at-risk. Furthermore, empirical estimates indicate that debt-at-risk is a key variable for predicting fiscal crises.
Presented by: Faizaan Kisat, International Monetary Fund
Estimating large-scale nonlinear macroeconomic models using the ensemble transform Kalman filter Abstract This paper shows how to estimate large-scale nonlinear Dynamic Stochastic General Equilibrium (DSGE) models using the Ensemble Transform Kalman Filter (ETKF). Estimating largescale DSGE models with many state variables, such as multi-sector and multi-country models or those with numerous shocks and real and nominal frictions, presents significant computational challenges. Failing to account for nonlinearities leads to inconsistent parameter estimates. However, standard nonlinear filtering methods become infeasible in high-dimensional settings due to the large number of particles required in particle filtering or the computational burden of tensorproduct-based discretization methods. This paper demonstrates that the ETKF, combined with data augmentation, provides a computationally efficient and accurate alternative. We illustrate this through a simulation study estimating a multi-country DSGE model and an empirical application that quantifies the importance of nonlinearities in estimating the role of sectoral shocks in driving aggregate fluctuations within a multi-sector model. For standard estimates of elasticities of substitution between goods in both demand and production, we find that sectoral shocks contribute approximately 36% more to aggregate fluctuations in the nonlinear economy compared to the linear economy.
Presented by: Amir Alipoor, Tilburg University
Estimating Heterogeneous DSGE Models Abstract This paper addresses the estimation of Heterogeneous Dynamic Stochastic General Equilibrium models based on the Mixture of Student-t by Importance Sampling weighted Expectation-Maximization. The proposed method can handle target distributions that exhibit non-elliptical shapes, such as multimodality and skewness, and is also well suited for parallel computing. Furthermore, to avoid weight degeneracy, a simple robustification of the algorithm is introduced. The proposed method is compared in a Monte Carlo exercise with standard Markov Chain Monte Carlo and the recently proposed Sequential Monte Carlo method. Monte Carlo results compute posterior moments for two canonical heterogeneous models of increasing complexity: a one-asset New Keynesian model and a two-asset New Keynesian model. The three methods deliver similar results, with the proposed method being faster and, therefore, could be a useful tool in the analysis of Heterogeneous Agent models.
Presented by: Mattia Alfero, University of Rome "Tor Vergata"
Session: E3: Empirical Macro XI March 31, 2026 13:30 to 15:10 Location: Room 514
Session Chair: Gianni Amisano, Federal Reserve Board
Disentangling the drivers of exuberant house prices Abstract This paper examines the macroeconomic drivers of U.S. housing price exuberance. Using a time-varying parameter VAR with stochastic volatility and recursive unit root tests, we identify episodes of explosive house price dynamics and attribute them to structural shocks—including credit supply, credit demand, monetary policy, and household expectations. Contrary to the notion that exuberance reflects non-fundamental factors, our findings show these episodes often stem from structural shocks, particularly those related to credit and policy. Conditioning on specific shocks allows earlier detection of exuberance compared to aggregate data. Moreover, the transmission of shocks to house prices changes significantly during exuberant phases.
Presented by: Marta Rodriguez, bank of Spain
Quantifying Demand Shocks in the Green and Digital Transition Abstract We use web-search data to construct indices that proxy the derived demand for metals - specifically cobalt, lithium, and nickel - which are key inputs in technologies driving the energy and digital transition. These indices are incorporated into monthly Structural Vector Autoregressive (SVAR) models of the global markets for these metals. Identification of structural shocks relies on a combination of zero, static and dynamic sign restrictions, allowing us to disentangle supply shocks from multiple demand-side shocks that drive the real price of metals. In particular, we isolate a demand component linked to the technological uptake of metals in the energy and digital transition. Our framework enables a quantitative assessment of the relative contribution of each structural driver to price dynamics and highlights the growing macroeconomic relevance of technology-linked metal demand.
Presented by: Andrea Bastianin, University of Milan
Micro-based SVAR Identification Abstract This paper develops a unified framework for integrating microeconomic evidence into the identification of structural shocks in vector autoregressive (SVAR) models. The approach allows information to flow between micro-level estimates and macroeconomic dynamics, bridging the gap between micro-based causal inference and aggregate time-series analysis. We introduce a total likelihood formulation that jointly estimates micro and macro models anchored by common structural parameters. The framework is flexible and modular, accommodating multiple micro studies or conventional identification schemes within a single coherent system. By linking micro evidence to macro impulse responses, it offers a transparent and data-driven strategy for improving the structural interpretation of aggregate shocks.
Presented by: Annika Camehl, Erasmus University Rotterdam
How to catch a star? Reliability of filtering estimates in linear state space systems Abstract In macroeconomic policy analysis, estimates of latent variables such as the natural rate of interest and the output gap are important inputs for policymakers. We develop reliability measures to assess the quality of estimates of latent variables in linear state space models that go beyond describing the level of uncertainty surrounding these estimates. Motivated by the (infeasible) population regression of the true state at time $t+h$ on time $t$ information and the time $t+h$ forecast, we propose the (conditional) correlation and the associated $R^2$ as intuitive measures to assess the predictability of the latent states. We relate this measure to the posterior variance of the state estimate. To help understand the source of the posterior uncertainty we propose the half-life and the contribution of the initial state to the conditional forecast error variance as additional diagnostics. Models that estimate the level of the hidden state very imprecisely may still imply predictability at relevant horizons. The Holston-Laubach-Williams (HLW) model features high state uncertainty, but reasonable predictability at medium-run horizons. Since it is not possible to assess whether the models accurately quantify the uncertainty surrounding the latent variables, researchers should evaluate the predictive accuracy of forecasts for observables.
Presented by: Gianni Amisano, Federal Reserve Board
Session: E4: Finance March 31, 2026 13:30 to 15:10 Location: Room 515
Session Chair: Fabio Parla, University of Palermo
Beyond Time Diversification: Clustering Private Equity Vintages for Resilient Portfolio Design Abstract This paper proposes a new framework to cluster private equity (PE) vintages by economic and structural features, such as region, industry focus, and macroeconomic regime at fund inception, rather than calendar year. By segmenting vintages that exhibit similar performance patterns and resilience, we challenge traditional vintage-year diversification and enable more robust portfolio construction and stress testing. We aim to (i) develop a systematic segmentation based on fund attributes, (ii) apply established PE performance models (Yale, Buchner, and a Nowcasting approach) within each segment, and (iii) introduce a stress testing protocol to assess segment resilience under adverse macro scenarios. Finally, we analyze diversification benefits and show that selecting vintages across segments, instead of purely by year, enhances cycle resilience and improves risk-adjusted outcomes for PE portfolio design and monitoring.
Presented by: Sara Boni, Free University of Bolzano-Bozen
Local Estimation for Option Pricing: Improving Forecasts with Market State Information Abstract We propose a novel local M-estimation framework for option pricing models that incorporates state-dependent information to improve out-of-sample forecasting performance. Our approach reweights historical observations based on current market conditions, using kernel functions with bandwidths selected via a validation procedure. Specifically, we apply this method to both GARCH and stochastic volatility models and utilize state variables such as VIX, realized volatility, and time to capture the market conditions. Our results show that local estimation improves the accuracy of forecasting S&P 500 option implied volatilities and underlying risk-neutral distributions substantially during low-volatility environments, while in high-volatility periods, improvements are primarily highlighted for deep out-of-the-money puts. The local estimators also outperform the non-local estimators in explaining future option returns. Our findings suggest that local information, when properly incorporated into the estimation process, can enhance the accuracy and robustness of option pricing models.
Presented by: Hyung Joo Kim, Federal Reserve Board
Geopolitical Risk and the U.S. Stock Market: A Comparative Portfolio Analysis Abstract This paper analyzes the asymmetric impact of geopolitical risk on U.S. sectoral equity returns from January 1985 to April 2025. Unlike previous studies that focus on aggregate market effects, we examine ten Fama–French portfolios to uncover heterogeneity in how different sectors respond to geopolitical risk (GPR). We further distinguish between geopo- litical risk acts (GPRA) and threats (GPRT) to assess whether markets react more strongly to realized events or to the anticipation of conflict. Using a nonlinear threshold model, we capture regime-dependent and asymmetric dynamics in the relationship between geopolitical risk and sectoral returns. This framework allows the sensitivity of stock portfolios to vary across market conditions. Our results show that U.S. equity portfolios are significantly and asymmetrically affected by GPR shocks, with the sign and magnitude of the response dif- fering across sectors and regimes. These findings highlight the importance of accounting for both sectoral structure and nonlinearity when assessing the effects of geopolitical tensions on financial markets. Keywords: US stock market, Geopolitical Risk, GPRA, GPRT, Nonlinear Model, Threshold models.
Presented by: Fredj Jawadi, University of Lille
Spectral climate risk Abstract In this study, we examine the return performance of a green-minus-brown (GMB) portfolio aiming to hedge climate risk in the United States. While existing studies analyze the effect of climate risk on returns in the time domain, we investigate how the sensitivity of portfolio returns to climate risk varies across frequency bands. To this end, we use the extended Wold decomposition of GMB risk-adjusted returns and shocks to a news-based index that proxies for climate change concern. In the second stage, a spectral climate risk factor analysis reveals that green stocks outperform brown stocks over short-term horizons (i.e., cycles shorter than approximately eight days). Furthermore, our results indicate that among the three main categories of climate change risks (physical, transition, and liability risks), greater concern about transition risk plays the most prominent role in explaining the observed outperformance of the green portfolio.
Presented by: Fabio Parla, University of Palermo
Session: E5: Forecasting II March 31, 2026 13:30 to 15:10 Location: Room 516
Session Chair: Michael Owyang, Federal Reserve Bank of St Louis
Structural forecast analysis Abstract This paper shows how the structural representation of a vector autoregressive model can support forecasting. We offer a unified framework between reduced-form forecast and structural analysis, and describe how the use of the latter can help form a narrative of two reduced-form objects of real-time forecasting: the forecast errors made relative to the outturn of the data, and the revisions of the full forecast made when new data become available. We conduct a real-time exercise on the UK focusing on the inflation surge that followed the pandemic. We show that the inflation forecast was revised up not only due to contractionary supply-side shocks, but also due to a mix of expansionary demand-side shocks and a change in the underlying unconditional mean.
Presented by: Michele Piffer, King's College London + Bank of England
Forecast Combination Using Random Subspace Abstract This paper investigates forecast aggregation using the random subspace regressions method (RSM). Through extensive Monte Carlo simulations, we demonstrate that RSM can outperform equal-weight averaging by repeatedly selecting, uniformly at random, small subsets of individual forecasts and combining them via OLS regressions. The optimal value of selected individual forecasts is also analyzed, as it plays a crucial role in the analysis. As an empirical application, the RSM is used to nowcast the US GDP and Industrial production growth.The results suggest that the RSM generally surpasses the simple mean, with especially large gains when time-series dynamics deviate from ”normal” times, such as during the COVID-19 recession.
Presented by: Boris Kozyrev, Halle Institute for Economic Research (IWH)
Does Uncertainty Predict Recessions? Abstract Previous studies have documented a correlation between increases in uncertainty and the onset of recessions. We consider whether adding uncertainty to a standard recession prediction model improves forecast accuracy in real time. To do this, we construct various vintage uncertainty measures for both the overall and for policy. We can then evaluate (1) if uncertainty improves accuracy, (2) which uncertainty measure(s) improve accuracy the most, (3) whether uncertainty's predictive content is a substitute for other information, and (4) whether aggregating the information content in uncertainty further improves accuracy. We find that uncertainty does increase forecast accuracy at most horizons. However, at shorter horizons, adding contemporaneous output variables such as GDP growth dampen some of uncertainty's predicitve ability. We also find that, alone, the EPU generally performs better than other uncertainty measures but, unsurprisingly, composite measures also do well. Finally, we find that averages of a number of predictive models seem to perform the best.
Presented by: Michael Owyang, Federal Reserve Bank of St Louis
Session: E6: Empirical Macro & Time Series March 31, 2026 13:30 to 15:10 Location: Room 517
Session Chair: Laura Coroneo, University of York
Are Hysteresis Effects Nonlinear? Abstract Do temporary aggregate demand shocks have lasting effects, and are they asymmetric between contractions and expansions? Using U.S. data from 1983:Q1-2019:Q4, we identify demand shocks with potential long-run consequences via a Bayesian SVAR and trace their propagation with nonlinear local projections. We find that negative shocks dominate in the short run, but positive shocks build up over time and by the medium run generate equally persistent effects on output. We investigate the mechanisms behind this result and argue that positive hysteresis is transmitted primarily through the labor market channel: expansions durably lower long-term unemployment and raise labor force participation. By contrast, the capital accumulation and R&D channels transmit predominantly negative hysteresis.
Presented by: Omar Pietro Carnevale, Queen Mary, University of London
Across the borders, above the bounds: a non-linear framework for international yield curves Abstract This paper presents a non-linear framework to evaluate spillovers across domestic and international yield curves when policy rates are constrained by the zero lower bound. Based on the sample of US and UK data, we estimate a joint shadow rate model of international yield curves, accounting for the zero lower bound, no-arbitrage conditions within and between government bond markets, and the global nature of some of the bond risk factors. Results indicate that the post-2009 US monetary policy transmission mechanism and its spillover effects on the UK yield curve are non-linear and asymmetric.
Presented by: Laura Coroneo, University of York
Session: E7: Empirical Macro IX March 31, 2026 13:30 to 15:10 Location: Room 526
Session Chair: Knut Are Aastveit, Norges Bank
Expectation formation in Large Language Models Abstract This paper investigates the ability of Large Language Models (LLMs), specifically ChatGPT, to form inflation perceptions and expectations. We compare LLM outputs to household survey data and official price statistics, mimicking the information set and demographic character- istics of the Bank of England's Ination Attitudes Survey. Our quasi- experimental design exploits the timing between ChatGPT's training period and the subsequent UK inflation surge. We find that Chat- GPT tracks aggregate survey projections and official statistics. At a disaggregated level, the LLM replicates key empirical regularities of household inflation expectations, particularly when conditioned on age, income, and gender. ChatGPT demonstrates heightened sensi- tivity to food and energy ination information, often leading to over- estimation of inflation perceptions. These insights could be used to evaluate the behaviour of LLMs more generally, to (cautiously) use LLMs for the generation and cross-check of official statistics, including counterfactual statistics, and the more efficient and effective design of surveys.
Presented by: Nikoleta Anesti, Bank of England
How Substantial is the Difference Between How People Understand Monetary Policy and Fiscal Policy? Evidence from an Experiment Abstract Both monetary policy and fiscal policy play large roles in shaping the direction of economies. However, the goals of fiscal policy and how it influences the economy is easier for the general public to understand because they interact with its consequences (such as fluctuations in spending and marginal income tax rates) frequently, while monetary policy is more difficult for the general public to understand because the mechanisms are less obvious and people interact with them far less frequently. Via an experiment, we show that people do in fact understand monetary policy less well than fiscal policy, but this understanding is greatly improved by educational tools that the Federal Reserve has on offer. We conclude that the efficacy of monetary policy can be improved, and the gap in understanding of monetary policy and fiscal policy can be closed, if these tools reach the general public. This has important policy implications, as monetary policy may be less effective if individuals do not respond appropriately due to limited understanding.
Presented by: Ezgi Kurt, Bentley University
Fiscal Narratives and Inflation Abstract This paper investigates how media narratives on fiscal policy shape household’s inflation expectations. We collect a large corpus of newspaper articles report- ing on fiscal policy from four major German newspapers spanning from 2006 to March 2025. Using a large language model (ChatGPT) we introduce a strategy to automatically identify different fiscal narratives in text and construct narrative indices out of this data. We then estimate the effect of these narrative indices on household inflation expectations and find that they all have a positive significant effect varying in size. Lastly, we measure how fiscal narratives affect the trans- mission of a government spending shock to the economy and find that some of the narratives have an amplifying effect while others dampen the impact.
Presented by: Farah Tohme, Goethe University
Do macroprudential policies stabilize household demand during a downturn? Abstract In this paper we explore whether less leverage stabilize household demand during economic downturns using administrative data on household balance sheets from Norway. Focusing on a policy-reform in 2011 which lowered household leverage combined with the subsequent 2014 oil price collapse as a natural experiment, we document how households with lower leverage had more stable spending patterns during the recession compared to other households, and that these differential spending patterns are not explained by a large set of pre-determined household characteristics. We explore the channels through which macroprudential policies stabilize consumption and show that the stabilizing benefits are primarily driven by spending on durables. Finally, we explore the interaction between macroprudential and monetary policies.
Presented by: Knut Are Aastveit, Norges Bank

 

This program was last updated on 2026-03-30 05:41:28 EDT