The risks grow that the passive and algorithmic transformation of equity markets could lead to a crisis.
The following article was originally published in “What I Learned This Week” on August 15, 2019. To learn more about 13D’s investment research, please visit our website.
For years in WILTW, we have dissected the meteoric rise of passive investing and algorithmic trading. The influence of “rules-based” strategies has only grown more extreme this year. According to estimates released by J.P. Morgan in late June, passive strategies now control 60% of U.S. equity assets while quant funds control 20% — a staggering 80% combined. Passive titans Blackrock and Vanguard now oversee $12 trillion, up from less than $8 trillion just five years ago. And based on a recent report by Thomson Reuters, algorithmic trading systems are now responsible for 75% of global trading volume.
This development has never been tested by a recession. And evidence continues to mount suggesting the algorithmic and passive transformation of markets will only accelerate and deepen pain. The U.S. equity market has proven more susceptible to herd behavior. It has grown more at risk of falling into liquidity traps. And it has seen safe-havens quickly turn into overvalued danger zones.
These weaknesses have manifested over the past year in extremes: periods of prolonged calm followed by violent market swings. Now, as the global economic outlook grows more uncertain, all investors must consider a question posed by IMF director Tobias Adrian earlier this year: “Will we just have to get used to more spikes in volatility, or have we just seen the tip of the iceberg, and the big problems underneath are about to come out?”
It’s rare markets offer such a vivid depiction of the difference between humans and machines than we’ve seen in recent weeks. As the S&P 500 hit new highs late last month, the proportion of “mom-and-pop” bears exceeded bulls, according to a survey by the American Association of Individual Investors. Discretionary investor equity exposure remained low, never breaking the decade average. And in recent weeks, traders loaded up on call options on the Cboe Volatility Index, anticipating market turbulence following the FOMC announcement.
Meanwhile, systematic strategies flooded into equities:
The distilled reason for this divergence appears clear: The humans saw increased risk of human unpredictability — a trade war one Trump tweet away from escalation and a Fed facing contradictory evidence for a rate cut. The machines saw a number: historically low volatility.
When the humans proved prescient and volatility spiked last week to its highest since January, the flat-footed machines unloaded swiftly. According to Wells Fargo estimates, volatility-targeting strategies sold roughly $50 billion in U.S. equities from Monday through Wednesday.
If volatility remains elevated, the unwinding will continue. Just between volatility control funds and commodity trend advisors (CTAs), computer-driven traders could unload a total of $70 billion in equities, Deutsche Bank has warned. And by the end of the month, risk parity funds — which are slower moving — could unload an additional $20 billion in developed market shares, according to Nomura analysis.
While discretionary investors may have limited their pain during this sell-off by pairing equity holdings, herd behavior has nonetheless exposed them to risk. Anticipating a downturn, investors have flooded into low-volatility funds. The iShares Edge MSCI Min Vol USA ETF and the Invesco S&P 500 Low Volatility ETF have seen inflows of roughly $8 billion this year, only slightly trailing the $8.33 billion in inflows to the biggest U.S. ETF, the Vanguard S&P 500 ETF. As a result of this crowding: “Low-volatility stocks are trading at almost three standard deviations above the mean. That means low-vol is more expensive than it has been nearly 99% of the time, relative to the mean, since 1990,” according to the Leuthold Group.
J.P. Morgan’s quant guru Marko Kolanovic has identified this crowding as an opportunity. An extreme divergence has opened up between the defensive stocks held by low volatility funds and value stocks. Kolanovic believes this offers, “a once in a decade opportunity to position for convergence”:
Yet, if the “low vol bubble” indeed pops, it could have ripple effects throughout the market. Last month, The Wall Street Journal published a report on an unintended consequence of the algorithmic era: “the gamma trap”. It appears a significant contributor to the market’s penchant for long periods of calm followed by violent swings. We quote the WSJ:
“The force is a byproduct of big investors embracing so-called low-vol investment strategies that promise income and smooth out gains and losses through the use of options markets…Gamma comes from the Greek letter used in mathematical formulas of options prices, and it measures how much the price of an option accelerates when the price of the security it is based on changes…When gamma is positive, options quickly get more valuable when the price of the related shares rise. The bank taking the opposite position to the investor then sells those shares. That damps volatility.When gamma is negative, it is the other way around, and banks buy shares when prices are rising and sell when they are falling…Banks have ended up being long gamma more often because of the strong demand from insurers and pension funds for strategies that generate income while limiting exposure to stock-market falls.”
What happens if that insurer and pension fund demand craters because low-volatility strategies have grown overcrowded, therefore not the safe-havens investors expect? “The gamma trap” is yet another example of what we see as the greatest threat posed by the algorithmic and passive transformation of markets: the disconnect between the “daily liquidity” of ETFs and the far-less-than daily liquidity of the assets underlying those ETFs. (See WILTW August 8, 2019, for our analysis of why this threat could cripple the corporate bond market.)
Due to post-GFC regulations, big banks have dramatically pared back their market-making capabilities. Instead, they now act like brokers, employing high-frequency algorithmic techniques to match supply to demand. Along with authorized participants — the liquidity providers for ETFs (see WILTW December 6, 2018) — it’s unclear whether modern market makers will provide the liquidity necessary to stabilize markets in the event of a crisis. The abundant liquidity provided by QE has delayed a test of this system for almost a decade.
If volatility continues to escalate, that test may soon come. Market depth appears to be sitting near historic lows — as measured by the volume of orders on the bid and ask sides for the S&P 500 E-mini futures contract (see chart below). The market is already walking a liquidity tightrope. And as The Daily Dirtnap’s Jared Dillian has pointed out: “There is an axiom in markets: volatility and liquidity are inversely correlated.”
With economic uncertainty intensifying, we recommend extreme caution. Stay away from where the passive herd has crowded. Watch algorithmic sell signals closely — for one, the S&P 500 falling below its 200-day moving average could trigger another unwinding. Or, maybe best, follow the wisdom offered last week by one of our closest advisors — an investing legend: there’s “too much risk for me to play for the last yard.”
