This is part of a 5-part series of various execution styles for volatility strategies - full list here.
(B) VIX Futures
VIX Futures are the most direct way to express a view on VIX, but for those new to it, the danger lies in not understanding what it is and how the prices move, as it does not function like a normal stock.
For a surface understanding of what VIX Futures are, you can head here. This post is for how to go long/short the VIX using VIX Futures.
B1 - Tactical Long Volatility
It can be as simple as buying the front month (nearest expiring month) Futures. When VIX spikes, the nearer Futures (front month and a few months beyond) also move up - the nearer the expiration, the greater the exposure to the VIX's moves.
However, there exists a trade-off between risk and reward, as the nearer months also have greater price decay rates - losses due to the contango in the futures term structure. Choosing a further-dated future means less decay rate but also less beta (sensitivity) to VIX.
The exact relationship need not be calculated - it's probably not worthwhile doing, simply because no one can even predict the magnitude of the next VIX spike, let alone how each future will respond to that. It is better to plan to limit your losses according to the current contango premium across the duration you wish to speculate on.
A chart can be constructed as shown below to give a broad view of historical starting contango values of the front month (red line, starting average around 2.6), and expressed as a percentage (blue line, starting average around 20%).
Ultimately, VIX spikes are difficult to predict, both in timing and in magnitude, and so it is difficult to make a good trade on a risk-adjusted basis. Unlike a stock where you can buy and hold, VIX futures will cost you dearly for calling a spike too early, due to its horrendous contango decay rates.
And so if you're then thinking of taking the opposite side of the trade, look to the B4 section of this page for harvesting contango.
B2 - Tactical Short Volatility
This is after VIX has spiked, and you're looking to take advantage of its reversion to a lower level. This decline in VIX levels is an eventual and inevitable thing - markets historically have always returned to a state of calm. The tricky part is estimating how long it will take, and whether VIX will go even higher in the interim.
If you are able to accept large drawdowns, the reward is that this strategy has proven to be quite profitable.
After a spike, it is difficult to predict if VIX will continue to go higher or start to decline. Recently (mid 2016 onwards) VIX spikes have proven to be very temporal, and markets return to a calm state within 2-4 weeks, giving away a very quick and fuss-free profit to those who sold VIX on each spike.
In other times (end 2011), VIX have remained at a higher level for months, providing a lot of pain for short sellers. Additionally, when VIX is higher, its swings tend to be greater too (volatility of volatility is greater when volatility is higher), with daily moves of about 15% being quite normal.
The below chart shows the historical closing prices of VIX front month futures and its 5-day percentage change (i.e. holding it for a week). 15-25% increases are not uncommon, and it can breach 75% as what it did in Aug 2015.
One more point to note - in a VIX spike, the front month may be in backwardation instead of contango. That means if you're holding on to a loss in your front month short, you cannot 'rollover' your short position to next month hoping to get the same price and giving your short more time for price to recede. Each VIX futures' pricing is the market's estimate of what VIX will be on each futures' expiry date. There will be a relation between successive futures, but not a strong one, as each future is estimating a different thing.
This suggests that when gains are made, it's better to keep it and wait for the next move rather than try to squeeze out all the profits. Unless it is part of a meta-strategy that fits within a strategic long-term short-volatility strategy.
B3 - Strategic Long Volatility
This is not economical, as over the long term the contango drag always overpowers the short term gain earned in a VIX spike. Just take a look at the performance of long-volatility ETFs.
B4 - Strategic Short Volatility
A perma-short volatility strategy will perform in a similar way (very high correlation) with the S&P500, simply because option values are correlated with the movements in the S&P. Employing this strategy yourself using futures should produce similar results to an inverse vol ETF - something like XIV. You just have greater flexibility to add your own active, speculative component to it.
Adequate capitalization is required, as you will likely have to ride out the massive drawdowns when VIX spikes occur.
To avoid the losses that may occur due to inability to rollover (see above), it's better to sell the VF2 (second month) or VF3 rather than the front month. That way it gives adequate time for VIX to recede and you can find a better rollover price. Selling further out futures also have the benefit of being less volatile during spikes - see below distribution of 5-day holding period returns of VF1 and VF6.
Links to other strategies: