For corporate finance to truly advance we need more genuinely testable models

• Published date: 01 November 2023
• Author: Matthew Spiegel
• Date: 01 November 2023
• DOI: http://doi.org/10.1111/fire.12346

DOI: 10.1111/fire.12346

FUTURE DIRECTIONS

For corporate finance to truly advance we need

more genuinely testable models

Matthew Spiegel

YaleSchool of Management, New Haven,

Connecticut, USA

Correspondence

Matthew Spiegel, YaleSchool of Management,

P.O. Box208200, New Haven CT 06520, USA.

Email: matthew.spiegel@yale.edu

April 23, 2023

Abstract

Corporate finance has turned into a field where researchers

produce what seem like a constant flow of disconnected

papers. Theories are never confirmed or refuted. At best,

empirical papers confirm models without a realistic alterna-

tive to refute. The problem is that today’s models are either

static or have firms that never interact directly with other

firms. Most industries are oligopolies. For firms in these

industries the competition’s actions in the product market

are likely of paramount importance. If corporate finance is

going to progress, we need papers with testable dynamic

oligopoly models. Models where firms compete directly with

each other.

KEYWORDS

corporate finance, hypothesis testing, industry structure, modeling

Wayback when you were a graduate student you took courses covering asset pricing theory. (For some readers I sus-

pect that was just 3 years ago. For my cohort it was, let me just say,much longer ago!) What do you remember about

the models? One aspect that always struck me was the structure. Peoplebought and sold stocks in what was generally

a perfectly competitive market. They did this along with consuming funds to optimize their lifetime utility.The stocks

themselves produced returns via a well specified return process. In principle, one could gather up data from the set of

available sources and test the theories. In a sense, the theories were built to fit data that might be available within a

setting that included many of the trading environment’simportant elements.

Were those asset pricing models completely realistic? No. People do not trade continuously.For most people, it is

not even close. Even professional money managers are not trading everything everyminute. If you go online around

2:00 p.m. Eastern time, it is easy to find stocks that havenot traded in the past hour. Markets are not perfectly compet-

itive. In manycases, market orders worth a few thousand dollars can climb up or down the limit order book. Even small

trades are subject to the bid-ask spread. Beyond that, stock returns are considerably more right skewedthan models

typically allow for.However, even with all these flaws one can take the available data and see which models offer the

best fit.

Financial Review. 2023;58:657–661. wileyonlinelibrary.com/journal/fire ©2023 The Eastern Finance Association. 657