Some people regard the bond market as the most mature financial market, providing discipline and stability where stock operators merely exhibit fear and greed. I’m not out to settle any internecine disputes here, but I do want to examine a strategy premised on the view that behavior in the bond market might tell us something useful about the equity market.
Conventionally, investor preference for higher yield correlates with a positive overall economic outlook. Brett Steenbarger looks at this phenomenon over the past year, specifically in terms of the relative performance of investment grade and high yield bond funds:
One quick and dirty way to evaluate investor sentiment and risk appetite is to compare the performance of investment grade corporate bonds (LQD; top chart) with that of high yield corporates bonds (JNK; bottom chart). When investors are bullish on the economy and perceive relative safety in the economic/financial environment, they will reach for yield and buy high yield debt over investment grade alternatives. Conversely, in times of perceived economic danger, investors will tend to favor the relative safety of investment grade debt over more speculative alternatives.
The implication is that we should expect to see bullish moves in equities confirmed by outperformance in high yield bonds relative to their investment grade peers, and perhaps vice versa. If you think that buyers of corporate debt are at least as likely to be well-informed as buyers of equities, an equity rally that isn’t accompanied by a relative reach for yield may warrant some caution. To get a clearer view of this relationship, I plotted the ratio of the high yield to investment grade corporate debt (using the same funds that Brett mentions, so JNK/LQD) against the S&P 500 since January 2008:
That’s a pretty interesting relationship – on first glance it doesn’t refute our basic premise about the relationship between rising equity prices and larger appetites for riskier corporate debt. Some traders may be content to treat this as a useful indicator for assessing the strength of equity moves over weekly or monthly time frames, and for that purpose it’s also worth looking at the behavior of the individual bond funds. But I’m less patient that than, so I tested three daily variations on this theme.
Transaction costs, slippage, and returns on cash are ignored; returns are logarithmically scaled.
- “Junk leadership” asks whether the high yield/investment grade ratio (again: JNK/LQD, hereafter “bond ratio”) increased or decreased today. If it increased (decreased), we go long (short) the S&P 500 today at the close and unwind the trade at the close tomorrow. The blue equity curve below tracks this approach. As you can see, this approach had a few winning streaks but isn’t even profitable overall, so I think it’s safe to rule out a daily high-yield leadership scenario.
- “Junk leadership SMA” gets long equities when that same bond ratio is above its 20-day simple moving average, and is otherwise short equities. The red equity curve below tracks this approach. Smoothing out the data driving our trade signals improves performance somewhat, but the strategy is still basically a lot of chop with no discernible edge whatsoever. My thought at this point was that, since the two key variables (the bond ratio and S&P) move together most of the time, instead of trying to squeeze alpha from such a synchronous relationship, why not look for divergences instead?
- “Junk/SPX divergence” asks whether the signs of the daily change in the bond ratio and S&P 500 were the same. If they were, i.e. if the bond market and equity markets “agreed,” no signal is generated. However, if the bond ratio was up – meaning that more investors were interested in high yield corporate debt than were buying investment grade debt – and equities were down on the day, we go long equities at today’s close and exit on tomorrow’s close; if the bond ratio was down and equities were up, we short equities. The orange equity curve below tracks this approach. In other words, when bonds and equities disagree, we’re choosing to believe the bond market. After a quiet start, this strategy responded well during the financial crisis of 2008 and has performed steadily since then, nearly doubling the starting equity over the 17 months examined.
I wouldn’t necessarily trade this strategy as-is; for one thing, I’d want to know whether these bond market moves are predictive over longer periods. The small sample size (the funds used in this study aren’t very old) also decreases confidence, as does the fact that these are still unusual economic circumstances and the extremely short-term relationship being studied here may not continue to hold during a quieter or bull market.
Bonus points to anyone who can explain the significance of the top image without Googling for it.