If one asks traders what was the common market theme over the past three months, among the various (often angry) answers, one will stand out: general prevailing chaos, manifesting itself in soaring volatility which coupled with record low liquidity and a sharp pick up in trading volume, resulted in disjointed, erratic price action that crushed all momentum and trend-following strategies (which these days is most of them).
Commenting on this phenomenon among others, yesterday JPMorgan quant Marko Kolanovic said, correctly, that "liquidity has become to a large extent driven by market volatility" reinforcing a negative feedback loop between volatility and liquidity, and as the most recent examples he cited the unprecedented drop in futures market depth or "record low liquidity" we discussed previously (shown in the chart below), the currency flash crash on Jan 2, or the equity market “upside crash” on December 26.
Yet one thing that Kolanovic ignored was the the third aspect of what has recently become the market's unholy trinity: trading volume.
Addressing this under-discussed facet of market function, overnight Goldman's derivatives expert Rocky Fishman pointed out that after nearly a decade of declining market volume, 2018 saw an abrupt reversal in this trend, with high volume growth in SPX E-mini futures (+21%), most S&P 500 ETFs (+40%), and SPX index options (+27%).
On the surface, this appears paradoxical: after all, falling realized liquidity would seem to contradict the observed strong growth in trading volumes of US index products. However, as Goldman points out, when volatility’s impact is considered, delta-one SPX product growth - or volume adjusted for vol impact - was actually low in 2018.
Indeed, when discussing market volumes one must always consider that they tend to be highly correlated with realized volatility, or as Goldman puts it, "there is a high positive correlation between realized volatility and volumes of SPX futures and the three major S&P 500 ETFs (SPY, VOO, IVV)."
The relationship shown in the graph below implies that E-mini SPX futures should trade $200bln notional/day when realized vol is at 10%, and around $260 bln/day when realized vol is at 20%. This relationship is particularly strong for the ETFs (0.88 correlation between monthly unit volume and realized volatility over the past 5 years, vs. 0.80 for E-mini futures).
From this it follows that the spike in volatility in the last quarter of 2018 contributed to high-volume months... however as shown above, delta-one volume was lower than volatility would have indicated. As Fishman notes, the concentration of high-volatility months (five of 2012-8’s ten most volatile months were in 2018) led to a concentration of high-volume months.
However, the last five years’ volume vs. volatility relationship for both futures and ETFs makes 2018’s delta-one volumes appear low for this level of volatility. Furthermore, July and August came with the lowest future and S&P ETF trading volumes in years, even though realized volatility was marginally higher than it was throughout 2017
Another consideration is that the sharp volume gains in 2018 were largely offsetting losses from the record low-volatility observed in 2017. According to Goldman, low volatility in 2017 contributed to reduced future and S&P ETF volumes in 2017, while 2018’s gains largely offset 2017’s losses.
Finally, showing the explicit co-relationships between volume, liquidity and volatility is the following chart which shows that the top-of-book depth, which as we showed previously is record low, tracks the ratio between volume and realized volatility. In other words, "volumes that are low for a given level of volatility may be limited by poor liquidity - or may be leading to poor liquidity" according to Fishman. So whichever way the arrow of causation works (it’s some of each), 2018 came with both poor liquidity metrics and delta-one volume numbers that were lower than would be expected for this level of volatility.
Why is this a concern?
Because in a time when the entire asset class of active investing is increasingly endangered by the encroaching threat of passive investing resulting in prominent hedge funds shuttering left and right and removing marginal market players, and where HFTs are now cannibalizing each other and unable to "provide liquidity", i.e. offset any sharp jump in trading volume, as a result of unprecedented commoditization of order flow, trading volumes - in both absolute and delta-one terms - will keep declining. This is how Kolanovic explained the recent trends observed in the market:
The depletion of market reversion forces was driven by a decline of value investors (as money moved to passive and systematic strategies), a shift of assets from public to private equity (private equity has a more favorable mark to market treatment, thus creating arbitrage between public and private equity), and a reduction of human risk taking activity after the 2008 crisis (e.g., block traders, prop desks, etc.).
Said otherwise, with volatility expected to keep rising as central bank liquidity withdrawal accelerates in the coming months, something BofA discussed two days ago when it noted that "a flattening in the yield curve over the last three cycles has preceded rising volatility by about three years" and thus "expects a more volatile backdrop for US stocks in the coming years", going so far as to suggest that volatility will "likely double by 2021"...
... absent an increase in overall market volume, liquidity will collapse to even more unprecedented depths as the volume/volatility ratio hits new all time lows.
Which takes us back to something Kolanovic said yesterday, namely that "Equity markets could benefit from a rethinking of the current state of liquidity provision and of market reversion forces." In this particular case, the JPM quant is absolutely correct, because if the current trend of rising volatility coupled with declining delta-one volumes continues, the result will be a market in which the top-of-book depth eventually collapses to zero and where even the smallest order has the potential to unleash chaos.
Which is good news for all those day traders who hope that the current illiquid, disjointed market state persists... and bad news for those investors who hope for a reversion to prior trends. Because absent another major central bank intervention that eliminates the endogenously rising market volatility, the real chaos is yet to come.