Liquidity — or, perhaps more to the point, a lack thereof — matters.
It matters in drawdowns and it matters in rallies. Failing to appreciate this leaves one bereft when it comes to explaining the ferocity of selloffs and the rapidity of rebounds after routs.
There are a number of ways to measure liquidity, but one thing we can say with some degree of certainty is that market depth never really recovered after the February 2018 correction which was, of course, defined by an “extinction event” in the VIX ETP complex.
“Naturally” (if that’s the right word), liquidity deteriorates in selloffs. But that simple, intuitive observation has morphed into something more pernicious in modern markets.
The relationship between volatility and liquidity “is very strong and nonlinear”, as JPMorgan’s Marko Kolanovic is fond of reminding the world. “Market depth declines exponentially with the VIX”, he reiterated in a 2019 note. “Given that an increase in volatility often results in systematic selling, this relationship is the key to understanding market fragility and tail events”.
This really is the crux of the matter when it comes to conceptualizing the “doom loop” dynamic behind most of the acute selloffs we’ve seen over the past several years.
“The supply of liquidity disappears exactly when a higher volume of futures is trying to make it through ever narrower pipes, thereby exacerbating spot moves, amplifying trends and increasing volatility all at the same time”, SocGen wrote late last month, in a volatility outlook piece for the second half. “With the VIX still above 30, order book depth on S&P 500 futures was still one-fourth of what it was in January, and given the empirical dependence on the level of the VIX, is unlikely to reach those levels unless the VIX moves below 20”. Note the reference to “empirical dependence”.
This same dynamic comes calling over and over again, and while more and more market participants are beginning to understand how it works on the way down, it’s also important to remember that a lack of market depth can also amplify upside moves.
Examples of lackluster liquidity colliding with bullish flow catalysts to create dramatic moves higher include the impact of rebalancing flows during the last week of December 2018 and the slow-motion melt-up that occurred during the first half of last year.
So, just how important was this during the recent crisis?
Answering that question could take us pretty far down the rabbit hole. Rather than go there again (regular readers have traveled down that rabbit hole with me on any number of occasions over the past four months), I thought it might be useful to highlight a few simple observations from Goldman.
In a new note, the bank spends some time modeling S&P returns based on flows, and part of that effort involves quantifying the impact of liquidity, which the bank notes “plays a role in understanding how investment flows impact price moves”.
“Flows that occur when the market is liquid have a smaller impact on price moves than flows that occur when the market is illiquid”, the bank’s John Marshall, Zach Pandl, and Rocky Fishman write, in a note dated July 2.
Referencing a flows-based model of returns, the bank says the model fit “was improved significantly by transforming each of the flow variables” using a single stock liquidity index.
Suffice to say that, as Marshall, Pandl, and Fishman go on to write, “scaling each flow variable by size tradable for a given market impact has been particularly important in recent years because of the significant and rapid decline in liquidity in Feb-2018, Dec-2018, and Mar-2020”.
Goldman looks at the recent selloff and compares the model fit both without a liquidity adjustment, and with one. Here is what the bank found (and do make sure you read the latter passage carefully):
Liquidity adjustments are important for explaining the asymmetric effects of flows on SPX returns in drawdowns vs rebounds. Using the recent drawdown as an example, the model that does not account for liquidity effects would have suggested an SPX decline of 23% from 19-Feb to 18-Mar, while the model including liquidity effects suggested a 33% decline (SPX actual decline was 29%).
The effect was even larger during the flow rebound; from 18-Mar to 19-May without liquidity effects would have suggested a rebound of 24%, while adjusting for liquidity effects suggested a 53% rebound. This shows how incorporating liquidity effects not only improves the fit through time, but also explains how inflows in an illiquid market (following a selloff) can drive a bigger increase than the drop caused by the same outflows that occurred when liquidity was stronger.
Again, that latter bit is crucial, and may well have quite a bit of explanatory power when it comes to the rebound off the March lows.
At this point, stating that many market participants have been “stunned” by the scope and rapidity of the bounce is something of a cliché. All manner of explanations have been offered to explain it, and I would suggest considering the above when you write your own history of what ended up being the most spectacular bear market rally of all time.
During selloffs, surging volatility diminishes market depth. When risk assets recover, subsequent buy flows/short-covering thus hit in an illiquid market. The impact of those flows is thereby magnified commensurate with the diminished state of liquidity.
The result: A turbocharged rally.