Yinxi Xie - Research

Abstract: We show that U.S. trade policy surprises account for a large share of the variation in U.S. Treasury convenience yields during 2025. Using transaction-level data from Polymarket’s tariff-related prediction markets, we construct a novel high-frequency measure of trade policy surprises. Leveraging narrow event windows around unscheduled trade announcements, we provide causal evidence that unexpected U.S. tariff increases weaken the convenience yield of U.S. Treasuries. Positive tariff surprises persistently reduce convenience yields across all maturities, with the underlying mechanisms differing by maturity through distinct safety and liquidity channels. Beyond Treasury markets, trade policy surprises also induce a persistent decline in equity prices, sustained increases in the VIX and text-based trade policy uncertainty, higher short-term inflation expectations, and a higher perceived probability of recession.

"On the Wedge between the PPI and CPI Inflation Indicators," with Shang-Jin Wei.

R&R at American Economic Journal: Macroeconomics. This version: June 2024. [NBER WP][Press: Brookings, NBER Homepage]

Abstract: While two strands of the literature suggest that producer price index (PPI) inflation, in addition to or instead of consumer price index (CPI) inflation, should be a targeting variable in a monetary policy rule, the distinction between the two is only important when they do not co-move strongly. Our first contribution is to document that their correlation has indeed fallen substantially since the start of this century. Our second contribution is to propose a model to understand this divergence based on expanding global supply chains. Our theory also produces additional predictions that are confirmed in the data. As such changes are structural rather than temporary, and so does the co-movement between CPI-PPI inflation, it implies that central banks need to take the change in the production structure of the economy into account when deciding their inflation target.

"Understanding Inflation Dynamics: The Role of Government Expenditure," with Chang Liu.

R&R at European Economic Review. This version: May 2025. [SSRN][BoC WP][BCB Conference video][Press: The Canadian Press]

Abstract: This paper studies the impact of government expenditure on inflation through the lens of an augmented Phillips curve derived from a structural New Keynesian model. Our estimation results, based on external instruments, show that the augmented Phillips curve has a flatter slope than the canonical specification, and government expenditure has a negative coefficient. Changes in government expenditure account for a substantial portion of inflation variations and provide new insights into the "missing disinflation" puzzle. We also observe a negative response of both inflation and inflation expectations to government expenditure shocks, which can be rationalized by the supply-side channel that we emphasize.

"Fiscal Policy: Financing and Indebtedness," with Jing Cynthia Wu, Shihan Xie, and Ji Zhang.

This version: November 2025. [SSRN]

Abstract: We conduct a large-scale information randomized controlled trial to study fiscal policy impacts. Surveying approximately 9,000 households across five eurozone countries with varying debt-to-GDP ratios enables a clean cross-country comparison. The key finding is that the fiscal multiplier depends jointly on the financing method and the country’s debt burden: multipliers are smaller in high-debt countries when debt-financed but remain similar across countries when tax-financed. Finally, we develop a New Keynesian model featuring fiscal discipline, which reproduces the empirical patterns observed in the survey and highlights their underlying economic mechanisms.

"Unconventional Policy and Inflation" with Jing Cynthia Wu and Ji Zhang.

This version: January 2026. [SSRN][NBER WP][Press: Yahoo! Finance]

Abstract: We study the effects of unconventional monetary and fiscal policies on inflation dynamics. Using multiple complementary empirical approaches, including event studies, vector autoregressions, and regional panel regressions, we find little evidence that these policies contributed meaningfully to inflation in the past decades. The key economic mechanism operates through a disinflationary supply-side channel in the Phillips curve, which offsets the upward pressure on inflation from the usual demand channel. We demonstrate this mechanism both theoretically and empirically.

"Fiscal and Monetary Policy Interaction under Limited Foresight."

Under revision

Abstract: Analyses of the interaction between monetary and fiscal policy often turn crucially on assumptions that are made about outcomes far in the future, sometimes infinitely far. This is a problematic feature of rational-expectations analyses, given the limited basis for assumptions about the distant future. This paper instead considers both short-term effects and long-run consequences of alternative monetary and fiscal policies under an assumption of bounded rationality. In particular, it assumes that explicit forward planning extends only a finite distance into the future, with anticipated situations at that horizon evaluated using a value function learned from past experience. Such an approach makes announcements of future policies relevant, but avoids the debates about equilibrium selection that plague rational-expectations analyses. The combined monetary-fiscal regimes that result in stable long-run dynamics are characterized, and the effectiveness of temporary changes in either type of policy as a source of short-run demand stimulus is analyzed. The effectiveness of a coordinated change in monetary and fiscal policy is shown to be greatest when decision makers' degree of foresight is intermediate in range (average planning horizons on the order of ten years), rather than shorter or longer.