The tremendous growth in the ETF market has created considerable concern about the risks inherent to the business model. Much of the coverage has focused on the liquidity mismatch: investors may sell their ETF holdings during the day, but the ETF itself may find it hard to sell its assets immediately.
This opens up the scenario that a fund may exhaust all its investor’s capital and have to liquidate its positions in a possibly illiquid and unfriendly environment. Today we present two such funds that at first glance do not look like they fall into a liquidity trap: XIV and SVXY. Rest assured, however, these two funds will suddenly and violently go to zero.
Both funds are very similar in structure: they maintain a short position in the front months of the CBOE VIX futures curve. Everyday they roll some of the short position from the front month to the next contract on the curve.
Lets examine the daily holdings of SVXY as of 11/18/2014:
As you can see, its structure is concise:
- Short 1670 Nov VIX Futures
- Short 31639 Dec VIX Futures
- $508,441,903 in cash
So how big a move in the front-end of the VIX futures markets would exhaust the $508 million in cash and force the fund into bankruptcy?
- VIX futures trade in 0.05 increments worth $50 each
- short 33309 total futures is $1,665,450 per 0.05 min tick
- $508,441,903 / $1,665,450 = ~ 305 min ticks = 15.26 VIX points
- current Dec VIX Futures price = 15.25
- This implies that a sudden move in the VIX futures to 29.5 would wipe out SVXY
Had either SVXY or XIV existed back in May 2010, it is unlikely they would have survived the flash crash:
On May 6th the front month VIX futures moved 15 points to a high above 40. This would have completely wiped out SVXY’s investors in one afternoon.
During the European crisis in the summer of 2011, VIX futures made another move that would have been catastrophic for investors in these funds.
October’s near crash was just a taste of whats to come for XIV and SVXY. The rally in volatility was slow motion compared to the 2 examples above. Nevertheless both funds saw their shares plunge by 40% before regaining some ground.
A universal rule for ETFs is that (at scale) its always more efficient to reproduce the funds holdings rather than own the fund. Fees are a big part of why this happens. Another is transaction costs and the convexity introduced by rolling contracts. Notice how both XIV and SVXY have partially rebounded, whereas the futures are unch?
When the spot VIX is low, and the VIX futures curve is normal and steep, short volatility funds like XIV and SVXY reap big rewards. Essentially, the expensive futures contracts decay into the cheap spot VIX. Since these funds are short the front end of the curve, they capture this decay into the spot.
However, when the spot VIX is climbing enough to invert the futures curve, these funds bleed cash. They lose money both through rising futures prices and through the inversion of the curve. The daily rebalancing from front month to next means the funds are constantly buying the front month calendar spread. When the curve is inverted this means buying at a high price and selling at a low price, mechanically, everyday.
Inverse volatility ETFs are likely to lose any upside correlation with the markets in the event of a “melt-up”, or rapid rise, in equity prices. Normally the VIX is invesely correlated with the market because prices tend to decline faster than they rise. In a melt-up scenario, price rises will lead to increased volatility, putting further pressure on these funds.
The ETF industry has yet to be tested in any trial by fire. Given the systemic nature of volatility markets, the effects of a forced liquidation could spill well beyond the VIX futures market. Liquidity providers would be hard pressed to handle tens of thousands of VIX futures contracts without moving the price significantly.
Moreover, this would overwhelm the capacity for hedging in the futures market itself; market makers would race to other markets to hedge their unwanted short volatility exposure. Given that this scenario would only occur when there is general panic in the markets, the collapse of an inverse volatility ETF would have unpredictable consequences.