Is the VIX Curve Too Flat?
For the last couple of weeks, VIX and VIX based ETFs and ETNs (UVXY, TVIX, VXX, XIV, etc) have only been on the fringe of radar screen. They were there, but they were only mildly interesting.
What I had noticed at the time was:
1. VIX ETPs were seeing shares outstanding increase as VIX dropped (doubling down)
2. Then last week as VIX was rising I commented on the fact that funds were not seeing large decreases in shares outstanding – a signal that the ‘hedgers’ were prepared to be resilient
3. Over the past few days I have had more and more discussions about VIX again and think the one thing worth pointing out for those who like to dismiss it as a ‘pure retail’ product is that UVXY alone has about 25% of the outstanding Nov VIX Futures contracts and almost 30% of the outstanding Dec VIX Futures contracts. That is a large percentage of any market – let alone one where the product doubles down on direction each night to maintain its leverage balance. It has a market cap of $800 million, but since that is 2x leveraged that represents $1.6 billion of VIX exposure. VXX is unleveraged and seems about 6 to 8 times as volatile as the S&P 500 in terms of realized vol. That to me, means UVXY controls roughly $10 billion of S&P 500 equivalent risk – less easy to dismiss – especially given the daily volumes, the outsized ownership of the futures and the overall complexity of the product.
Unlike Treasury VIX which I hit on last week (which I think played a roll in Treasury and Credit Market price action last week), the absolute level of VIX isn’t particularly scary – its not unusually low or at fear levels.
But the “Kiss of Death” for the VIX products is the steepness of the VIX curve not the absolute levels in any case. That curve hammers the VIX based products (which is why the ETFs and ETNs are inevitably forced to do extremely large reverse splits). I fully expected the VIX curve to be steep and was going to highlight that as a problem, but it was flat – eerily flat.
Flat Like August 2015?
I started to analyze the data. The shape of the VIX curve is correlated to the absolute levels of VIX. Whenever VIX spikes (about 20) it often becomes inverted.
Low levels of VIX tend to have the steepest curves as people hedge some event on the horizon even if there is nothing to worry about today.
So I took all the days since January 2015 when the VIX had closed between 15 and 17.5 (Friday’s close of 16.2 is a rough midpoint).
There were 110 such days. 3 days in October, including Friday, ranked in the top 14 flattest curves. So we have a cluster of relatively flat curves given the current level of VIX.
More striking was that of the remaining 11 days that were flatter, 6 occurred in July and August 2015 – including August 19th.
VIX in the summer of 2015
Is the combination of a potentially deep and committed hedging community, coupled with self-reinforcing daily rebalancing where the election news cycle seems to be getting more surreal if anything going to lead to another spike in VIX?
I don’t know, but it is yet another thing to be cautious of.