Home ECONOMY End-of-Month Liquidity in the Treasury Market

End-of-Month Liquidity in the Treasury Market

25


Trading activity in benchmark U.S. Treasury securities now concentrates on the last trading day of the month. Moreover, this stepped-up activity is associated with lower transaction costs, as shown by a smaller price impact of trades. We conjecture that increased turn-of-month portfolio rebalancing by passive investment funds that manage relative to fixed-income indices helps explain these patterns.

Trading Volume Concentrates on the Last Day of the Month

Since 2020, trading activity in benchmark Treasury notes and bonds has been roughly 33 percent higher on the last trading day of the month, on average, as shown in the chart below.

Trading Volume Is Higher on the Last Day of the Month

Source: Authors’ calculations, based on data from BrokerTec.
Notes: The chart shows the average percent deviation of trading volume on each day of the month as compared to the average for the same day of the week for the two weeks preceding and following that day. Days of the month are plotted relative to the last day of the month, with 0 being the last trading day and 1 being the first trading day. Volume is for the most recently auctioned two-, three-, five-, seven-, ten-, twenty-, and thirty-year nominal securities and the sample period is January 1, 2020, to July 31, 2024.

Moreover, as shown in the next chart, this end-of-month effect has been growing over time and was essentially nonexistent in the daily data before 2015.

End-of-Month Effects Have Been Growing over Time

Source: Authors’ calculations, based on data from BrokerTec.
Notes: The chart shows the average percent deviation of trading volume on each day of the month as compared to the average for the same day of the week for the two weeks preceding and following that day for the indicated time periods and benchmark Treasury notes. The sample period is January 1, 2005, to July 31, 2024.

To gauge the end-of-month patterns as far back as possible, we base our analysis on data from an interdealer broker whose coverage goes back to the early 2000s. Treasury TRACE, in contrast, better measures the breadth of trading in the market, including for the dealer-to-customer market and seasoned securities (see this post), but starts in July 2017. While there is no reason to think that our analysis is biased despite not viewing all market activity, month-end patterns may differ in other segments of the market.

Our analysis controls for day-of-week effects. This could matter because Friday is the last trading day of the month about three times more often than other weekdays (Friday is the last trading day for months that end on Saturday or Sunday). That said, activity on Friday is comparable to that of other weekdays, with benchmark note and bond volume about 4 percent lower than average. By comparison, volume on Monday is 12 percent lower than average, whereas volume on Tuesday, Wednesday, and Thursday is higher than average by 2 percent, 5 percent, and 7 percent, respectively.

Our analysis does not control for differences in end-of-month effects across months. Most notably, trading tends to be 16 percent lower on the last trading day of December because of the shortened holiday trading hours, lower trading desk staffing levels, and possibly positions that have been realigned in advance of that day. It follows that volume has been 37 percent higher on the last day of the month since 2020 if Decembers are excluded. We don’t find significant differences in end-of-month patterns for other months.

The end-of-month effect is robust to security type, as shown in the chart above. Two- and five-year notes are issued monthly, on the last day of the month, which might induce some monthly pattern in trading activity. However, the effect is roughly as strong for the ten-year note, which is issued mid-month. (Moreover, even for the two- and five-year notes, the more relevant auction days are concentrated two to three days before the end of each month, as shown in this paper’s Figure A1). Patterns are somewhat stronger for the less actively traded benchmark securities, which explains why the month-end effect in the first chart above is somewhat bigger than the effects observed in the second chart for the two-, five-, and ten-year notes (for the same 2020-2024 sample period).

Why Do These End-of-Month Effects Occur?

End-of-month effects have been studied across various asset markets and geographies, focusing mostly on prices and returns. For example, Ariel (1987) and Lakonishok and Smidt (1988) find higher U.S. equity market returns in the last few days of the month. Hartley and Schwarz (2019) and Etula et al. (2020) identify higher end-of-month U.S. Treasury returns and attribute them to price pressure from institutional investors’ trades and portfolio rebalancing. The former paper shows that net month-end purchases of Treasuries by insurers exceed net purchases on any other day of the month. Moreover, it finds that insurers that benchmark their performance more closely to indices show greater net purchases of the securities that are added to the index and that these purchases are concentrated on the end-of-month rebalancing date.

Evidence on end-of-month price patterns and returns, however, doesn’t immediately translate to higher than usual volume on the last trading day of the month. While not studying the impact on the aggregate market, Dick-Nielsen and Rossi (2019) examine the effects of corporate bond index rebalancing, also occurring on the last day of the month, and find that the volume of bonds that are excluded from the index is four to five times higher than normal.

Similar channels might be at play in the U.S. Treasury market. The high concentration of volume on the last trading day of the month and the increasing concentration over time coincide with the growth of passive funds that track index changes. For example, although still small as a fraction of U.S. Treasuries outstanding, exchange-traded funds (ETFs) that track Treasuries grew more than ten-fold between 2013 and mid-2024 as shown in the chart below, surpassing the two-fold growth of Treasuries outstanding over the same period. Asset managers are increasingly managing relative to indices that are rebalanced at the end of each month and this may be causing investors to increasingly trade at that time.

U.S. Treasury ETF Assets Under Management Are Growing Rapidly

Source: Authors’ calculations, based on data from ETFG and etfdb.com.
Notes: The chart plots the monthly average of assets under management of U.S. Treasury exchange-traded funds (ETFs). U.S. Treasury ETFs include the sixty-six ETFs included in the “Treasuries ETFs list” on etfdb.com and are described as ETFs that invest primarily in U.S. Treasury notes of various lengths. The sample period is January 1, 2013, to July 31, 2024.

How Is End-of-Month Liquidity Affected?

The relationship between trading volume and liquidity is not a simple one. Volume and volatility are positively correlated, and volatility and liquidity are negatively correlated (see this study, for example). One therefore might expect a negative relationship between volume and liquidity, and that’s what is seen at times of market turmoil, such as around the near-failure of Long-Term Capital Management (see this paper), during the 2007-09 financial crisis (see this paper), and during the COVID-19-related disruptions of March 2020 (see this paper).

In the case of month-ends, however, the purported reasons for the higher end-of-month activity are not information-based, and the higher volume is not associated with higher volatility. Higher volume that arises independent of volatility is associated with improved liquidity, consistent with larger Treasury issues, and the most recently issued Treasuries, being more actively traded and more liquid (see this paper and this paper).

It follows that liquidity tends to be markedly better on the last trading day of the month. Since 2020, price-impact coefficients for benchmark notes have been about 26 percent lower, on average, as shown in the chart below, indicating better liquidity.

Price Impact Is Lower on the Last Day of the Month

Source: Authors’ calculations, based on data from BrokerTec.
Notes: The chart shows the average percent deviation of price impact on each day of the month as compared to the average for the same day of the week for the two weeks preceding and following that day averaged across the benchmark two-, five-, and ten-year Treasury notes. Days of the month are plotted relative to the last day of the month, with 0 being the last trading day and 1 being the first trading day. Price impact is calculated for each day and security as the slope coefficient from a regression of one-minute price changes on one-minute net order flow (buyer-initiated trading volume less seller-initiated trading volume). The sample period is January 1, 2020, to July 31, 2024.

Moreover, this end-of-month liquidity improvement has been increasing in magnitude over time, in a manner akin to that for trading volume, as shown in the next chart.

End-of-Month Price Impact Effects Have Been Increasing in Magnitude over Time

Source: Authors’ calculations, based on data from BrokerTec.
Notes: The chart shows the average percent deviation of price impact on the last trading day of each day of the month as compared to the average for the same day of the week for the two weeks preceding and following that day for various time periods and benchmark Treasury notes. Price impact is calculated for each day and security as the slope coefficient from a regression of one-minute price changes on one-minute net order flow (buyer-initiated trading volume less seller-initiated trading volume). The sample period is January 1, 2005, to July 31, 2024.

Similar but weaker patterns are observed for other measures of market liquidity. Quoted depth at the inside tier has been about 6 percent higher on the last trading day of the month since 2020, on average, implying better liquidity, as compared to 9 percent lower between 2005 and 2009 (percent differences are first calculated for each of the two-, five-, and ten-year notes, and then averaged across them). Bid-ask spreads have been about 1 percent narrower on the last day of the month, on average, suggesting slightly better liquidity, as compared to 2 percent wider between 2005 and 2009. The weak end-of-month effects for spreads in particular are likely attributable to minimum tick sizes, which cause spreads to vary little outside times of market stress (see this paper, for example).

Implications

Based on this post’s findings only, one might conclude that the last trading day of the month is an especially good time to trade because of the day’s higher trading volume and lower transaction costs. However, the evidence of periodicity in returns from other studies suggests that advantageous times to trade vary for other reasons and differ between buyers and sellers. These monthly patterns also change over time, as shown in this post, warranting close watching of these patterns going forward.

Henry Dyer is a research analyst in the Federal Reserve Bank of New York’s Research and Statistics Group.

Portrait: Photo of Michael Fleming

Michael J. Fleming is the head of Capital Markets Studies in the Federal Reserve Bank of New York’s Research and Statistics Group. 

Photo: portrait of Or Shachar

Or Shachar is a financial research advisor in Capital Markets Studies in the Federal Reserve Bank of New York’s Research and Statistics Group.

How to cite this post:
Henry Dyer, Michael Fleming, and Or Shachar, “End‑of‑Month Liquidity in the Treasury Market,” Federal Reserve Bank of New York Liberty Street Economics, September 24, 2024, https://libertystreeteconomics.newyorkfed.org/2024/09/end-of-month-liquidity-in-the-treasury-market/.


Disclaimer
The views expressed in this post are those of the author(s) and do not necessarily reflect the position of the Federal Reserve Bank of New York or the Federal Reserve System. Any errors or omissions are the responsibility of the author(s).



Source link