Momentum is perhaps the most basic of trading techniques, but it wouldn't seem to apply to a segment such as high-yield bonds. In tech or pharma, a stock price theoretically can continue rising forever, so it makes some sense that good news about a stock can keep compounding on itself, leading to ever-higher prices. But that doesn't fit the bond market.
When bond prices go up, the yield goes down. Hence, the bond is demonstrably less attractive at the higher price. In addition, the price of the bond can't keep rising forever. Even government bonds have struggled to get below 0% yield. Certainly credit-risky bonds have a logical limit as to how low the yield can get. Indeed, it would seem more like junk bonds should be a mean-reverting market -- that is, a market where periods of above-average performance are followed by periods of below-average results.
But here's the reality. Junk bonds show very strong short-term momentum characteristics. The mean-reverting factor eventually does catch up with them, but only in very long cycles. This is relevant not just for those who might want to get involved in trading a relatively low volatility market, but also for those thinking about selling high yield at tight spreads. The absolute level of spread isn't much of a forward indicator of returns, whereas momentum is.
First, let's take a look at where we are. Below is the spread of the Bloomberg Barclays High Yield Very Liquid Index. This is the index used by the SPDR High Yield Bond ETF (JNK) . The "spread" is just the average yield gap between each junk bond in the index and a maturity-matched Treasury. In essence, it is the amount of yield you are getting paid to take the credit risk of these bonds.
Source: Bloomberg Barclays Indices
There we are, basically at post-recession lows. Reasonably, one could conclude that you just aren't getting paid enough to take the credit risk of junk bonds. But history shows that buying and selling high-yield debt based on where spreads are at the outset isn't consistently a winning strategy. The chart below is a scatter plot of the total return of the Bloomberg Barclays High-Yield Index six months forward (vertical axis) vs. the initial spread (horizontal axis), measured every month since 1994. To make it easier to see, I divided these into quintiles and color-coded them. The larger red dots are the medians of each quintile.
Source: Bloomberg Barclays Indices
The widest two quintiles -- i.e., the ones with the most initial spread -- offer the best median returns (blue and yellow dots). But strangely, the tightest quintile has significantly better returns than either the orange or grey quintiles in the middle. This is literally telling you that buying junk bonds when they are at their worst valuation is a better bet than buying them when they are at their average valuation. What is going on here?
You'll find the answer if you think about the underlying risk of any given junk bond: default. As long as the company stays in business, they pay you back and you make your return. That is your only real risk. If you are invested in a portfolio of junk bonds, such as an ETF, any given company's default doesn't matter that much. In that case, you are more focused on the general trend for defaults. Of course, defaults don't happen at random. They are clustered around recessions. That tells us that a diverse portfolio of junk bonds is basically going to trade on the odds of a near-term recession.
Here's where momentum comes in. The economy doesn't just flip from growth to recession on a dime. We go from growth to peak to slowdown to recession to recovery in a pretty consistent pattern. We don't know the timing, duration or severity of each step, but we know the sequence. Once you think about it this way, the quintile chart makes sense. The tightest quintiles are periods where the market is confident that a recession isn't around the corner. Sometimes the market is wrong, but usually it takes more than six months to go from everything being hunky-dory to a recession seeming imminent. So, buying when spreads are pretty tight usually means solid returns to junk bonds. By contrast, in those middle quintiles, it is often the case that the market is growing worried about a recession coming. Consequently, you'd hit those kind of spreads as you moved from peak to slowdown. Sometimes the slowdown doesn't actually materialize (such as what happened in late 2015 with the big oil price drop), but if it does, junk bond performance is going to be poor.
Junk bonds also show pretty good short-term momentum characteristics, too. Using daily data from 2000 to today, if you went long high yield every day that the spread was below the 75-day simple moving average, you'd have a 484% cumulative return, or an 18.4% annualized rate. Even if we exclude the highly volatile 2008-09, you'd still be up 274% in the aggregate, or 15.32% annualized, following this simple momentum strategy.
Note that the strategy doesn't work that great on the short side. When the spread is above the simple moving average, the cumulative return is 4%, or 0.73% annualized (again, excluding 2008 and 2009), so it isn't worth shorting. But what that does tell you is that nearly all the return from junk bonds comes during positive momentum periods.
Even if you are not a short-term trader, thinking through the consequences of this analysis can help you invest. Right now, waiting for the junk bond market to back up a bit before buying might be a losing strategy. The market probably isn't going to back up unless a recession looks more likely. By that time, you won't want to be a buyer.
This commentary originally appeared on Real Money Pro on July 19. Click here to learn about this dynamic market information service for active traders.