Race for the Exits
Vol #85 - June 17th, 2022
Liquidity is the ability to get into, and more importantly out of, a trade when you need it.
Simple metrics like volume, bid-ask spread, or posted size are a good indicator of instantaneous liquidity. High volumes, tight spreads, and large size demonstrate the ability to execute large trades with relatively little slippage.
Liquidity is deepest when participants are most confident in their pricing. When model inputs and macro backdrop align to give traders confidence, they are willing to deploy large amounts of capital for very thin margins.
As the hydro etymology suggests, liquidity is very fluid, and can drain in an instant. Like much about the markets, this process is more geometric than linear.
The compounding effects of a liquidity event mean that spreads reach multiples of their “normal” level and prices move exponentially more than usual. Swirl this with human psychology and the events become even more dramatic.
More than 90% of stocks in the S&P 500 dropped yesterday. That’s happened 5 times in the last 7 days. There is no historical precedent for this since 1928 - that’s how furious the selling pressure is.
The crowded movie theater analogy is often used here to describe the perils of liquidity, and what happens when it gets stressed. In normal circumstances, patrons file into the theater through the doors and find their seats one by one. Even though they’re all arriving for the same show, there’s dispersion in the exact moment they pass through the door. Whether it’s temperament about being on time or delays at the popcorn stand, there is a randomness to their arrival times.
Taking a brief mathematical detour, this dynamic is typically modeled by the Poisson Process, named for the French mathematician Siméon Denis Poisson. A leading academic of the early 19th century (his name is one of 72 inscribed on the Eiffel Tower), he investigated statistical distributions, thermodynamics, and fluid mechanics. Given this background, it’s surprising he never found his way to the Bourse.
A Poisson process is one in which the average time between various events is known, but they occur randomly over time, and are independent of each other. Call centers are some of the most common users of this modeling, but it extends to many other events, such as the frequency at which customers open brokerage accounts.
During normal liquidity scenarios, the randomness of buying and selling and timing of customer orders means we can think of orderflow as being relatively random. Jack might be selling stocks to pay for his retirement, while Jill is buying stocks because she just got her biweekly paycheck. On a day to day basis most of the movement in the market can be assigned to this type of orderflow distribution. Marginal moves in the market are coyly described as “more buyers than sellers.”
There are broad underlying trends like earnings growth or quantitative easing that might provide an underlying trend to the market, but in high liquidity and low volatility environments, participants more or less offset each other.
But when the amplitude of sell flow dwarfs the buy flow, you see much more choppy price action and lower liquidity. The imbalance of flows has created much less price certainty, and pushes the market dramatically. This happens in both directions, and is often the cause for bear market rallies.
A key point in using Poisson distributions is the independence of various trials. Inbound calls at the Air France desk might be random, until there’s a weather pattern over Europe that results in hundreds of canceled flights. The movie theater has an orderly and random filing in, but if there was a fire, it would be a flood for the exits. Headline news like war and inflation have the same effect on the market.
Market practitioners talk frequently about “exit liquidity” in this context. Sure the trade seems good now, but how wide are those exit doors if a spark alights? (This was a major component in the death spiral of the stablecoin UST. Depositor balances relative to the liquidity exit points were too large when events got stressed.)
Orderflow in the market right now is far from random. The market wide contagion has eliminated the relative independence of both participants and asset classes. Investors everywhere are licking their wounds, and some of the largest participants sorely remember the lessons of the 1970s as they try to get out while they can.
At the heart of the market the most active participants are measuring their activity in milli, micro, and picoseconds. Human time is much slower than that, and the extended sustained sell off reflects that. The slow cycle of reading the news, discussing with your peers, reaching out to your broker or advisor, all takes time that is measured in days and weeks. As price action reflects this, the cycle compounds on itself.
Liquidity can ultimately be both a blessing and a curse for public equity markets. While it’s fantastic that customers can access markets anywhere from their cell phones, it also makes them prone to making shorter term decisions. Confetti from Robinhood helps titillate these emotions, but broadly the accessibility of the sell button opens up an escape valve that we may be too quick to reach for.
Less liquid markets force a certain discipline on investors. It’s much harder to panic sell commercial real estate than it is Microsoft stock. The accessibility of equity and options markets makes them attractive, but prone to the fickle swole of popular delusions and manias.
What the right value of the S&P 500 is today, is a question without an answer. The purist will tell you it’s the price on the screens right now. Analysts will look at future cash flow or systematic factors to say whether we’re under or overvalued. Optimists will tell you to close your eyes and buy the dip.
Value is closely tied to liquidity. The more liquidity there is, the more certain we can be about value. When liquidity is being drained, it’s time to become less confident about prices in the short term.