Hedging Tail Risk with the VIX

Felix Salmon is doubtful about whether it is possible to hedge tail risk, and I wholeheartedly agree with the data he cites showing that, of eight major asset classes, only volatility and managed futures offer genuine non-correlation to market returns. In fact, I'll go a step further: I'm not that enthusiastic about the benefits of managed futures, at least in their current form. As a registered commodity trading advisor, I've seen the sorts of strategies that most of my peers are using, and I think it's safe to say that "managed futures" would be better classified as "long/short trend following." Trend following is a fine approach, and is certainly better than being 100% long-only, but I doubt whether most trend-following managers really deserve a 20% incentive fee. If there are managed futures ETFs or ETNs with reasonable expenses, they may be worth a look, but otherwise, I agree that managed futures are out of reach for most investors.

That leaves volatility as the only asset class with a realistic chance of letting you hedge tail risk.

Felix says, "The problem with trying to invest in asset classes like volatility or managed commodity futures, of course, is that it's expensive and difficult to do so. You can't just go out and buy an ETF." But, of course, you can buy two such ETFs (technically ETNs): their tickers are VXX and VXZ, and they've become extremely popular over the last year. Since VXX wasn't launched until January 2009, it isn't possible to show its effectiveness as a hedge in 2008. However, I think a reasonable proxy for our purposes would be the second-month VIX futures contract. In the chart below, I've split a portfolio 90%/10% between dividend-adjusted SPY shares and the second-month VIX futures contract, rolling the contract the day after expiration (i.e., rolling on the day it becomes the front-month contract). The equity curve of the VIX position is shown on a separate axis.

As you can see, even a modest 10% allocation to a long volatility position offered some protection in the fall of 2008: where the buy-and-hold portfolio was down 35% from its 2006 inception, the hedged portfolio declined only 22%. I debated whether to show other fixed allocations, but decided against it, since it isn't realistic to ask an equity investor to devote 30% or more of his capital to hedging in perpetuity. Not to get too deep into the details here, but I think the real solution is to keep some form of minimal hedge always in place, but then to scale hedge positions up and down as volatility ebbs and flows, on the principle that truly abnormal volatility tends to beget more volatility. The chart below applies just such an approach (unoptimized, and with only two parameters).

Lest you think these returns were achieved by simply ratcheting up the VIX exposure, note that the average allocation over the period shown was 13% - only slightly higher than the original study. Roughly similar results can be achieved with the very liquid VXX product. I'm not the greatest fan of the Barclays volatility ETNs: the roll yield that the ETNs pay to keep a constant duration actually makes them relatively expensive to hold, and a good target for short-sellers who don't care about tail risk or have their own solutions. But the roll yield problem would be somewhat alleviated by the variable allocation approach I just mentioned, since very large allocations to volatility assets are called for less than 20% of the time, and less than 5% of capital is allocated to hedging over 25% of the time.

Actively rebalancing a futures-based portfolio hedge on a daily basis is well beyond the capacity of many investors, but trading in and out of some VXX shares a few times a month strikes me as a fair demand. The age of buy-and-hold-with-blinders-on is clearly over, and investors should now consider it a fiduciary duty to find some way to manage tail risk. On this point, I think I agree with Felix: if you can't find a way to credibly and economically insure yourself against catastrophe, you're better off leaving the asylum that is financial capitalism entirely.