October was supposed the month when volatility returned. Or maybe not. Actually no one believed that.
Seasonality made for some good “watch out” posts, but the reality of the situation is that between the central bank/market communication loop, the dynamics and perverse incentives that loop creates, and the vol. suppressing effect of indiscriminate flows, there’s no hope of vol. catching anything more than a fleeting bid.
Besides, as Goldman wrote a couple of weeks ago, October seasonality isn’t really a thing during low vol. regimes, a factoid that’s either interesting or completely tautological depending on how predisposed you are to cynicism. Here’s the chart:
Meanwhile, when it comes to the potential for an N-vol spike to cause problems, flows into levered long VIX ETPs have offset outflows from short products to ensure that the rebalance risk remains near record highs. Here are the updated charts on that:
But will it ever matter? Will, the vol. sellers ever, as Bloomberg’s Cameron Crise puts it, “get their comeuppance”?
Maybe, and here’s Cameron to explain why…
While October has a reputation for stock-market fireworks, this year the month has mirrored the rest of 2017: a grind higher in stocks that’s almost devoid of volatility on a close-to-close basis. It certainly feels like the market’s behavior this year, and indeed for much of the post-crisis period, has been abnormal relative to the standards of history. Central bank monetary policies offer a convenient scapegoat to explain the apparently abnormal market returns. While QE and forward guidance have certainly helped dampen volatility, in reality the S&P 500 has had a slightly higher standard deviation of returns in this decade than it did in the 1990s. However, the distribution is markedly different. Today’s markets feature a much skinnier body — and ominously for vol sellers, fatter tails.
- The relentless grind higher in stocks this year, and indeed for much of the post-crisis period, seems remarkable — but is it? Clearly volatility has declined since the last decade; the daily standard deviation of returns since 2010 is 0.94% versus 1.41% from 2000-09. However, volatility was even lower in the 1990s, with a daily SD of 0.89%. So why does the market tone feel so different these days?
- It turns out that the nature of trading has, in fact, changed. Relative to the 1990s, a much higher percentage of daily returns have come within half a standard deviation of average. At the same time, there have been a larger proportion of moves exceeding 2.5 SDs as well
- However, before you rush to blame central bank manipulation, a word of caution: in aggregate the concentration of returns (relative to the normal distribution) within half a standard deviation of average is only slightly higher than it was in the last decade: 15.7% to 15.5%. However, these numbers contrast greatly with the 1990s (11.1%) and 1980s (13.5%). Unfortunately, the 2000s provided ample evidence of what happens when a skinny body sprouts a fat tail
- What about other markets? Surely QE and forward guidance have generated the same impact on 10-year yields? Quite the contrary! In fact, the normalized distribution of changes in the U.S. 10y yield since 2000 more closely resembles a normal distribution than any of the previous three decades. That the 1980s features the most equity-like distribution of returns suggests that the study may simply be capturing the strong trend from very high yields to more moderate ones (with a lower daily change).
- As for the dollar, it probably exhibits the most consistent return pattern of any market — at least when measured by the DXY. There is little notable change in the concentration of “nearly unched” returns over the decades, though interestingly there has been a substantial increase in “fat tail” events relative to the last decade. That sort of makes sense; in a generally low-vol environment, things like SNB peg nonsense and sterling flash crashes really make a dent. The other interesting observation is how much flatter the distribution is relative to equities. Compared to the normal distribution, the current decade features an “excess population” within half a standard deviation of average of 7.7%, less than half of that of the SPX. Conspiracy theorists, make of that what you will.
- That market returns have a skinny body and fat tails should come as no surprise to anyone. That the current decade features an unusually high concentration of “nearly unchanged” S&P days should be obvious as well. Selling vol into oblivion is a great strategy as long as it works. Yet the data suggests that the nature of the market hasn’t really changed since the last decade. Vol shorts will get their comeuppance, even if it’s later rather than sooner