The incredible cost of short selling
Acadian Asset Management
Systematic investing built on 35+ years of data-driven insights. Equity, alts, credit, sustainable strategies.
We live in extraordinary times. As discussed in Stupidity is our destiny: Historic closed-end fund overpricing, last month the U.S. equity market had a closed-end fund with a premium over 2,000%. Here’s another doleful milestone: the cost of shorting stock is now at historic highs. As of 2024, we routinely observe annualized shorting costs above 100% for hundreds of individual U.S. equities, and we occasionally observe stocks with costs above 1,000%. We haven’t seen anything like this since 1931. Turning from the extremes to the whole distribution, Daniel et al (2024) look at data from 2001 to 2023 and show that both the mean and median cost of shorting have steadily increased, especially for smaller-cap names.
Now, it remains true that most large-cap U.S. stocks are cheap to short, and on a value-weight basis the cost of shorting individual names remains low. Thus, the fact that we have many expensive smaller stocks says little about the valuation of the aggregate market; it only tells us that there exists a dark underbelly of grotesquely overpriced smaller stocks.
What does it mean to have a shorting cost of 1,000%? Suppose Alice wants to short stock ABC. She needs to borrow ABC from existing owner Bob. Bob will charge Alice a fee, which is paid daily by Alice to Bob and expressed as an annual rate. If stock ABC has price of $100 and Bob demands a fee of one dollar per day, Alice’s daily cost is 1% of the price, which annualizes to more than 1,000%. In order for Alice to profit from shorting ABC, the price needs to fall by more than $1 per day.
What are the implications of high costs? In my view, high shorting costs are both a symptom and a cause of overpricing. If ABC costs 1,000% to short, then (a) that deters many would-be Alices from shorting and (b) that means there is at least one brave Alice out there who is willing to bet that ABC will fall more than $1 per day. Thus today’s high costs tell me that there are some horrifically overpriced stocks out there, and our dysfunctional stock loan market is an impediment that prevents rational traders from correcting this overpricing.
Let’s take a specific example: Bed, Bath, and Beyond (BBBY), which delisted from Nasdaq in October 2023. BBBY was part of the meme stock madness of January 2021, when its price peaked around $35. On its journey from $35 to zero, BBBY was extremely expensive to short. At one point in February 2023, for example, when BBBY was around $5, the cost of shorting was 443%. If you correctly believed in February 2023 that BBBY would soon be worthless, it would have been very expensive to act on this belief. My contention is that today the market contains many more BBBYs; stocks that should have a price near zero, but instead have a high price due to high shorting costs.
How widespread are these ginormous shorting costs? On a typical day in May 2024, we observe more than two hundred U.S. common stocks with shorting costs over 100%, of which perhaps five have market cap over $1B and another 30 have market cap above $100M (so the majority have cap under $100M). Taken as a whole, this group has the sketchy characteristics you’d expect: controversial stocks, meme stocks, penny stocks, stocks teetering on bankruptcy, and stocks with very low float. These are the “frauds, fads, and failures” that short sellers believe are overpriced. We also see extreme shorting costs on some ADRs, ETFs, and closed-end funds, including the previously discussed closed-end fund with a huge premium.
How surprising are these numbers? Very surprising. D'Avolio (2002) discovered that a handful of stocks in the period 2000-2001 had double-digit costs and commented, “The fees that short sellers pay for these stocks are startling …e.g., 55% for Krispy Kreme…”[1] Well, if double-digit costs are “startling,” what are triple-digit costs? Flabbergasting? How about four-digit costs? Mind-blowing? My mind is certainly blown by today’s extreme costs.
On an equal-weight basis there’s been quite a shift in the distribution of borrowing costs. If you define a “special” stock as a stock that costs more than 1% to borrow, then the fraction of the market that is special has risen from less than 10% of names in 2001 to about 50% in 2023, according to Daniel et al (2024). The term “special” comes from a previous era when most stocks were cheap to short; as of today, it is no longer very special to be “special” in U.S. equity markets.
Let’s use the term “shortflation” to describe the rise in costs over the past twenty years. Why is shortflation happening? I’ll discuss various explanations, but one thing is clear: however we got here, today there are some stocks out there that are incredibly expensive to short.
In this post, I want to explore the causes and consequences of high shorting costs. I’ll take you on a harrowing tour into the heart of darkness: the U.S. stock loan system. We’ll learn more about Alice and Bob and their tortured relationship. I’ll touch lightly on many topics; for more complete coverage of the burgeoning academic literature, I recommend you read the excellent discussion in Daniel et al (2024).
Disagreement, short-sale constraints, and overpricing
The cost of borrowing stock is only one of the many constraints that short sellers face. These constraints include both costs and risks. In this section, I want to briefly sketch an important idea in economics: disagreement + short-sale constraints = overpricing. This overpricing could be for individual stocks or for the entire stock market.
Why does Alice want to short ABC, while Bob wants to be long? They must disagree about something. Disagreement is at the heart of any stock market; markets are places where bulls and bears trade with each other to express their different views.
As discussed in Lamont (2004), our system is just not set up to encourage traders to go short, and short sellers are a hated minority that is frequently abused by just about everyone. For example, the management of ABC may harass Alice or encourage Bob not to lend to her. Examples of this type are discussed in Lamont (2012), and in April 2024 we saw another example in the news of a firm battling with short sellers.
A recent development is mobs of retail investors who coordinate short squeezes on social media, as discussed by Allen, Haas, Nowak, Pirovano, Tengulov (2021). If Alice shorts stock ABC, she faces the risk that its price will rise due to a meme that triggers a wave of buying.
Let’s consider an extreme form of short-sale constraints: suppose shorting is illegal (as currently in Korea). In that case, Alice is unable to express her views by shorting ABC, and ABC can become overpriced. If everyone agreed with Alice, it wouldn’t matter that she can’t short. The problem is when negative information or opinion is prevented from getting into market prices. Thus, as emphasized by Miller (1977), we need both disagreement and short-sale constraints in order to get overpricing. For more on this topic, see Lamont (2004), Reed (2013), and especially Daniel et al (2024).
Suppose short selling is illegal and that stock ABC is overpriced. Alice can only fume and refuse to buy it, but Bob decides to hold it. Should we conclude that Bob is an idiot? As an empirical matter: yes, in my opinion Bob is most likely an idiot, or maybe Bob is busy at his dental practice and doesn’t realize that ABC is overpriced, or maybe Bob has goals other than maximizing his wealth through trading.
As a theoretical matter, though, it is possible that Bob is totally rational, he is aware that ABC is overpriced, but he decides to hold it anyway. Why? Because Bob plans to sell ABC to someone else tomorrow at a profit. Amazingly, it is possible to construct a version of this story in which all market participants are rational, everyone is aware that ABC is overpriced, but everyone is willing to hold it anyway. This result, formalized by Harrison and Kreps (1978), is a version of the “greater fool theory.”
What I am calling “overpricing” could also be called a “rational speculative premium.” Lamont (2004) gives a simple numerical example. More recent research has de-emphasized the “rational” explanation and turned to irrational motives for speculative trading, including overconfidence. See Daniel et al (2024) for a discussion of this literature. The idea is that Bob disagrees with Alice because he wrongly believes he is smarter than she is; Bob, after all, was in the top ten percent of the bottom half of his class in dental school.
What do overconfident traders do? They trade a lot, and they believe they have the ability to outwit other market participants. Put overconfident traders into a market where shorting is difficult, and you have the necessary ingredients for a speculative bubble. As discussed in No we are not in a bubble yet, you can’t have a bubble without trading volume, and as discussed in The unholy trinity of bubbles: valuation, volatility, and volume, this trading volume perhaps reflects a “greater-fool” mentality held by overconfident investors.
As discussed in Lamont (2004), short-sale constraints were an important part of the tech stock bubble of 1999-2000. The main constraint was not the expense of borrowing stock, but broader constraints that prevented pessimists from expressing their views. For example, in 1999, you could easily short Nasdaq futures, buy puts or otherwise bet against the broad stock market, but few did (see for example Lamont and Stein (2004)). However, in some specific episodes including the Palm/3Com case in Lamont and Thaler (2003), shorting costs played a huge role.
If short selling is constrained, how does overpricing get corrected? One answer is issuance. If ABC is overpriced, one way to increase supply is for Alice to short sell ABC, thus creating more shares for optimists to buy. If this channel is not available, then company ABC itself can increase supply by issuing shares. In this sense, short selling and issuance are substitutes. We need smart money to correct mispricing, and the smart money can be either Alice or ABC. In the tech stock bubble of 1999-2000, we saw a huge wave of issuance (and a decline in short selling), and thus I say that issuance is the Third Horseman of the bubble apocalypse in No, we are not in a bubble yet.
Let me summarize. First, on an aggregate level, short-sale constraints allow the whole stock market to be overpriced and allow speculative bubbles to occur. These constraints are not costs of borrowing stock; they are broader constraints that deter betting against the whole market. Second, at the individual stock level, stock-specific constraints, including shorting costs, allow overpricing; hence my concern with shortflation.
Dark markets
Before plunging into the murky waters of the stock loan market, I want to talk about a market transaction we can all understand: buying a latte.
We live in a golden age of coffee. If I want to buy a latte, I can walk into any one of the 54 Starbucks in Boston. I am not required to haggle over the price, Starbucks invariably has lattes available, and it’s easy to find a location near me.
Now consider the following dystopia. The government outlaws all advertising and exterior signage on buildings. It further mandates that every retail establishment must move to a new random location every week. Now, to obtain a latte I must go from door to door, entering each building and inquiring whether they serve lattes. Once I find a vendor, I may discover that they’ve run out of milk, or that the only available latte is decaf skim pumpkin spice. The baristas become surly, and I must haggle with them over the price; sometimes (like the Soup Nazi from Seinfeld) they refuse to serve me altogether. One Starbucks may sell a latte for $5, while the one across the street charges $50.
If I get tired of traipsing around Boston in search of a latte, I must hire intermediaries. A network of “prime latte brokers” will arise to buy and sell lattes, and each Starbucks employs a “latte agent” to negotiate with these brokers as well as a “latte custodian” to hold the latte so that it does not spill during the negotiations. In return for their services, each of these middlemen gets to take a sip of my latte, so that by the time the cup gets to me it is half empty (or, as my prime latte broker will insist, half full). While I am drinking my latte, sometimes my broker will suddenly grab the cup out of my hands and run away, because the latte has been “recalled” by the latte agent.
I’ve just described the stock loan market. A normal person would describe this market as “messed-up,” “dysfunctional,” or “just plumb loco;” economists would call this a “search market with frictions,” an “over-the-counter” (OTC) market, or a “dark market.” I like “dark market” because of the ominous connotation: in my view, dark markets are bad places where bad things happen.
Dark markets vs. centralized markets
Here’s Duffie (2012) describing dark markets:
Rather than trading through a centralized mechanism, participants in an OTC market negotiate terms privately with other market participants, often pairwise. OTC investors…may be unaware of prices that are currently available elsewhere in the market…In this sense, OTC markets are relatively opaque; investors are somewhat in the dark about the most attractive available terms and who might offer them.
The polar opposite of a dark market is a centralized market. An example is the market for buying U.S. equities: it is an “all-to-all” market (all can sell and all can buy) with prices determined publicly instead of by private bilateral negotiation. Thus during trading hours we can say the price of ABC is $100; we don’t have to say “ABC has a price of $10 in Boston and a price of $100 in Cleveland” or “ABC has a price of $10 from Goldman Sachs and a price of $100 from JP Morgan.”
Let’s talk about search. Search models have been studied by economists for decades, and we think search is important both for understanding individual markets (what determines the price of lattes?) and the entire economy (why do we experience recessions with high unemployment?). Search frictions determine where we are in the spectrum between pure competition (where each seller just passively accepts the market price) vs. monopoly (where sellers have market power and deliberately set prices high).
Let’s go back to the coffee dystopia where I have to wander, uncaffeinated, around Boston. Suppose after 30 minutes of searching, I find a Starbucks. But this particular location demands $20 for a latte. I can either accept their offer, or trudge away in hopes of finding another location with a better deal. The higher the search cost, the more likely I am to pay the high price demanded. In this sense, higher search costs create monopoly power, while lower search costs push us towards the competitive outcome.
So, what is the actual situation with buying a latte? Unlike the U.S. equity market, the market for lattes in Boston is not a centralized market with all-to-all trading. Starbucks stands ready to sell me a latte, but stubbornly refuses to buy a latte from me. So the latte market is an OTC search market characterized by sequential bilateral one-way trades. Still, unlike the hypothetical dystopia, the actual market for lattes in Boston is pretty efficient: we have minimal price dispersion, easy ability to transact, and no middlemen required. When there are few frictions and the cost of search is low, dark markets can closely approximate the outcome of competitive, centralized markets.
Why are some markets dark?
Would it be better to have a centralized New York Latte Exchange regulated by the Starbucks Exchange Commission? No. It’s just not practical to have centralized markets in everything. As every prime broker or their lobbyist will gladly explain, the world needs customized contracts and trades that can only be supplied by OTC markets. Setting up a centralized market entails costs, and most markets are not big enough to justify these costs.
Most of our major life decisions, both economic and non-economic, involve dark markets. If we want to find an employer to employ us, or a romantic partner to romance us, we don’t use centralized markets. Instead, we use bilateral search. However, it’s probably true that in recent years technology has lowered search costs and has moved many of these activities closer to the competitive equilibrium. Dating apps such as Tinder or Bumble may not (yet) have a central limit order book, post-trade analytics, or a consolidated audit trail, but new technology has surely pushed us closer towards the competitive centralized outcome.
As discussed by Weil (2020), U.S. financial markets are mostly dominated by dark markets and not centralized markets. Markets sometimes switch from centralized to dark. For example, in the 1920s and 1930s, corporate bonds in the US mostly traded on a central exchange; now they mostly trade OTC. Similarly, in the 1920s and 1930s, there was a centralized market for stock loans on the NYSE, as described by Jones and Lamont (2002).
It’s not always obvious what drives different market structures, or why various attempts over the years to create centralized markets sometimes fail and sometimes succeed. One theory involves politics. Daniel et al (2024) mention a failed recent attempt to set up centralized stock loan in the U.S. Let’s go back to the coffee dystopia. Who benefits from the absurd legal system? It’s the middle men. You can rely on the latte lending agents, latte custodians, and latte prime brokers to lobby strenuously against reform.
The stock loan market
In any market, whether dark or centralized, supply and demand play the central role in determining prices and quantities. In the stock loan market, the price that equilibrates supply and demand is the fee that Bob charges Alice. The arrangement between Bob and Alice is (in theory) renegotiated every day, so the fee fluctuates daily and either Bob or Alice may withdraw anytime.
First, consider supply. Suppose stock ABC is widely held by institutions. In this case, the supply of lendable shares will be high, because institutions typically have stock lending programs. In contrast, if ABC is mostly held by retail investors, the supply of lendable shares will be low.
Now consider demand. If ABC is controversial, or perceived to be overpriced, short sellers will want to borrow it, so demand will be high. If ABC is boring and fairly priced, no one will want to borrow it, so demand will be low.
Thus in the stock loan market, it will be expensive to short overpriced, controversial stocks held by retail investors. One way to describe the stock loan market is to say it works great when you don’t want to use it (most stocks) but works terribly when you do want to use it (overpriced stocks).
For large-cap stocks, there are tons of unused supply and the fee is small. Looking at aggregate U.S. equities in 2024, ballpark figures, the whole market has capitalization of $50T, of which about $16T is lendable supply (in other words, about 32% of the market is held by institutions with lending programs).[2] Of this $16T, only 2.7% is actually lent out, and the value-weight average fee is about 80bp. So the big picture is that there is plentiful supply and low shorting cost on a dollar-weighted basis.
Bizarre outcomes
The good news is that for large-cap stocks, it is cheap and easy to short. The bad news is that when we look at smaller stocks, we observe high costs and inexplicable outcomes.
Let me start by elaborating on the lending process. So far we’ve discussed Alice (the borrower) and Bob (the beneficial owner who is willing to lend). Alice and Bob, like two ships passing in the night, will never actually meet or know of each other’s existence. They will be forever divided by the additional characters now crowding on to the stage:
So when Alice wants to short ABC, she contacts her brokers Eleanor and Fritz, one of whom may contact Denzel, who may supply the shares held by Bob.
Here are some scenes we see with these actors on stage:
Recommended by LinkedIn
Why are these weird things happening? Short explanation: dark markets. Anything can happen.
There is nothing necessarily sinister or surprising about price dispersion. Just as we see different passengers on the same airplane paying dramatically different prices for their seats, we see different short sellers paying different costs for the same stock. It is also true that securities lending is just one part of Alice’s relationship with her prime broker. If Alice generates a lot of revenue for Fritz, Fritz will allocate the cheapest loans to Alice.
Why is it so expensive to borrow stock? I am not sure. What we have here is a mystery with a list of different suspects who all have means, motive, and opportunity. Perhaps Bob is acting monopolistically. Or perhaps his agent Denzel is acting monopolistically on his behalf. But many of the outcomes I’ve listed seem inconsistent with monopolist lenders, and more just head-scratching. Maybe Fritz is bribing Denzel to get access to scarce shares and Bob is actually the victim. Perhaps Denzel and Fritz and Eleanor are all colluding. Who knows?
The stock loan market has historically been a place where prices failed to equilibrate supply and demand. When prices for some reason do not adjust, other forces come into play. In 1989, Business Week reported that “Authorities and brokerage executives alike suspect that some stock-loan aides accept big-money kickbacks, sexual favors, and drugs in exchange for routing business,”[3] while in 2007 the SEC charged 38 individuals with bribery/kickbacks in the stock loan market.
Shortflation
There is no doubt that today’s stock borrow costs are the highest in U.S. recorded history. Let me just briefly review some of the previous high costs recorded in the literature.
First, Jones and Lamont (2002) looked at the period 1926-1933. We found the 2% of the stocks were special, with the highest cost being 783.5% for International Shoe in September 1931 (that’s a premium of 782% plus the interest rate of 1.5%).
More recent results:
Daniel et al (2024) find that as of 2023, more than 15% of the universe has a fee higher than 10%. For stocks with market cap under $100M, they find fees had risen to an average of more than 30% as of 2023.
Are these fees economically significant, or are they just meaningless quotes that do not represent actual transactions? I have two answers to this question. First, if Alice wants to short ABC, but she is deterred by a quoted 1,000% fee, that is a meaningful event even if she decides not to short. Indeed, one explanation for the observed decline in short interest in smaller-cap names is the rising cost of shorting.
Second, we do observe large dollar aggregate amounts. Consider again the example of BBBY (Bed, Bath, and Beyond). BBBY is a complicated case, involving an attempt to issue shares (issuance is a substitute for short selling) as explained by the indispensable Matt Levine.[4] But briefly, at the beginning of February 2023, the company had a market cap of around $400M. During that month, BBBY generated $16M in lending revenue (payments from Alice to Bob) according to S&P Global.[5] Taking the holders of BBBY as a whole, that’s an annualized “dividend yield” of 48%. That’s a big number.
Supply and demand
Shortflation defies traditional economic analysis, since we observe rising costs during a period in which supply went up (more lenders of stock) and demand went down (fewer short sellers).
Looking first at supply, over the past 20 years, did lenders withdraw their shares from the market? I don’t think so. The number of lenders fell a bit around the GFC but has since recovered. If anything, the quantity of lenders has gone up.[6]
Next, demand. Have shorting costs gone up because there is a plethora of hedge funds seeking to go short? No. Short sellers are a vanishing breed, and the quantity of short selling is far lower than it was prior to the GFC. As can be seen in a recent SEC report,[7] short interest as a percent of shares outstanding peaked around 2008, especially for smaller-cap names. Short interest declined precipitously following the GFC, along with many other types of arbitrage activity. Short interest took another hit from COVID and the meme stock madness of January 2021.
So, high supply, low demand, put those two things together, and you should see low utilization. That’s exactly what we observe. D'Avolio (2002) reports utilization of 7% in 2001, while today U.S. equity loan utilization is a modest 2.7%.[8] In other words, if Bob has $100 worth of shares he wants to lend, he is able to lend only $2.5 worth on average.
Explaining shortflation
So supply goes up, demand goes down, plenty of unused inventory, yet cost is up? How can we explain this? Let me offer some theories.
First, perhaps the composition of the market has changed. Maybe the types of stocks that now exist are different. For example, the portion of biotech stocks has risen, and maybe biotech stocks are more likely to be subject to disagreement.
Second, perhaps it is the investors, and not the firms, that have changed. Perhaps retail investors are producing crazy mispricing that short sellers are striving to correct, in the process driving up shorting costs for some names.
Third, perhaps various regulatory crackdowns in the securities-lending market have made short selling more costly. Blocher and Ringgenberg (2018) say that due to changes in options regulations, “market quality has deteriorated: price efficiency is lower and stocks are more overpriced.”
Fourth, perhaps something in the stock-loan market has changed, so that lenders are behaving less competitively. This theory says shortflation is due to greedflation.
Fifth, maybe we are moving away from a backward system with fixed prices towards a more open system where prices actually equilibrate supply and demand. Perhaps twenty years ago, there existed many overpriced stocks that were just impossible to short: shorting cost = infinity. Today, these same stocks can be shorted for 1000%. Progress! Under this theory, the higher number of expensive-to-short stocks is a victory for market functionality, not a defeat.
In a dysfunctional system, many observed prices are meaningless. Consider a communist economy where the government fixes prices and there are widespread shortages. The price of cabbage at the store is fixed at 10 kopeks, but there is never any cabbage available in the store; the shelves are always empty. You can, however, buy cabbage on the black market for 50 kopeks. Suppose we switch to a free market. Now the store has cabbage freely available at a price of 20 kopeks. We observe a rise in the legal price of cabbage from 10 to 20, but the true price of cabbage has actually fallen. Maybe something similar happened in the stock loan market.
Living with huge costs
Given that we now live in a world with gigantic shorting costs for some stocks, how should we respond?
Consider the following scenario. We are sure that ABC is totally worthless, will never pay any dividend, and in 100 days will go bankrupt with delisting price of zero. In a frictionless market, the price of ABC should be zero today. However, we observe ABC has price of $100 and a daily shorting cost of $1.
Given our beliefs, what should we do? Should we short ABC? I don’t think so. While we are sure that ABC will fall $1 a day on average, we are not sure how fast this will occur. In order to benefit from shorting, we need the price to fall more than $1 a day during the period we are short.
Should we buy ABC? If we can lend it out at $1 a day for 100 days, we will at least break even. Now our profits are a mirror image from shorting; we win if the price falls less than $1 a day. Unfortunately, the fee might rise or fall over time, there is no guarantee it will remain at $1.
However, in practice, there is not perfect symmetry between the fees the borrower pays and the fee the lender gets. First, let’s not forget about Denzel, the lending agent. When Alice pays $1 a day to short, Bob only gets fifty cents, with Denzel getting the rest. Second, there is the issue of utilization. If Bob has 100 shares to lend, Denzel may only be able to find borrowers for 50 of those shares. In this case, Bob only gets a net 25 cents a share for the 100 shares he is willing to lend; Alice is meanwhile still paying $1 for the shares she is borrowing.
So, as a practical matter, what should we do? Well, being Denzel seems pretty nice, since he gets revenue with zero price risk. But, other than that, it’s a big mess. I don’t want to go long, since my price will fall $1 per day and will only be offset by 25 cents of lending revenue. I don’t want to go short, because that is break-even at best.
One thing is clear: if you must own ABC, you should be lending it out. It’s got a “dividend yield” of more than 1,000% annualized, of which you will receive more than 250% annualized. And yet most retail investors do not or cannot lend their shares. Indeed, in some cases, they proudly withhold their shares from the stock loan market in order to hurt the evil short sellers. That can be an expensive form of populist protest; similar to publicly burning your U.S. currency.
If, after reading this post, you are confused and frustrated, you are not alone. The stock loan market is messed-up. Maybe less messed-up than previously, but still a freakish place where many outcomes make little sense.
Perhaps someday, shorting a stock will be as simple as buying a latte. Until that blessed day occurs, we must survive in a dystopian system with hundreds of horrendously overpriced stocks, but nothing much that anyone can do about it.
You can now subscribe to Owenomics via our website!
References
Allen, Franklin, Marlene D. Haas, Eric Nowak, and Angel Tengulov. "Market efficiency and limits to arbitrage: Evidence from the Volkswagen short squeeze." Journal of Financial Economics 142, no. 1 (2021): 166-194.
Allen, Franklin, Marlene Haas, Eric Nowak, Matteo Pirovano, and Angel Tengulov. "Squeezing shorts through social media platforms." Swiss Finance Institute Research Paper 21-31 (2021).
Blocher, Jesse, and Matthew C. Ringgenberg. "Stock options, stock loans, and the law of one price." Vanderbilt Owen Graduate School of Management Research Paper 3087563 (2018).
Daniel, Kent D., Alexander Klos, and Simon Rottke. "Optimists, pessimists and stock prices." Available at SSRN 4700311 (2023).
D’Avolio, Gene. "The market for borrowing stock." Journal of Financial Economics 66, no. 2-3 (2002): 271-306.
Duffie, Darrell. Dark markets: Asset pricing and information transmission in over-the-counter markets. Vol. 6. Princeton University Press, 2012.
Harrison, J. Michael and David M. Kreps, 1978, Speculative investor behavior in a stock market with heterogeneous expectations, Quarterly Journal of Economics 92. 323-336.
Jones, Charles M., and Owen A. Lamont. "Short-sale constraints and stock returns." Journal of Financial Economics 66.2-3 (2002): 207-239.
Lamont, Owen, Short Sale Constraints and Overpricing, in Short selling: strategies, risks and rewards, (Frank J. Fabozzi ed., 2004).
Lamont, Owen A. "Go down fighting: Short sellers vs. firms." Review of Asset Pricing Studies 2.1 (2012): 1-30.
Lamont, Owen A., and Jeremy C. Stein. "Aggregate short interest and market valuations." American Economic Review 94, no. 2 (2004): 29-32.
Lamont, Owen A., and Richard H. Thaler. "Can the market add and subtract? Mispricing in tech stock carve-outs." Journal of Political Economy 111, no. 2 (2003): 227-268.
Miller, Edward M., 1977, Risk, uncertainty, and divergence of opinion, Journal of Finance 32, 1151-1168.
Reed, Adam V. "Short selling." Annu. Rev. Financ. Econ. 5, no. 1 (2013): 245-258.
Weill, Pierre-Olivier. "The search theory of over-the-counter markets." Annual Review of Economics 12 (2020): 747-773.
Endnotes