Informed trading in government bond markets

Robert Czech, Shiyang Huang, Dong Lou and Tianyu Wang

Government bond yields serve as a benchmark for virtually all other rates in financial markets. But what factors drive these yields? One view is that yields only move notably when important news hit the market, for example monetary policy announcements. Others suspect that some investors have an information advantage due to their access to costly information (e.g. data providers) or more accurate interpretations of public information. In a recent paper, we show that two investor groups – hedge funds and mutual funds – have an information edge in the UK government bond (gilt) market, and that these two investor types operate through different trading strategies and over different horizons.

At first glance, it seems difficult for any investor (or investor type) to have an information advantage in the deep and liquid government bond market. Previous studies have indeed struggled to identify such an information edge, not at least due to a lack of comprehensive trading data. In fact, one needs to observe the identity of both counterparties to identify whether some investors are better informed than others. We use the rich regulatory ZEN database to obtain detailed information, including the identities of both counterparties involved in the trade, on virtually every transaction in the UK gilt market in the period from August 2011 to December 2017.

Equipped with this rich dataset, we sort all gilts into different groups based on the net purchases of hedge funds or mutual funds. We then test whether the group of gilts heavily bought by either type outperforms the group heavily sold in the following days or months.

We find that the group of gilts heavily bought by hedge funds outperforms the group heavily sold by 1.3 basis points (bps) on the following day, and by 2.9 bps in the following week (left-panel of Figure 1). Interestingly, this effect is completely reversed after one month. This indicates that the outperformance of hedge funds is not driven by superior knowledge of the gilts’ fundamental value, because such information would have a more persistent impact on returns.

Figure 1: Long-short portfolio returns based on daily order flows of hedge funds/mutual funds

The group of gilts heavily bought by mutual funds, in contrast, does not outperform the group heavily sold in the first ten days. However, we discover substantial outperformance over a longer horizon: the cumulative return grows to 6.5 bps by the end of month one, and to 15.6 bps by the end of month two (right-panel of Figure 1). If we extend the time period to the following twelve months, we see no evidence of reversal: the cumulative return after one year is nearly 1.3%.

We next turn to the sources of the information advantage of hedge funds and mutual funds. First, we analyse whether funds can predict the order flows of other investor types. We find that hedge fund trading is a strong predictor of future mutual fund trading. We hypothesise that hedge funds observe public data on capital flows in and out of mutual funds and infer which gilts mutual funds are most likely to buy in the next week. Hedge funds then profit from this by trading ahead of these predictable order flows. In contrast, mutual fund trading has no predictive power for future order flows of other investor types.

Second, we also link the funds’ outperformance to macroeconomic news announcements and future changes in short-term and long-term government bond yields. We find that mutual funds are able to predict future changes in short-term interest rates, and profit from this ability by shifting the duration of their gilt portfolios. Furthermore, both hedge funds and mutual funds earn significantly higher returns on days with macroeconomic announcements (such as inflation/labour statistics announcements or Monetary Policy Committee (MPC) meetings). For example, hedge funds earn nearly twice as much on announcement days (2.50 bps) than on non-announcement days (1.28 bps). Even more strikingly, mutual funds earn 3.62 bps/day on macro-announcement days and only 0.5 bps/day on other days. Interestingly, both types of funds seem to earn higher returns on labour/inflation statistics announcement days than on monetary policy announcement days.

As an important cross-check, we extend our analysis to non-dealer banks, insurers and pension funds. Perhaps unsurprisingly, the net purchases of these investor types have insignificant (and sometimes even negative) predictive power for future gilt returns. This could be due to the fact that those investors have different objectives (such as liability matching in the insurance sector).

All in all, our paper reveals that hedge funds and mutual funds have a superior ability to collect, process, and trade on information that is relevant for future gilt returns. We also highlight the differences in their approaches to earning above-average returns in the government bond market. Through their active trading, these professional money managers help to incorporate relevant information into gilt yields and accelerate the price discovery process in one of the world’s most important financial markets.

Robert Czech works in the Bank’s Financial Stability Strategy and Risk Directorate, Shiyang Huang works at Hong Kong University, Dong Lou works at London School of Economics and Tianyu Wang works at Tsinghua University.

If you want to get in touch, please email us at bankunderground@bankofengland.co.uk or leave a comment below.

Comments will only appear once approved by a moderator, and are only published where a full name is supplied. Bank Underground is a blog for Bank of England staff to share views that challenge – or support – prevailing policy orthodoxies. The views expressed here are those of the authors, and are not necessarily those of the Bank of England, or its policy committees.