Tue, Oct 12, 2010 | Jared Woodard
This is the third post in the series I’m writing to introduce the VIX Portfolio Hedging (VXH) Strategy. The most important feature of the VXH strategy is that it offers protection against severe market declines. Almost equally important, however, is its performance during normal market environments.
A hedging strategy that profits during a market crash is no help at all if it imposes heavy costs the rest of the time; this is one of the chief limitations of conventional strategies. People suffer from more cognitive biases than it is possible to keep track of, and biases involving recency, optimism, and overconfidence mean that the real challenge for a hedging strategy isn’t just to protect against unexpected market shocks; the challenge is also to protect against the likely future folly of irrational human animals. Even though we clearly live in a crisis-prone world, and even though taking the full, unhedged brunt of a “fat tail” month can erase years of gains, it is ultimately people, not history, that make allocation decisions, which means objective probability is not the only relevant factor. A hedging strategy that is too expensive will quickly become a hedging strategy that is ignored. When evaluating any hedging strategy, therefore, it is essential to ask: how would the strategy perform in a crisis-free world?
Before looking at the VXH performance, we should establish some baseline for comparison. I’ve already shown that conventional hedging methods don’t work – not well, anyway, so those won’t do in this case. Since the key feature that makes VXH so cost effective is that it is designed to vary the level of allocation to hedging positions in response to the market environment, a suitable baseline would be a fixed-allocation version of the same strategy. The equity curve of a fixed-allocation version of the strategy is below.
The flat line from September through November in the middle of the chart marks the financial crisis of 2008 – the period in which every hedging strategy made enormous profits. If we set the returns from that period to zero, we get a clearer sense of what the costs of a strategy are during normal market environments. In the case of a fixed allocation to VIX futures, the results aren’t pretty. Although the history of the product is much shorter, investors holding VXX shares have experienced the same trend.
Let me emphasize one point, however, about why this chart can be misleading. The decline in implied volatility from early 2009 through early 2010 can only be understood as a direct effect of the crisis: you obviously can’t watch implied volatility fall from 60% to 20% over the course of a year or two unless you first see it rise to 60%. So, as dramatic as the decline in volatility in 2009 and 2010 has been, it should be regarded as a mirror image of the financial crisis – drawn out over time, perhaps, as in a funhouse mirror. In that sense, it’s a bit unfair to show the losses incurred by a strategy during a decline in volatility without including the gains that preceded them.
But no matter: the point of this exercise is to show that a variable approach to allocation provides a dramatic improvement, even excluding the period during which the strategy is designed to profit:
Notice that the equity curve of the variable-allocation VXH Program (dark blue) nearly touched 100 during the May 2010 flash crash. That means that even in spite of the severe decline in implied volatility subsequent to the financial crisis, and even if we exclude the outsized gains achieved during the crisis, it’s still the case that as of May 2010, this hedging strategy would not have imposed any significant costs since its 2006 inception. And at current levels, $100 allocated to this version of VXH (stripped of the gains from fall 2008) would be worth almost exactly what you would have earned by allocating $100 to the S&P 500 at the same starting point – and with dramatically less downside volatility. It would be going much too far to think about this or any hedging strategy in terms of a “free lunch”; capital devoted to hedging can’t be allocated to risky assets, and the strategy is certainly expected to impose some costs over time during strong bull markets. But VXH survives the “crisis-free world” test better than any other hedging strategy I’ve encountered, which means it has the intangible-but-valuable property of not expecting too much from cost-sensitive investors. When you consider the sizable and perpetual costs of option hedges expiring worthless, the prospect of a hedging strategy that does not impose such costs becomes even more compelling.
The notion of allocating capital to a strategy gradually is something I’ve written about before. VXH is successful precisely because it varies exposure in response to changes in the market. Having some minimal long volatility exposure constantly in place offers a level of protection that is distinct, in fact, from the large-scale economic crises that most observers worry over. During the “flash crash” of May 6, 2010, for instance, the strategy signal was only at about 10%, but even that was sufficient to cover half of the losses incurred by a target equity portfolio on the day. One-day shocks like May 6 or September 11 don’t give investors enough time to maneuver into hedge positions, which is why having a variable, low-cost strategy permanently in place makes a lot of sense.
The VXH strategy is designed for use with VIX and Mini-VIX futures, but for investors managing smaller accounts it can also be implemented using the VXX ETN. In the next post in this series, I’ll show how the VXH strategy is able to overcome the well-known limitations of VXX.