Stocks quoted in this article:
Well, while you were watching your NCAA brackets go to dust, the Chicago Board Options Exchange (CBOE) was busy celebrating 10 years of CBOE Volatility Index (VIX) futures.
There were 120 financial professionals that attended. I wasn't one of them -- it's a bit of a commute to get to Chicago from New Jersey (it took me a full 24 extra hours to get home the last time I tried it … really). But fortunately, the CBOE took some notes on the presentations. Included was a summary of comments by "Buzz" Gregory, a managing director at Goldman Sachs.
Let's go over some of them.
- VIX spot is not directly tradable.
- Access to the VIX market comes from trading listed VIX products. The performance of VIX ETFs, ETNs, and VIX options is dependent on the listed VIX futures market, so understanding the dynamics of VIX futures is crucial for successful trading.
I frequently get asked about trading "spot" VIX. It's just not possible. There are all sorts of day-of-the-week, time-of-day, and time-of-cycle quirks that are basically impossible to eliminate from the VIX you see on the board as it's presently calculated.
When the CBOE revised the VIX methodology a decade ago, they measured "time" by the hour instead of the day. An option has more time value on the open of trading than it does on the close, simply because there's one fewer trading day at that point. The old VIX didn't consider that, the new VIX does. That certainly helps smooth the VIX, but it's still not perfect. Hours after the regular-market close are simply worth less than hours within the regular trading day, so VIX drifts from open to close, just less dramatically.
And that's minor relative to the Friday and Monday effects. All things equal, traders reduce bids ahead of Friday and holiday-eve closes, and then on Monday, time catches up. If you had a "spot" VIX, you could simply buy it on Friday's close and sell it on Monday's open, and gain an average of about 2% per week. Of course, no one would ever take the other side of that trade, which is all a long-winded way of saying the VIX really can only sensibly trade in futures form.
- VIX futures began trading in March 2004 and have now been tested in both low and high volatility regimes.
This is an important point. The past decade has seen record highs* in VIX (*it wasn't invented yet in 1987, back-calculated VIX was higher in 1987), as well as single-digit VIX. VIX futures withstood both extremes. When VIX exploded, futures traded at steep discounts most of the time. Those discounts were correct … in fact, VIX futures were generally good shorts even at serious discounts to "cash." In fact, the best trade of all was buying actual near-term S&P 500 Index (SPX) options at gigantic implied volatility versus shorting longer-term SPX paper or VIX calls or VIX futures. The realized volatility in the market was so high that you could work off your near-term options by trading, and sit with your longer-term volatility short.
On the flip side, VIX futures have persistently overpriced "low" VIX. I've noted this a million times or so; they always expect a VIX rally over the next half year, but all they get are blips and fails.
It's important to note that iPath S&P 500 VIX Short-Term Futures ETN (VXX) started in 2009, and has really only seen a low-VIX regime. It will presumably do well in a high-VIX regime characterized by persistent backwardation in VIX futures, though we'll have to wait and see exactly how well.
- The relationship between spot VIX and VIX futures levels is highly correlated to the term structure of S&P 500 implied volatility.
- If the term structure is steeply upward sloping the VIX future will typically be trading above VIX spot.
- The steeper the term structure, the higher the basis ( Basis = future - spot ).
- High basis environments have tended to be profitable for shorts, bad for longs
OK, guess I delivered the punch line on this already. But the low VIX has often led to very steep term structures, which have almost universally been good to short into. There's no cost of capital reason for VIX futures premiums as in futures on other financial products. It's purely sentiment. That sentiment is relatively muted now, but it's still predictive of a lift in volatility. Like he says, the VIX tends to mirror the term structure in index options. The difference, though, is index options can benefit from either spikes in realized volatility or persistent directional moves like we saw in 2013. VIX futures do not; they need an actual lift in implied volatility. Short-term spikes help, but not all that much, especially in longer-term VIX futures, which are already pricing in those spikes.
Disclaimer: Mr. Warner's opinions expressed above do not necessarily represent the views of Schaeffer's Investment Research.