It’s all about “loops” and “spirals” these days.
On Tuesday some folks were out warning that if the NAV discount on high yield ETFs gets too wide, it could conceivably lead the arb crowd that’s implicitly responsible for making sure everyone’s worst ETF fears aren’t realized to act irrationally thus short-circuiting the safety valve in a devastating “spiral.” You can about that in the high yield context here and in the EM bond context here.
When it comes to “loops” the discussion very often centers around the doomsday-ish self-feeding dynamic that’s been embedded in markets thanks to the interplay between central bank forward guidance, VIX ETPs, and vol.-sensitive, programmatic strats.
Hopefully you don’t need a refresher on this, because if you do it likely means you haven’t been paying much attention to how shifts in market structure are creating systemic risks that no one understands, but just in case, the idea is that thanks to the low starting point on the VIX, a nominally small spike could force inverse and levered VIX ETPs to panic buy VIX futs into said spike, thus exacerbating the situation and ultimately forcing CTAs, vol. control funds, and risk parity to deleverage into a falling market.
You can call that far-fetched if you want and you can also call the concerns overblown, but what you can’t do is argue that no one is talking about precisely that dynamic because they most assuredly are.
What you generally want to keep track of with this is the ETP rebalance risk (vega-to-buy on a N-vol. spike in the VIX futures curve) and the extent to which the systematic crowd is exposed. Deutsche Bank tracks this, and their latest read on the feedback loops is not what you would call “comforting.”
Implicit in the above is the idea that VIX ETPs cannot themselves tip the first domino, only that they can make things worse once the first domino is toppled. As Deutsche writes, the rebalance risk is still very large. To wit:
Impact of VIX ETPs remains high even as vol has risen slightly. VIX ETP providers would need to buy ~$140mm vega on a 5-vol spike in the VIX futures curve, which is ~65% of the average daily 1st/2nd/3rd VIX futures volume over the last 2 months. Even a 2-vol move would induce $70mm of buying. That 2-vol move would not represent a significantly high level of volatility historically, and may not require a large spot move to realize. A 5-vol move represents more of a tail risk given current volatility, however we saw a 2-vol move in the weighted average VIX future on May-17 and Aug-10.
That’s been a constant for some time now, but as DB goes on to write, CTA exposure (and implicitly, the risk associated with that exposure) is now at historically high levels. “Positioning of CTAs may pose a risk to SPX,” the bank writes, before elaborating as follows:
CTAs beta to SPX has risen to the top of its 3 year range over the past month, so there is some risk that the funds could sell a large number of futures into a declining market if there was shock and subsequent deleveraging.
As for vol. control funds, DB estimates they would need to sell $9-12 billion in equities during a 3% selloff. Here’s the feedback loop dashboard:
On the bright side, we’re in the midst of the second-longest streak without a 3% drawdown since WWII, so you know, it should be fine. Hell, we can’t even get a 0.5% move lower these days…