The Paradox of Financial Fire Sales: The Role of Arbitrage Capital in Determining Liquidity

• DOI: http://doi.org/10.1111/jofi.12584
• Author: JUNGSUK HAN,JAMES DOW
• Date: 01 February 2018
• Published date: 01 February 2018

THE JOURNAL OF FINANCE •VOL. LXXIII, NO. 1 •FEBRUARY 2018

The Paradox of Financial Fire Sales: The Role

of Arbitrage Capital in Determining Liquidity

JAMES DOW and JUNGSUK HAN∗

ABSTRACT

How can fire sales for financial assets happen when the economy contains well-

capitalized but nonspecialist investors? Our explanation combines rational expec-

tations equilibrium and “lemons” models. When specialist (informed) market par-

ticipants are liquidity-constrained, prices become less informative. This creates an

adverse selection problem, decreasing the supply of high-quality assets, and lowering

valuations by nonspecialist (uninformed) investors, who become unwilling to sup-

ply capital to support the price. In normal times, arbitrage capital can “multiply”

itself by making uninformed capital function as informed capital, but in a crisis, this

stabilizing mechanism fails.

IN A FIRE SALE,SELLERS ARE FORCED TO SELL assets at deep discounts because no

one is willing to buy them at fair prices. Sellers can be forced to sell because

of financial distress, credit market frictions, regulation, margin calls, etc.1

Why do fire sales happen? What makes investors avoid buying assets that are

apparently cheap? One explanation is offered by Shleifer and Vishny (1992): if

industry experts with higher private valuations do not have enough liquidity,

assets are bought at a discount by nonexperts who cannot use them efficiently.

This argument applies naturally to real assets rather than financial securities.

But since financial securities typically require that the holder only collects

cash flows, not operates the assets, there should not be significant differences

in private valuations.2

∗James Dow is from London Business School and Jungsuk Han is from Stockholm School of

Economics. We thank three anonymous referees, an Associate Editor,and Bruno Biais (the Editor)

for very helpful comments. We also thank Brandon Daley, Marco Di Maggio, Alex Edmans, Julian

Franks, Simon Gervais, Vincent Glode, John Kuong, Pete Kyle, Paolo Sodini, Per Str¨

omberg, and

J¨

orgen Weibull, as well as seminar participants at AFA 2016, Brevan Howard/FTG conference

2016, Cambridge Corporate Finance Theory Symposium 2015, FIRS 2016, Jackson Hall finance

conference 2016, BI Norwegian Business School, INSEAD, Korea University, London Business

School, Sungkyunkwan University,Sungkyunkwan GSB, Stockholm School of Economics, Univer-

sity of Gothenburg, University of Mannheim, and Yonsei University. Dow declares that he has not

received any financial support for the research described in this paper, and he also has no relevant

or material financial interests that relate to the research described in this paper. Han declares

that he has not received any financial support for the research described in this paper, and he also

has no relevant or material financial interests that relate to the research described in this paper.

1See, for example, Shleifer and Vishny (2011)forasurvey.

2Some differences in valuation for financial securities do exist. For example, some holders are

able to repo the security and others are not. Also, expert investors may understand the risk of

DOI: 10.1111/jofi.12584

229

230 The Journal of Finance R

Allen and Gale (1994), Shleifer and Vishny (1997), Gromb and Vayanos

(2002), Brunnermeier and Pedersen (2009), and Duffie (2010) give an alterna-

tive explanation based on limited arbitrage capital. They argue that complex

securities are normally held by specialized investors who can understand

them properly, so that if those investors become severely capital-constrained,

the available pool of specialized capital is too small to pay full value for the

outstanding stock of assets. It is true that if all specialized investors are

capital-constrained, they are not able to buy undervalued assets. But what

about nonspecialized investors? Many episodes described in the literature as

fire sales (such as the Long-Term Capital Management crisis and the collapse

of the mortgage-backed securities (MBS) markets during the financial crisis

of 2007 to 2008) occurred even though there were plenty of well-capitalized

investors somewhere in the world economy (e.g., Warren Buffett, or sovereign

wealth funds). Why would not well-capitalized outside investors want to step

in and buy undervalued assets, thereby preventing large price drops? This

seems paradoxical.

A possible resolution of the paradox is based on “lemons” problems as de-

scribed by Akerlof (1970). Nonspecialized investors could be exposed to adverse

selection perpetrated by industry insiders. This could happen if insiders choose

to hold high-quality assets, and sell only low-quality assets, causing a classic

lemons problem whereby anyone who is forced to sell high-quality assets would

suffer a loss. But this explanation has a major flaw because it does not explain

why the lemons problem would suddenly increase during a crisis. In a crisis,

asset sellers should have less discretion on selling, and thus liquidity short-

ages should mitigate the lemons problem (e.g., Malherbe (2014)). The paradox

therefore remains.

We give an intuitive answert o thisquestion by examining therole of informed

trading in preventing adverse selection. We argue that informed traders who

can buy assets help make market prices informative. This prevents high-quality

assets from trading at prices similar to low-quality assets. Hence, informed buy-

ers remove the incentive for sellers to sell low-quality assets. In case of a large

shock to the market, however, this mechanism breaks down. A severe need for

liquidity that affects specialized traders prevents them from using their private

information to bid up undervalued assets. This can make prices uninformative,

leading to adverse selection whereby sellers predominantly supply overvalued

assets to the market. This, in turn, leads uninformed agents, who are potential

buyers for those assets, to withdraw from the market even though they are

not wealth-constrained. The result is a market freeze for high-quality assets:

they are not traded, except when their owners are subject to a severe liquidity

shortage and therefore forced to sell at fire-sale prices.

To formalize this argument, we develop an information-based theory of fire

sales using a noisy rational expectations equilibrium framework with endoge-

nous adverse selection. Using this framework, we explain the role of informed

a security better, allowing them to hedge their overall portfolio more effectively. However, such

differences in private valuations should be minor.

The Paradox of Financial Fire Sales 231

trading in fire sales and market freezes. Unlike the traditional literature on

adverse selection, which features only informed sellers, we highlight the role

of informed traders who compete to exploit mispricing and thereby make the

price more informative. In addition, we address the question of why fire sales

occur even when, somewhere in the economy, there are traders with enough

capital to correct prices. We show that a liquidity shortage for informed traders

can deter uninformed, well-capitalized traders from buying, which can lead

asset prices to fall. This result sheds light on the paradoxical nature of fire

sales whereby capital moves out of the market when it is needed most and

would seemingly earn higher returns. Finally, we explain why a market freeze

happens at the same time as fire sales. Our paper therefore sheds light on the

“double whammy” situation, in which fire sales and low trading volume occur

together (see, for example, Tirole (2011) for a discussion of the financial crisis

of 2007 to 2008).

Consider a two-period model with some participants who are informed but fi-

nancially constrained, and some who are unconstrained but uninformed. There

exists a marketable asset with risky payoffs whose value is known only to the

informed participants.3There are two types of informed participants: a dis-

tressed seller who needs to raise liquidity, and arbitrageurs who aim to make

trading profits. The seller is forced to meet liquidity needs, which he can do

either by selling his holdings of the marketable asset or by liquidating a non-

marketable asset (such as profit-generating operations).4This choice leads to

a lemons problem: the seller sells the marketable asset only when it is overval-

ued (a lemon) or when liquidating the nonmarketable asset is very costly.There

are two types of uninformed participants: hedgers and risk-neutral investors.

Hedgers trade to hedge against their risk exposure, which is correlated with

the payoff of the marketable asset. The randomness in hedgers’ risk exposure

creates trading noise that prevents full revelation of informed participants’

private information.

Under normal market conditions where arbitrageurs have enough liquidity,

the price reveals the fundamental value of the asset because informed trading

volume overwhelms the impact of noise. This allows the seller facing a liq-

uidity shortage to fund itself by selling the marketable asset at intrinsic value

regardless of the quality of the asset. This, in turn, makes uninformed investors

willing to absorb the supply of the asset without worrying about adverse se-

lection. In periods of crisis where arbitrageurs are liquidity-constrained, prices

are less informative. This makes the seller willing to supply only low-quality

assets to the market, except in cases in which it is effectively forced to sell be-

cause the alternative of liquidating the nonmarketable asset is very costly.This

adverse selection problem makes uninformed investors unwilling to absorb the

supply of assets unless there is a drop in price to reflect a lemons discount. It

3In practice, they are likely to be financial intermediaries such as banks and hedge funds.

4The nonmarketable asset will produce low levels of output if it is liquidated early. Similar

assumptions are standard in the literature, for example, the assumption of illiquid assets held by

banks in Diamond and Dybvig (1983).