货币政策传导的流动性状态依赖性（英）

The liquidity statedependence of monetarypolicy transmission by Oliver Ashtari-Tafti, Rodrigo Guimaraes, Gabor Pinterand Jean-Charles Wijnandts Monetary and Economic Department September 2025 JEL classification: E43, E44, E52, G12 Keywords: monetary policy, long-term real rates, limits toarbitrage, segmented markets BISWorking Papers are written by members of the Monetary and EconomicDepartment of the Bank for International Settlements, and from time to time by othereconomists, and are published by the Bank. The papers are on subjects of topicalinterest and are technical in character. The views expressed in them are those of theirauthors and not necessarily the views of the BIS. This publication is available on the BIS website (www.bis.org). ©Bank for International Settlements 2025. All rights reserved. Brief excerpts may bereproduced or translated provided the source is stated. The Liquidity State Dependence ofMonetary Policy Transmission Oliver Ashtari-TaftiLSERodrigo GuimaraesBoEGabor PinterBIS Jean-Charles Wijnandts BoE Abstract We show that monetary policy shocks move long-term government bond yields onlywhen market liquidity is high and arbitrageurs are well capitalized. Thisliquidity statedependenceoperates entirely through real term premia, not expectations. Using noveltransaction-level data on the US Treasury market, we find that arbitrageurs trade about40% more duration during FOMC meetings in high-liquidity periods. We proposeways of enriching standard term-structure models to rationalize our evidence thatconstraints on arbitrage capital suppress transmission. The results introduce newempirical moments for theories of limits to arbitrage, and underscore the role ofliquidity conditions in shaping the effectiveness of conventional monetary policy. Keywords:monetary policy, long-term real rates, limited arbitrage, segmented markets JEL Classification:E43, E44, E52, G12 1Introduction Monetary policy influences economic activity through its impact on short- and long-terminterest rates, with the latter linked to the former by no-arbitrage conditions (Woodford,2003). A large literature shows that limits to arbitrage—arising from capital or risk con-straints on intermediaries—can weaken these links. Yet little is known about whether suchfrictions systematically shape the strength of monetary policy transmission along the yieldcurve. We ask: does the impact of short-rate shocks on long-term bond yields vary withmarket liquidity and the availability of arbitrage capital? Using aggregate time-series data, we show that monetary policy shocks move long-termbond yields only in high-liquidity states, when arbitrageurs are better capitalized. In theseperiods, a 1 percentage point shock to the 1-year nominal forward rate raises the 10-yearforward by about 0.4 percentage points, with statistically significant effects persisting outto 15 years. This is an even stronger degree of monetary non-neutrality than documentedin the previous literature.1In contrast, the response beyond five years is close to zero inlow-liquidity states. The state dependence operates entirely through real rates—and withinthose, real term premia—with no effect on inflation compensation. The effects are highly persistent: in high-liquidity states, the impact on long-term realrates remains statistically significant for more than three months after the shock, andpasses through to mortgage rates. These results suggest that the availability of arbitragecapital shapes not only financial-market transmission but also real economic outcomes. To arrive at these results, we identify monetary policy shocks using high-frequencychanges in interest rate futures around Federal Open Market Committee (FOMC) announce-ments, following Nakamura and Steinsson (2018). Liquidity states are defined using theyield-curve noise measure of Hu et al. (2013), which captures deviations from a smoothyield curve and is tightly linked to the availability of arbitrage capital. We show that thismeasure is better explained by proxies for arbitrage capital, such as fixed-income arbitragehedge fund returns, than by alternative candidates such as macroeconomic conditions or volatility. This evidence makes clear that the state-dependent transmission we uncover isclosely tied to arbitrageurs’ capacity to intermediate. To link these aggregate results to trading behavior, we use transaction-level data on theUS Treasury market from MiFID II reporting by UK-based institutions. We identify counter-parties as arbitrageurs or other traders based on regulatory categories and trading patterns,following the preferred-habitat framework (Vayanos and Vila, 2021). Arbitrageurs tradeabout 40% more long-end duration (11–30 years) around FOMC meetings when liquidity ishigh—providing direct evidence that arbitrage activity drives the state dependence. We alsodocument that some investors display upward-sloping demand for duration, consistentwith the idea that “reach-for-yield” beh