Does maturity matter? The case of treasury futures volume

• DOI: http://doi.org/10.1002/fut.22039
• Date: 01 October 2019
• Author: Alex Petkevich,Doina Chichernea,Kershen Huang
• Published date: 01 October 2019

J Futures Markets. 2019;39:1301–1321. wileyonlinelibrary.com/journal/fut © 2019 Wiley Periodicals, Inc.

|

1301

Received: 15 May 2018

|

Accepted: 8 June 2019

DOI: 10.1002/fut.22039

RESEARCH ARTICLE

Does maturity matter? The case of treasury futures volume

Doina Chichernea

1

|

Kershen Huang

2

|

Alex Petkevich

3

1

Reiman School of Finance, Daniels

College of Business, University of Denver,

Denver, Colorado

2

Department of Finance and Economics,

Huizenga College of Business and

Entrepreneurship, Nova Southeastern

University, Fort Lauderdale, Florida

3

Neff Department of Finance, College of

Business and Innovation, University of

Toledo, Toledo, Ohio

Correspondence

Doina Chichernea, Daniels College of

Business, University of Denver, 2101 S

University Blvd, Denver, CO 80210.

Email: doina.chichernea@du.edu

Abstract

We find that Treasury futures volume contains information about future

economic and financial market conditions. Short‐and long‐term volumes are

economically different: A relatively higher volume in short‐term (long‐term)

Treasury futures is counter‐cyclical (procyclical), preceding worse (better)

economic and financial conditions. Further, we construct a single factor from

futures volumes of different maturities that forecasts the performances of

Treasury securities and the corporate debt and equity markets, as well as

macroeconomic conditions. Our results are consistent with the notion that

futures volumes from different market segments reflect differences in beliefs

and contain different information about future financial and economic activity.

KEYWORDS

financial markets, macroeconomic conditions, treasury futures, volume

JEL CLASSIFICATION

G11; G12; G13

1

|

INTRODUCTION

Economists have traditionally relied on the prices of futures contracts to make judgments about the future market and

economic activities. Recent studies suggest, however, that futures volume and open interest contain information that is

not captured in futures prices and can be useful predictors. For example, Hong and Yogo (2012) build a theoretical

model showing that, in the presence of hedging demand and limited risk absorption capacity in the futures markets,

movements in open interest and volume are more informative signals than futures prices; they find that movements in

open interest and volume are procyclical and predict both future economic activity and asset prices.

1

A widely accepted interpretation of volume is that it reflects investors’disagreement in beliefs, such that heterogeneous

expectations about future economic conditions cause investors to trade with each other. Xiong and Yan (2010), for instance,

show that a reasonable amount of dispersion in priors can generate endogenous wealth fluctuations through the speculative

positions among economic agents. More recently, this line of literature has been growingly rich in emphasizing the term

structure of disagreements. To date, the role of disagreement at different maturities has been discussed in the context of federal

funds rates and GDP output (Andrade, Crump, Eusepi, & Moench, 2016), inflation (Andrade et al., 2016; Capistrán &

Timmermann, 2009), bond yields (Crump, Eusepi, & Moench, 2018), and exchange rates (Cao, Huang, Liu, & MacDonald,

2018). Evident from these studies, disagreements in the short‐and long‐term segments of a market are economically different.

In this paper, we build on the literature above and study the information in the trading volume of the US Treasury futures

market, conditioning on maturity. We posit that trading volumes contain information about upcoming financial markets and

1

The mechanism driving this predictability in the aforementioned model is essentially underreaction to news and asset prices. The underreaction typically arises due to the graduate diffusion of

information in the market (Hong & Stein, 1999, 2007). The authors argue that futures open interest and volume likely capture underreaction because it is highly related to both contemporaneous and

future news and returns.

macroeconomic activities, based on the argument that volume reflects investors’disagreement. We ask the following research

question: What do short‐and long‐term futures volumes tell us about the future financial market and economic conditions?

We are motivated by the idea that economic uncertainty and dispersion in beliefs come hand in hand. Under the

fundamental framework of Patton and Timmermann (2010), market participants make forecasts and take actions based

on a weighted average of information updates and prior beliefs. Since private information is of lower value for the

forecasting of longer term variables, holding all else constant, disagreement is stronger for long horizons. During bad

times, gathering information becomes relatively difficult due to greater uncertainty, which leads to a heavier reliance on

prior beliefs (Van Nieuwerburgh & Veldkamp, 2006; Veldkamp, 2005). To the extent that prior beliefs are sufficiently

dispersed, one would expect higher volumes at the short end of the market as the state of the economy deteriorates.

2

This is consistent with the “flight from maturity”phenomenon documented in Gorton, Metrick, and Xie (2014): In

expectation of bad times, trading flies from longer maturities to shorter ones, reflecting the overall increase in

uncertainty.

Oppositely, in expectation of the better market and macroeconomic environments, investors face lower uncertainty

and anticipate higher levels of economic activity and productivity (e.g., see Hong & Yogo, 2012). As a result, they have

the incentive to shift to longer term assets.

3

Aside from testing the implications of the “flight to maturity”argument, examining the maturity dimension of

Treasury futures volume also allows us to better understand the informational role of trading contracts at different

maturities. As investors transmit information regarding the level of heterogeneity in their beliefs through market

activities, the resulting volumes that are observable in different market segments (short‐term vs. long‐term)

consequently provide us with different predictions for future market and macroeconomic conditions.

Based on the above rationale, we hypothesize that an increase in relatively short‐term Treasury futures volume

precedes decreasing asset prices and deteriorating general economic conditions. Conversely, an increase in long‐term

futures volume is followed by higher levels of economic activity and in turn better financial and macroeconomic

conditions. Considering volume as a reflection of disagreement, our hypothesis is consistent with models such as that

presented in Cujean and Hasler (2017), which shows that during bad times there are spikes in disagreement across

shorter maturities, accompanied by increases in volume.

To test our hypothesis, we collect a unique sample of US Treasury futures volume for different maturities from

Bloomberg. We primarily rely on time‐series techniques (e.g., vector autoregressive [VAR] models) to test whether

trading activity in short‐and long‐term interest rate futures can predict asset prices and macroeconomic conditions. We

summarize the three key findings in this paper as follows:

1. The short‐term (long‐term) Treasury futures volume is negatively (positively) associated with future financial

market performance. We find very consistent predictive power of futures volume for the equity markets. For

corporate bond markets, the effect is evident mostly in junk bonds, where uncertainty is most relevant. On the flip

side, the short‐term (long‐term) Treasury futures volume is positively (negatively) associated with lead Treasury

securities performance.

2. The short‐term (long‐term) Treasury futures volume is counter (pro)‐cyclical. We find that an increase in relatively

short‐term futures volumes signals deteriorating future economic conditions, while an increase in relatively long‐

term volumes is associated with improving future economic trends (our macroeconomic variables include the

growths in GDP, consumption, industrial production, and unemployment).

3. We construct a single, common factor using futures trading volumes that has predictive power across the Treasury

and capital markets, as well as for macroeconomic conditions. This factor, defined as the difference between the

long‐and short‐term futures trading volumes, strongly predicts lower Treasury returns, better performance by low‐

rated bonds, and higher equity returns; it is also significantly procyclical in predicting the growths of GDP,

consumption, industrial production, and unemployment.

In sum, we identify the economic difference between trading volumes of short‐and long‐term Treasury futures as a

source of information about the future financial market and macroeconomic conditions. Controlling for other

2

In the extreme case, the no‐trade theorem states that if agents do not have differences in prior beliefs, transactions will not occur based on purely informational updates, even when information is

asymmetric (Milgrom & Stokey, 1982; Morris, 1994).

3

In the face of the recent subprime mortgage crisis, there is vast evidence about investors changing the maturity of their trading positions based on their expectations about markets and

macroeconomic activity. For example, Gorton et al. (2014) show that maturity of the debt instruments shortens dramatically during the crisis due to this “flight from maturity”effect. Similarly,

Krishnamurthy (2010) also argues that liquidity preferences caused shortening of the maturity in the debt markets.

1302

|

CHICHERNEA ET AL.