Insurers and Lenders as Monitors During Securities Litigation: Evidence from D&O Insurance Premiums, Interest Rates, and Litigation Costs

• Date: 01 September 2019
• DOI: http://doi.org/10.1111/jori.12231
• Author: Dain C. Donelson,Christopher G. Yust
• Published date: 01 September 2019

INSURERS AND LENDERS AS MONITORS DURING SECURITIES

LITIGATION:EVIDENCE FROM D&O INSURANCE

PREMIUMS,INTEREST RATES,AND LITIGATION COSTS

Dain C. Donelson

Christopher G. Yust

ABSTRACT

This study examines whether directors’ and officers’ insurers and lenders

effectively monitor securities litigation and respond through pricing before

case outcomes are known. By “monitoring,” we refer to tracking case

progress and obtaining information from the insured (defendant) firm and

its counsel prior to case resolution. We find that insurers and lenders increase

rates, and that this effect is almost completely isolated to firms with cases that

eventually settle. We confirm that this response is reasonable as settled cases

are associated with lower future earnings, while there is generally no relation

between future earnings and dismissed cases. As direct costs appear low, our

results suggest that most costs are indirect in the form of reputational

damage. Overall, our results suggest that researchers and policymakers

interested in litigation should focus on settled cases, which are the only cases

with material long-term costs.

INTRODUCTION

This study examines whe ther directors’ and office rs’ (D&O) insurers and le nders

effectively monitor securities litigation and respond through pricing before

case outcomes are known. Num erous studies focus on the r ole of D&O insurers

as ex ante monitors by pricin g policies based on facto rs such as corporate

governance (Core, 2000 ; Gupta and Prakash, 2012) a nd litigation risk (Gil lan and

Panasian, 2015). However, no studies examine the role of D&O insurers during

litigation. Thus, we first examine how D&O insurers respond to securities class

Dain C. Donelson is at the McCombs School of Business, University of TexasAustin. Donelson

can be contacted via e-mail: dain.donelson@mccombs.utexas.edu. Christopher G. Yust is at the

Mays Business School, Texas A&M University, College Station, TX 77843. Yust can be contacted

via e-mail: cyust@mays.tamu.edu. We thank Keith Crocker (the editor), two anonymous

referees, Matt Ege, Jen Glenn, John McInnis, and Brian Monsen for helpful comments and

suggestions. We thank Antonis Kartapanis for research assistance. We gratefully acknowledge

research support provided by the Red McCombs School of Business and the Mays Business

School. All errors are our own.

© 2017 The Journal of Risk and Insurance (2017).

DOI: 10.1111/jori.12231

1

663

Vol. 86, No. 3, 663–696 (2019).

actions that will be dismis sed versus those that will be se ttled while the cases are

pending.

This issue is important for two reasons. First, more than other types of

insurance, D&O insurance premiums are based on judgment, and risk

assessment is critical (Baker and Griffith, 2010). Thus, insurers must interpret

the signal from the filing of a securities class action prior to its resolution: does it

reveal information about the client’s level of risk, which should trigger a

premium increase, or does it simply reflect bad luck (i.e., the appearance of

managerial misconduct when no misconduct actually occurred), which should

notresultinanincrease?

Second, Baker and Griffith (2007, 2010) suggest that D&O insurers are not effective

monitors with respect to loss prevention (i.e., encouraging actions that would reduce

securities litigation risk) or managing defense costs during securities litigation itself.

However, Baker and Griffith (2007, 2010) also state that insurers manage settlements,

suggesting they monitor case details. By “monitoring,” we refer to tracking case

progress and obtaining information from the insured defendant firm and/or its

counsel prior to its resolution. While a case is ongoing, insurers receive private

information from defendants and their counsel (Baker and Griffith, 2010). Suits that

reflect bad luck are unlikely to represent substantial risk to the insurer for either the

current or future cases. Whether insurers can determine case merits is important

to the insurance economics literature as it improves our understanding of how

professionals price risk.

Specifically, we first test the effect of securities litigation on D&O insurance

premiums. We focus primarily on price responses because we are able to

observe price responses but cannot observe any direct or indirect influence the

insurer may have on case outcomes. Our sample uses securities litigation

alleging fraudulent conduct (i.e., Rule 10b-5 violations) to provide the strongest

test as this litigation likely imposes the greatest risk and costs to firms. It is not

currently clear whether these experts price cost differences during litigation

because these tests require the joint hypotheses that they can determine case

merits prior to case resolution, they are aware of differences in the cost for the

two types of cases, and they price such cost differences. Examining whether

D&O insurers respond differentially to settled versus dismissed cases before

resolution is a strong test of monitoring because most litigation-related

information is nonpublic, and the insurer could only adjust pricing based on

private information from the client.

Premiums are relativel y small, as even firms in the highest quintile of li tigation risk

have average premiums of only ar ound 0.21 percent of market value of e quity or

raw premiums of $2.4 millio n. Thus, although we find si gnificant increases in

premiums for litigation that eventually settles (compounded by the fact that

limits also increase, whi ch will further increase pr emiums), the cost is still low.

For dismissed cases, ther e is only a significant incr ease in premiums for the

relatively rare dismissed cases that survive more th an 2 years. This is consistent

with insurers quickly d etermining whether litigation is meritorio us, likely through

case monitoring.

2THE JOURNAL OF RISK AND INSURANCE

2The Journal of Risk and Insurance

664

To corroborate our results with respect to D&O insurance pricing, we also examine

how lenders respond to securities litigation. Like insurers, lenders have incentives to

find out if suits against their clients are likely to be dismissed or settled because

lenders monitor clients for issues that endanger repayment and have access to private

firm information (Fama, 1985). While Deng, Willis, and Xu (2014) find that interest

rates of bank loans are affected by securities litigation, they find no difference

between settled and dismissed cases. However, they wait until after case outcomes are

known to examine rate changes, rather than shortly after case filing. We find a short-

term interest rate increase, but only for settled cases. Overall, our findings regarding

D&O insurance premiums and interest rates suggest that both insurers and lenders

anticipate cost differences between settled and dismissed cases and promptly react to

firm securities litigation.

We then examinewhether this differential pricing betweensettled and dismissed cases

is consistent with litigation cost patternsacross case outcomes. Evidence of consistent

cost patterns would validate the pricing decisions by D&O insurers and lenders to

restrict price increases to settledlitigation. The cost of litigation is also important in its

own right as the United States is viewed as overly litigious (Brownand Schmit, 2008),

the alleged cost of litigationis high (U.S. Chamber Institute for Legal Reform (USCILR),

2014a),and firms seek to manage litigation costs (Graham, Harvey, and Rajgopal, 2005).

Many assume that both dismissed and settled cases impose large costs. For example,

Kim and Skinner (2012, p. 294) state that firms are concerned with filings, rather than

case outcomes, due to “legal, reputational, and time costs.” Large costs for dismissed

cases would be a concern as the Private Securities Litigation Reform Act of 1995

(PSLRA) was designed to minimize costs in such cases. Prior litigation cost research

focuses on stock returns at the lawsuit filing date (e.g., Griffin, Grundfest, and Perino,

2004; USCILR, 2014a), which should capture the total loss in theory. However, filing

returns are subject to measurement issues, including difficulty separating litigation

costs from other costs (Griffin, Grundfest, and Perino, 2004).

We thus estimate costs with both returns and financial statement information. In

contrast to returns, financial statements allow us to separate the direct costs, such as

settlements funded by the defendant firm or increased D&O insurance premiums,

and the indirect costs, such as reputational harm through lost revenue from

customers who are hesitant to deal with the firm (Karpoff, Lee, and Martin, 2008).

Finally, reputational harm captured in the (discounted) future abnormal reduction in

earnings (i.e., the difference in earnings between the sued firm and an matched firm)

should equal damages captured in stock returns (see Penman and Sougiannis, 1997)

but minimizes the over- and understatement issues with filing date returns.

Nonetheless, while financial statement data improve the ability to measure litigation

costs, they are still imperfect because the very information released at the end of the

class period that results in a large stock price drop and triggers litigation may also

result in revisions of future earnings and cash flow expectations. Thus, we use three

alternative samples. The first two minimize potential bias by providing upper and

lower bounds on cost estimates. The final sample, which includes only sued firms,

avoids these issues because all sued firms have similar price drops. Thus, while no

sample is perfect, we are able to “triangulate” results across tests.

INSURERS AND LENDERS AS MONITORS DURING LITIGATION 3

INSURERS AND LENDERS AS MONITORS DURING LITIGATION 3

665