美联储-交易成本与场外国债市场的指示性流动性

Trading Costs v. Indicative Liquidityin the Off-the-Run Treasury Market*Oleg Sokolinskiy†July 7, 2025AbstractThis paper estimates trading costs in the off-the-run Treasury market using compre-hensive transactions data and machine learning techniques. The analysis reveals severalkey findings that enhance the understanding of the off-the-run Treasury market liquid-ity. First, the indicative bid-ask spread is shown to be a biased measure of liquidity, evenwhen not considering transaction volume. Specifically, bid-ask spreads systematicallyoverstate trading costs of more seasoned Treasuries, and the liquidity of benchmark, on-the-run securities affects how off-the-run bid-ask spreads map to trading costs. Second,the paper demonstrates that trading costs may scale non-monotonically with transactionvolume, which suggests selective, opportunistic liquidity-taking. Additionally, transac-tion size has greater effect on off-the-run securities’ trading costs when benchmark, on-the-run liquidity is lower. Finally, indicative bid-ask spreads may notably overstate trad-ing costs for larger orders of relatively less liquid securities. These findings contributeto our understanding of actual liquidity in the off-the-run Treasury market, while high-lighting the limitations of a traditional liquidity measure. By providing a more nuancedview of trading costs, this study contributes valuable insights for supporting financialstability and optimal asset allocation.Keywords:liquidity, Treasury market, off-the-run, effective bid-ask spreadJEL codes:G10, G12This article represents the views of the author, and should not be interpreted as reflecting the views ofthe Board of Governors of the Federal Reserve System or other members of its staff. I am grateful for helpfulcomments from and discussions with David Bowman, Dobrislav Dobrev, Maximilian Dunn, Sebastian InfanteBilbao, Peter Johansson, Don Kim, Edith Liu, Andrew Meldrum, Marius Rodriguez, and Min Wei.Board of Governors of the Federal Reserve System. Email: oleg.v.sokolinskiy@frb.gov1 1IntroductionThe U.S. Treasury market is the deepest fixed income market in the world. Market liquidity– ability to transact significant amounts without temporarily dislocating prices – allows theTreasury market to serve as a benchmark for valuation and a source of near risk-free collat-eral. However, Treasury market functioning became a concern at the height of the COVID-19crisis (e.g., see Logan, 2020; Fleming and Ruela, 2020; Duffie et al., 2023).1SVB collapse and the April 2025 trade tensions led to two more episodes of severely strainedTreasury market liquidity. Furthermore, the supply of Treasury securities has been increas-ing relative to the primary dealers’ ability to intermediate in this market (see Duffie, 2023,2025). As a result, the functioning of the Treasury market is likely to remain a key focus forboth regulators and investors.The Treasury market has two major segments that differ in their microstructure and, con-sequently, liquidity. First, there are on-the-run Treasuries – most recently-issued securitiesfor a given maturity date. On-the-run Treasuries have much higher trading volumes thanother Treasuries, with a significant proportion of trades being on centralized markets withfirm quotes.2However, on-the-run Treasuries are only a small fraction of the notional of alloutstanding marketable Treasury debt. Second, there are off-the-run Treasuries – all Trea-sury securities that are not on-the-run. In contrast to on-the-run securities, off-the-run Trea-suries trade infrequently in a decentralized, bilateral market providing indicative quotes.Due to this microstructure of the off-the-run Treasury market, its liquidity is difficult togauge.In this paper, I suggest an estimate of the effective the bid-ask spread measure (see Dem-1Adrian et al. (2025) considers Treasury market liquidity over a longer timeline starting from the GFC.2A firm quote with an associated volume is a commitment of a liquidity provider to trade in that size ifrequested to do so.3Indicative quotes are not firm commitments to trade, and the price of a potential trade is may be set uponfurther negotiations. The March 202332 setz, 1968) for the off-the-run Treasury market liquidity largely based on actual trades.4The core components of the suggested method for the assessment of actual trading costsin the off-the-run Treasury market are (i) reliance on off-the-run Treasuries’ transaction dataalong with firm quotes for on-the-run Treasuries and (ii) state-of-the-art machine learningapproach to fitting the term structure of benchmark, on-the-run rates. The method relieson the on-the-run Treasuries’ quotes as the source of most recent information on changes inthe benchmark interest rate term structure. In turn, periodic snapshots of indicative quotesenable the estimation of idiosyncratic price components of particular off-the-run securities.The machine learning-based method of Filipovi´c et al. (2022) for fitting the