In early 2021, a subreddit community took revenge upon what it saw as greedy and ruthless capitalists who were shorting a retail game store; which carried fond memories for many of the chatroom users. For much of 2020, the price per share of GameStop was under $5, but as this group gathered together they pushed the price up to $325 in January 2021 – a staggering 6,400% increase.
While the motivation was clear, what is less clear is why a market, which many regard as being reasonably efficient, strayed so far away from fundamentals, writes Morningstar’s Mark Murray.
There is a vast literature of psychological evidence highlighting human irrationality, which is one of the cornerstones of behavioural economics. However, while this may be a necessary condition for the existence of market mispricing, it is by no means sufficient. Efficient market theorists can easily allow that humans make irrational decisions; all they require is that the resulting mispricing is brought back to fair values by rational arbitrageurs.
But the main problem with this argument is that real world arbitrage is risky and costly and, therefore, limited in bringing prices back to fair values. Hence mispricing is persistent in the market. Economist Andrei Shleifer, author of Inefficient Markets, has identified numerous reasons why this is the case.
The first reason is that in order for arbitrage to be effective there needs to be either perfect substitutes to the mispriced asset, which is a security with identical cash flows in all states of the world, or very close substitutes that have cash flows which are very similar to the mispriced asset in all states of the world.
In the real world this is rarely the case and, therefore, arbitrage is risky and costly. For example, if an investor thought that the broad US stock market was overvalued in 2021, they could not arbitrage this as a close substitute to create a short position which did not exist – namely, an asset with identical cash flows in all states of the world which was not overvalued.
Another reason arbitrage is limited is that even if a close substitute does exist, the arbitrageur still bears risk. That is the risk that the fundamentals of the company shorted improve and the company of the long position deteriorates. If an investor thinks that shares in Exxon are expensive relative to BP, he/she can long BP and short Exxon and thereby eliminate general risk to the energy sector and benefit when prices converge to fair values. However, he/she still bears the risk that Exxon does surprisingly well and BP surprisingly poor thereby making the arbitrage transaction risky.
Even if perfect substitutes exist, the arbitrageur still faces the risk that the mispricings get worse before they converge to fair values. This can lead to short-term losses for the arbitrageur who may face institutional or client pressures, especially if the mispricings last for long periods of time. This may be made worse if he/she faces margin calls which mean he/she may have to close the position at a loss before the prices return to fair values.
These theoretical assumptions are increasingly being supported by a growing empirical literature on the limits to arbitrage. Many studies have looked at variables as proxies for arbitrage risk, including the amount of institutional ownership and idiosyncratic volatility, and have found higher returns once sorted based on these. The paper Arbitrage Asymmetry and the Idiosyncratic Volatility Puzzle, for example, combines 11 common mispricings identified in academic literature and combines them into one mispricing measure. The authors find, in their sample period 1965-2011, the most overpriced stocks have a monthly excess return of -0.81%, but when only stocks which are hard to arbitrage are selected this increases to -1.89%.
We see a similar effect on the most under-priced stocks, with a monthly excess return of 0.28%, which increases to 0.56% for the hardest to arbitrage stocks. The impact, however, on the under-priced stocks relative to overpriced stocks is less pronounced due to arbitrage asymmetry. That is, arbitrage should eliminate more under-pricings than overpricings as a result of more investors capital in the former than latter.
With this in mind, was there a risk-free way to arbitrage the obvious mispricing generated by this determined chatroom?
This article was written for Portfolio Adviser by Mark Murray, senior investment analyst at Morningstar Investment Management.
