If you've spent more than 5 minutes doing research on the cadre of exchange-traded products available on the VIX futures, you are aware of the well-known cost to holding a position in the long and double-long products like VXX/VIXY and TVIX/UVXY. To say these products are long-term money losers is a slight understatement. All you really need to do is pull up a since-inception chart on any of those above products to see what I mean. (Sidebar: As bad as VXX is over the long term, it's not all a short seller's paradise…many people may not realize that a long VXX position performs just as well on a risk-adjusted basis during the shorter vol-spike periods as a short VXX position does during the much longer, drag-me-to-hell-my-position-is-going-to-zero periods. Another writing for another time.)
Now it's one thing to say these products cost you money to buy and hold, but a good to question to ask is exactly how much do these products cost you as a form of equity portfolio insurance? (Now the simple answer is way too much. Which is why any good professional options trader who halfway understands these products is simply going to short the crap out of them. But let's quantify regardless.) To do this, you can simply look at a price chart like I mentioned above, but that analysis becomes muddied by movements in the VIX, which is the pricing basis for the futures contracts these products hold. We need an analysis purely of the cost of this insurance policy that adjusts for the fact that it occasionally "pays off" during periods of heightened market volatility.
To this end, I did a simple study that isolates the return components on a buy-and-hold VIX futures position between movement in the VIX itself and the cost of holding a futures to expiration and rolling the position into the next month of the expiration cycle. This latter component is what we want to look at. Last year was a particularly instructive example of just how expensive holding these products can be. The VIX began 2012 at 23.40 and ended the year at 18.02, a drop of 5.38 VIX points. Not entirely damning until you consider that an equivalent VIX futures buy-and-hold registered a drop of 34.65 VIX points, which implies that the cost of carrying that position was a whopping 29.27 VIX points. Another way of looking at this, is that for your insurance policy to at least be break-even (forget about actually hedging the portfolio) at the end of the year, the VIX would have needed to end the year at 52.67. I can assure you this, your equities would be in some pretty severe red for the VIX to be that high, and what's mind blowing about these numbers is that is the VIX level where this trade begins to hedge your equity portfolio's year-end value.
I'll spin it to you another way. Using historical estimates on the statistical relationship between the VIX and the S&P 500, let's apply a beta of -4.0 for the VIX relative to the S&P 500. Roughly speaking, we can hedge $4 of equity exposure with $1 of VIX exposure. Based on the value of the VIX of 23.4 at the beginning of 2012, I could have fully hedged roughly $94,000 of equity exposure for the year by buying a single VIX future at that time. Over the course of the year, this hedge would have cost me $34,650, or roughly 36.9% of the value of my equity holdings. By comparison, if you held the S&P 500 for the year, you would have earned an even 16.0%. This hedged portfolio should really give you nothing, but in reality it loses almost 21%. Not exactly a reliable hedge. Of course, my beta estimate for the VIX could be off, but the empirical beta from this example would have to be at -9.5 for the hedge to work properly in 2012. A beta of 9.5 seems, well, a little high.
If you want to think of these products as insurance for the long-term, you are more than welcome to do that. But it's insurance you want to be selling, not buying. The reason is that the premiums for this insurance historically have been so severe that even large "payout" events don't even eat all the way into the collected premiums…an insurer's dream (30%+ years like 2012 add up pretty quickly). To be fair, 2008 represented the worst case scenario on this sort of equity insurance policy. If you look at a VIX futures buy-and-hold all the way back to the inception of tradeable VIX futures, 2008 did mark a period where the insurer takes losses on his own money. But at that point the accumulated profits were so large that any losses of principal were of a limited nature and were eventually made back and then some. For the full data set going back to March of 2004, the VIX has increased 1.28 points…essentially unchanged over the last 10 years. By comparison, the futures have yielded a loss of 111.53 points for the whole period.