When stripped of the technicalities, the premise is pretty simple as these things go. Levered and inverse VIX ETPs would be forced to rebalance on a nominally small vol. spike and to the extent that rebalance exacerbates said spike, vol. control funds, CTAs, and risk parity could be forced to deleverage into a falling market.
This has become one of the rallying cries of the "Chicken Little" contingent and it's a topic that JPMorgan's "Gandalf" (Marko Kolanovic) has been keen on warning about for quite some time. The quant crowd isn't fond of the notion that they'll be an aggravating factor in the next plunge and the VIX ETP crowd doesn't generally understand the role they're playing in what amounts to a buildup of systemic risk.
There are plenty of people who will tell you why none of what I've just said is possible and indeed, there are some folks who will claim it doesn't make much sense at all, but I would submit that those rebuttals will one day be looked back on as something akin to "famous last words." So you know: "time stamp it."
Well in a sweeping new note out Tuesday, Goldman takes a look at this dynamic asking "would automatic flows exacerbate a potential selloff?"
"In moments of stress, products that de-risk when risk rises (including volatility control strategies and trend-following strategies) can create feedback loops," the bank notes, echoing our "doom loop" characterization. Here's some further color:
Volatility control strategies typically sell equities when vol rises above a threshold, but with vol well below-target a prolonged pickup in vol would likely be needed. The S&P 500 Managed Risk – Moderate Index, which consistently would have had 7-10% 12-month rolling realized vol in its history, has now had 5% realized vol over the past 12 months. A quick selloff would drive this index and the large funds that have similar strategies to start realizing near-target vol, rather than above-target, so we would not expect them to de-risk in an initial pickup in vol.
Growth in levered, short VIX ETPs has increased their potential feedback loop impact. Over the past few months, the combination of product inflows (to long VIX ETPs, as a hedge), strong performance (of short ETPs, following +135% YTD performance), and low VIX futures prices (allowing more futures to "fit" inside each ETP share) have brought the gross VIX ETP market size to a new high (the net size with offsetting long and short products is less long than usual).
That second paragraph underscores the notion that when it comes to the "doom loop," VIX ETPs are the biggest risk. Here's Goldman again:
The automatic flow that has changed the most of late – and that poses the most immediate risk – is VIX ETPs, after recent growth of the leveraged and inverse products.
Large levered and inverse VIX ETPs help make vol move faster. When VIX futures are rising, levered and inverse products buy VIX futures, creating a feedback loop. Levered funds buy VIX futures when they rise, because their NAVs rise twice as fast as the value of their underlying VIX futures in percentage terms, leaving the futures position insufficient for the now-larger ETPs. Similarly, inverse funds cover short VIX futures when they are rising.
A 3-point VIX futures spike in prices would drive ETP issuers to buy more than 50% of average volume of VIX futures. A 3 point jump would leave vol below-average, yet create demand for over $100mm vega (100k VIX futures), potentially driving vol up further. This number would have been around $50mm vega on a typical day in 2016. The amount of rebalancing has been boosted by low VIX futures levels, making an N-point move a larger percent of the current level of VIX futures. This effect has been largely untested in 2017 because of the lack of large SPX moves.
So there you go.
When the proverbial shit hits the fan, don't blame risk parity and the trend followers. Rather, point the finger at the "Target manager next door".
Nessun commento:
Posta un commento