Monday, November 10, 2014

The Beauty of Butt Ugly Markets

In the wake of much discussion of "V-bottoms", given the August and October experiences, I decided to take a look at V candidates and their prospective returns.  Above we see an overbought/oversold indicator, which consists of 20-day price change adjusted for the market's volatility.  So what we're seeing is how much directional movement up and down we're getting per unit of market realized vol.  Notice how this normalizes the overbought/oversold levels across very different markets from 2010 to the present.

When the volatility-adjusted OBOS measure is quite high, it means that we've traveled a good distance upward with relatively little corrective movement along the way. When we have a very low reading, it means that the market has come down a great deal with relatively little upward bouncing.  Those latter occasions are truly butt ugly markets:  ones in which short term oversold conditions stay oversold for weeks.  By the time we get readings that are very low in the OBOS measure, we can identify a V candidate without hindsight bias.

As you can see above, we are in overbought territory on the measure, but not at peak levels that have typified recent market cycles.

So how do overbought and oversold readings affect future market returns?

The top quartile of OBOS readings are the relative beauty pageant winners:  those markets have been trending strongly higher over a number of weeks.  Over the next five trading days, they have averaged a gain of only +.04%, significantly lower than the average gain of +.29% for the remainder of the sample.  Over the next 20 days, they have averaged a gain of +.38%, meaningfully lower than the average return of +1.20% for the rest of the sample.

And the bottom quartile of OBOS readings?  Those are the butt ugly markets and the V candidates.  Over the next five trading sessions, they have averaged a gain of +.46%, quite a bit higher than the average gain of +.15% for the remainder of the sample.  Across the next 20 days, the ugly markets have averaged a sizable gain of +2.12%, far higher than the average gain of +.62% for the rest of the sample.

So there you have it:  over the past several years, if you had bought the most attractive markets when things have looked best, you have dramatically underperformed.  If you had bought the ugliest markets when things have looked worst, you have dramatically outperformed.  Now, of course, there have been historical periods when ugly markets have gotten uglier before giving good returns; the 2008-2009 period comes to mind.  I strongly suspect, however, that ugly markets below a threshold VIX level are pretty good buying candidates on average, as those are most likely to represent ugly shakeouts in otherwise attractive markets.

Further Reading:  Why Trading Might Be the Most Difficult Job in the World