10 accounting theory questions.

M Y a s o

I s s S

E F p

THE ACCOUNTING REVIEW American Accounting Association Vol. 86, No. 2 DOI: 10.2308/accr.00000031 2011 pp. 367–383

Accounting Scholarship that Advances Professional Knowledge and Practice

Robert S. Kaplan Harvard University

ABSTRACT: Recent accounting scholarship has used statistical analyses on asset prices, financial reports and disclosures, laboratory experiments, and surveys of prac- tice. The research has studied the interface among accounting information, capital markets, standard-setters, and financial analysts and how managers make accounting choices. But as accounting scholars have focused on understanding how markets and users process accounting data, they have distanced themselves from the accounting process itself. Accounting scholarship has failed to address important measurement and valuation issues that have arisen in the past 40 years of practice. This gap is illustrated with missed opportunities in risk measurement and management and the estimation of the fair value of complex financial securities. This commentary encour- ages accounting scholars to devote more resources to obtaining a fundamental under- standing of contemporary and future practice and how analytic tools and contemporary advances in accounting and related disciplines can be deployed to improve the profes- sional practice of accounting.

Keywords: accounting research; professional practice; accounting education; field studies; risk measurement; pension risk; fair value measurement; options pricing model.

I. INTRODUCTION y talk is a response to a request from a 28-year-old, newly minted doctoral graduate soliciting advice about how to select research topics and teaching assignments for her forthcoming career as an accounting academic.

You are embarking on an exciting five-decade faculty career as a business school professor. ou have acquired knowledge of contemporary accounting research issues and excellent disciplin- ry training in mathematics, economics, finance, statistics, and the behavioral sciences. But you hould understand that you are joining the faculty of a professional school, not a graduate school r academic department. Your profession could be viewed as “accounting” or somewhat more

am indebted to two HBS colleagues: Robert Merton, who inspired me with the opportunities to advance accounting cholarship and practice by building upon his pioneering work in financial economics; and Rakesh Khurana, for discus- ions about the role for scholarship in professional schools. I also benefited from discussions with Scott Richard, Wharton chool, about the feasibility and desirability of fair value measurements for financial instruments.

ditor’s note: This commentary, based on a lecture at the 2010 American Accounting Association Annual Meeting in San rancisco, CA, was invited by Senior Editor Steven Kachelmeier, consistent with the AAA Executive Committee’s goal to romote broad dissemination of the AAA Presidential Scholar Lecture.

Submitted: September 2010 Accepted: October 2010

Published Online: March 2011

367

b t m fi

fi k a w t m

b

m f i fl t h i a d p p r s

a o c c p a a s p

�

368 Kaplan

T A

roadly as “management” �Khurana 2007�. Among all the management disciplines, accounting is he closest to a professional discipline—many of the students we educate get certified and must

aintain professional standing through continuing education programs. These accounting and nancial professionals are our clients.

As a scholar in a professional school, you will educate current and future professionals on the eld’s common body of knowledge. You will learn how to dispense the existing common body of nowledge in ways that your students can put into action during their professional careers. All ccounting and management professors have this responsibility. But also important, though less idely known, accepted, and advocated, academic scholars at a professional school should con-

ribute to advancing the profession’s body of knowledge, especially when innovation is high and ajor changes are occurring in the practice environment of the profession.

For building and sustaining a successful scholarly career in a professional school, you should e continually thinking about three fundamental questions:

1. What are the big issues faced by our practice community? 2. What are the comparative advantages that accounting scholars can bring to address these

fundamental accounting and management issues? My partial list for areas in which ac- counting scholars should have a comparative advantage, relative to scholars in other managerial disciplines, includes measurement, reporting and disclosure, auditing, finan- cial analysis, and management control.

3. How does our research advance knowledge in these core areas of our discipline?

II. ACCOUNTING RESEARCH: 1968–2010 During the past 40� years, research by management and accounting scholars has produced

any important insights and advanced our understanding of the environment in which accounting unctions. Much of this scholarship has focused on how information and markets interact, includ- ng the information content of accounting numbers, the role of accounting accruals versus cash ows, voluntary and involuntary managerial disclosures, efficiency and anomalies in markets, and

he impact of accounting choices made by managers. Some scholarship has given us insights in ow financial analysts do their work and their influence on markets. I have been particularly mpressed by the development of the accounting-based valuation model �Ohlson 1995; Feltham nd Ohlson 1995�. This is a particularly good, although somewhat rare, example of where aca- emic accounting scholarship has created an important innovation that advanced professional ractice. Academics have contributed to standard-setting practice by testing the reactions to pro- osed and implemented standards and communicating the implications of capital-market-based esearch to standard-setters. Accounting scholarship has also informed and advanced the profes- ional work of auditors and tax professionals �AAA Research Impact Task Force 2009�.

Most of these research contributions have come from statistical studies of phenomena, using rchival data from financial statements, proxy statements, market prices of equity, debt, and ccasionally other financial instruments, credit ratings, loan documents, bankruptcies, conference alls, and analyst forecasts. Accounting scholars have also performed statistical analysis on data ollected from carefully designed laboratory experiments and simulations, and from surveys of ractitioners. One characteristic of all this research is the existence of multiple tables within the rticle populated with estimated coefficients, at least a few of which have between one and three sterisks near them, attesting to their degree of statistical, although not necessarily economic, ignificance �McCloskey 1998�. Scholars have also studied, both analytically and empirically, the roperties of management contracts and incentives.

Table 1, extracted from the annual reports of The Accounting Review over the past two years Kachelmeier 2009, 2010�, summarizes the research methods used by accounting scholars in the

he Accounting Review March 2011 merican Accounting Association

p A o a �

p b m a i p m t e n

e f T h p n i i b h o r c p

R

E E A S F T

S

Accounting Scholarship that Advances Professional Knowledge and Practice 369

T

apers submitted and accepted during the two-year period 2008–2010 to the American Accounting ssociation’s leading research journal, The Accounting Review. The table shows the predominance f statistical analysis on archival, mostly financial, data, and also the extensive use of labor- tory experiments, surveys, and analytic models. For future reference, note the low market share 1 percent� of field and case studies.

Overall, I think this is a good record, in fact a very good record, of accomplishment. Com- ared to 45 years ago, we have had a great expansion of research performed and published by a roader academic accounting community, and some of this research has helped us understand, uch better than in 1968, the role of accounting reporting and disclosure in markets and contracts,

nd how people process and use accounting information. But we are likely encountering dimin- shing returns in several of our research areas. The increment in our knowledge from the 371st aper on accruals versus cash flows, earnings management, or voluntary disclosure is undoubtedly uch lower than the contribution from the first five to ten papers that introduced each topic into

he academic accounting literature. Accounting scholars, like many of their academic colleagues, xhibit strong herding effects; they follow where others have already gone rather than forging a ew path by studying a new issue in an innovative way.

III. THE GAP BETWEEN ACCOUNTING SCHOLARSHIP AND PRACTICE If academics in professional school limit their research agenda to issues that can be ad-

quately addressed by a narrow set of generally accepted research methods, then they must wait or phenomena to happen to generate sufficient archival data for them to analyze statistically. hus, much of accounting �and management� research for the past 40 years has been reactive. It as studied, evaluated, and explained existing practice, but has not contributed to advancing that ractice �the accounting-based-valuation model, and improvements in auditing and tax being otable exceptions�. This situation would be adequate if little change or innovation was occurring n the profession. But over the past 50 years, some major discontinuities have occurred and the mplications of these changes have yet to make their way into the common body of knowledge eing studied by academic scholars and taught to future professionals. Among these changes are uge increases in the volume, velocity, volatility, and complexity of transactions, the globalization f markets leading to trading of assets in different time zones and currencies and to issues in eporting and managing across unclear organizational boundaries, and the introduction of new and omplex securities and derivative contracts that shift risks among companies and market partici- ants. Society has increased its expectations about corporate governance and oversight of execu-

TABLE 1

Submissions and Acceptance to The Accounting Review (2008–2010)

esearch Method # Submitted # Accepted % Submitted % Accepted

mpirical-Archival 933 93 75% 65% xperimental 175 24 14% 17% nalytic 96 18 8% 13% urvey 30 6 2% 4% ield and Case Study 15 2 1% 1% otal 1249 143 100% 100%

ource: Kachelmeier �2009, 2010�.

he Accounting Review March 2011 American Accounting Association

t � a

i p f r m

f i

o a t a a t

a s � F i b

p t

1

2

370 Kaplan

T A

ive compensation. A major management technology innovation has occurred for pricing risk Black and Scholes 1973; Merton 1973�, which has been overlooked and underappreciated by ccounting scholars for the past four decades �much more on this later�.

By responding slowly if at all to major new challenges and opportunities in the environment n which accounting is practiced, accounting scholars have become less familiar with emerging rofessional challenges and opportunities. Few have helped to craft new solutions and approaches or people practicing in the profession. For example, we failed to see in a timely fashion the eporting, valuation, and disclosure implications from the new types of mortgage lending and the assive securitizations of these loans.

Interestingly, accounting scholars are not the only academics who have become disconnected rom practice. Paul Krugman �1994, 43–44� describes a similar situation in development econom- cs, which he called the “hollowing of Africa” effect:

The paper, “The Evolution of European Ignorance about Africa,” describes how European maps of the African continent evolved from the 15th to the 19th centuries … The coastline of Africa was first explored, then plotted with growing accuracy, and by the 18th century that coastline was shown in a manner essentially indistinguishable from that of modern maps … On the other hand, the interior emptied out. The weird mythical creatures were gone, but so were the real cities and rivers. In a way, Europeans had become more ignorant about Africa than they had been before … improvement in the art of mapmaking raised the standard for what was considered valid data. Second-hand reports of the form “six days south of the end of the desert you encounter a vast river flowing from east to west” were no longer something you would use to draw your map. Only features of the landscape that had been visited by reliable informants equipped with sextants and compasses now qualified.

Today’s accounting academics, after 40 years of using rigorous social science research meth- ds, know much more about the coastline of accounting, its interface between accounting reports nd capital markets, analysts, auditors, regulators, boards, and the media. But they know substan- ively less than academics of 40 years ago about the “interior” of leading-edge professional ccounting and finance practice. We have experienced a “hollowing of professional practice” mong accounting scholars that you and a new generation of accounting faculty could produc- ively fill.

More than 30 years ago, Roethlisberger, a social scientist and management scholar who was coauthor on the Hawthorne effect experiments, described the research preferences of academic

ocial scientists to work on topics that can be explored using statistical and mathematical methods Roethlisberger 1977, 390�. He illustrated his points using the Knowledge Enterprise diagram in igure 1.1 Roethlisberger observed that the clinical approach of acquiring knowledge by describ-

ng and classifying the activities of skilled practitioners is frequently bypassed by researchers ecause clinical research is considered less elegant than deductive research.2

Donald Schön �1992, 54� also expressed concern that systematic study and description of rofessional practice holds a less exalted view in academia than deductive theory and hypothesis esting:

In the varied typography of professional practice, there is a high, hard ground which overlooks a swamp. On the high ground, manageable problems �N.B. think of simple principal-agent models� lend themselves to solution through the use of research-based theory and technique. In the swampy lowlands, problems are messy and confusing and incapable of technical solution … �But� the

Summaries of the activities at the various stages of the knowledge tree—description, classification, measurement, discovering relationships, theory building, and theory testing—can be found in Kaplan �1986�. On academic scholars’ desire for elegance, I add a comment by Einstein �1920, 5 �Preface��, “leave elegance to the tailor,” which he attributed to the 19th century physicist, Ludwig Boltzmann.

he Accounting Review March 2011 merican Accounting Association

p m t l f b

S

Accounting Scholarship that Advances Professional Knowledge and Practice 371

T

problems of the high ground tend to be relatively unimportant to individuals or to society at large, however great their technical interest may be, while in the swamp lie the problems of greatest human concern … Some researchers … pursue an agenda driven by evolving questions of model- ing theory and techniques, increasingly divergent from the contexts of actual practice.

I am less pessimistic than Schön about whether rigorous research can inform professional ractice �witness the important practical significance of the Ohlson accounting-based valuation odel and the Black-Merton-Scholes options pricing model�, but I concur with the general point

hat academic scholars spend too much time at the top of Roethlisberger’s knowledge tree and too ittle time performing systematic observation, description, and classification, which are at the oundation of knowledge creation. Henderson �1970, 67–68� echoes the benefits from a more alanced approach based on the experience of medical professionals:

both theory and practice are necessary conditions of understanding, and the method of Hippocrates is the only method that has ever succeeded widely and generally. The first element of that method is hard, persistent, intelligent, responsible, unremitting labor in the sick room, not in the library … The second element of that method is accurate observation of things and events, selection, guided by judgment born of familiarity and experience, of the salient and the recurrent phenomena, and their classification and methodical exploitation. The third element of that method is the judicious

FIGURE 1 Knowledge Enterprise Diagram

The phenomena

Analytical (scientific) knowledge

Clinical knowledge

Skill

Levels

Characteristic Statements (theories) Methods Products

Knowledge of

acquaintance

Practice and

reflection

How‐to‐do‐it statementsand

aphorisms

Conceptual schemes

Classification

Observation and Interviewing

Descriptive cases and syndromes

Taxonomies

Elementary concepts

Empirical propositions

General propositions

Creative and inductive leap of imagination

Operational definitions; rigorous measurement

Definition of concepts and variables

Elementary measurement

Statementsof the form x varies with y

Statementsof the form x varies with y under given conditions

Deductive systems

(For the development of knowledge, read from the bottom up; for the practice of knowledge, read from the top down.)

ource: Roethlisberger (1977, 393).

he Accounting Review March 2011 American Accounting Association

g p

a o l c h

I

p s e t o c t

fi s a i a s s

i r r B t 2 s c i b i a p

372 Kaplan

T A

construction of a theory … and the use thereof … �T�he physician must have, first, intimate, habitual, intuitive familiarity with things, secondly, systematic knowledge of things, and thirdly an effective way of thinking about things.

More recently, other observers of business school research have expressed concerns about the ap that has opened up in the past four decades between academic scholarship and professional ractice. Examples include:

Historical role of business schools and their faculty is as evaluators of, but not creators or origi- nators of, business practice. �Pfeffer 2007, 1335�

Our journals are replete with an examination of issues that no manager would or should ever care about, while concerns that are important to practitioners are being ignored. �Miller et al. 2009, 273�

In summary, while much has been accomplished during the past four decades through the pplication of rigorous social science research methods to accounting issues, much has also been verlooked. As I will illustrate later in these remarks, we have missed big opportunities to both earn from innovative practice and to apply innovations from other disciplines to important ac- ounting issues. By focusing on these opportunities, you will have the biggest potential for a ighly successful and rewarding career.

ntegrating Practice and Theory: The Experience of Other Professional Schools Other professional schools, particularly medicine, do not disconnect scholarly activity from

ractice. Many scholars in medical and public health schools do perform large-scale statistical tudies similar to those done by accounting scholars. They estimate reduced-form statistical mod- ls on cross-sectional and longitudinal data sets to discover correlations between behavior, nutri- ion, and health or sickness. Consider, for example, statistical research on the effects of smoking r obesity on health, and of the correlations between automobile accidents and drivers who have onsumed significant quantities of alcoholic beverages. Such large-scale statistical studies are at he heart of the discipline of epidemiology.

Some scholars in public health schools also intervene in practice by conducting large-scale eld experiments on real people in their natural habitats to assess the efficacy of new health and afety practices, such as the use of designated drivers to reduce alcohol-influenced accidents. Few cademic accounting scholars, in contrast, conduct field experiments on real professionals working n their actual jobs �Hunton and Gold �2010� is an exception�. The large-scale statistical studies nd field experiments about health and sickness are invaluable, but, unlike in accounting scholar- hip, they represent only one component in the research repertoire of faculty employed in profes- ional schools of medicine and health sciences.

Many faculty in medical schools �and also in schools of engineering and science� continually nnovate. They develop new treatments, new surgeries, new drugs, new instruments, and new adiological procedures. Consider, for example, the angiogenesis innovation, now commercially epresented by Genentech’s Avastin drug, done by Professor Judah Folkman at his laboratories in oston Children’s Hospital �West et al. 2005�. Consider also the dozens of commercial innova-

ions and new companies that flowed from the laboratories of Robert Langer at MIT �Bowen et al. 005� and George Whiteside at Harvard University �Bowen and Gino 2006�. These academic cientists were intimately aware of gaps in practice that they could address and solve by applying ontemporary engineering and science. They produced innovations that delivered better solutions n actual clinical practices. Beyond contributing through innovation, medical school faculty often ecome practice thought-leaders in their field of expertise. If you suffer from a serious, complex llness or injury, you will likely be referred to a physician with an appointment at a leading cademic medical school. How often, other than for expert testimony, do leading accounting rofessors get asked for advice on difficult measurement and valuation issues arising in practice?

he Accounting Review March 2011 merican Accounting Association

i m p

e e

I

y c s m i a c a p c

s i

w a f l a

c d fi t

q

Accounting Scholarship that Advances Professional Knowledge and Practice 373

T

One study �Zucker and Darby 1996� found that life-science academics who partner with ndustry have higher academic productivity than scientists who work only in their laboratories in

edical schools and universities. Those engaged in practice innovations work on more important roblems and get more rapid feedback on where their ideas work or do not work.

These examples illustrate that some of the best academic faculty in schools of medicine, ngineering, and science, attempt to improve practice, enabling their professionals to be more ffective and valuable to society.

mplications for Accounting Scholarship To my letter writer, just embarking on a career as an academic accounting professor, I hope

ou can contribute by attempting to become the accounting equivalent of an innovative, world- lass accounting surgeon, inventor, and thought-leader; someone capable of advancing profes- ional practice, not just evaluating it. I do not want you to become a “JAE” �Just Another Epide- iologist�. My vision for the potential in your 40� year academic career at a professional school

s to develop the knowledge, skills, and capabilities to be at the leading edge of practice. You, as n academic, can be more innovative than a consultant or a skilled practitioner. Unlike them, you an draw upon fundamental advances in your own and related disciplines and can integrate theory nd generalizable conceptual frameworks with skilled practice. You can become the accounting ractice leader, the “go-to” person, to whom others make referrals for answering a difficult ac- ounting or measurement question arising in practice.

But enough preaching! My teaching is most effective when I illustrate ideas with actual cases, o let us explore several opportunities for academic scholarship that have the potential to make mportant and innovative contributions to professional practice.

IV. RISK MEASUREMENT AND MANAGEMENT I �and probably many others� became aware of the gaps in risk management practice in 2007

hen important institutions began to fail unexpectedly. As this trend continued through 2008, I felt lot like Yogi Berra when he said, “It’s déjà vu all over again.” The 2007/08 risk management

ailures took me back to the early 1980s �Kaplan 1983; Johnson and Kaplan 1986�, when we earned, to our surprise, that cost accounting and performance measurement, highly important ccounting functions, were being done poorly by many enterprises.

Risk management is a great issue for accounting academics. It builds upon our fundamental omparative advantages as an academic discipline: expertise in measurement, valuation, reporting, isclosure, management control, auditing, and governance. The topic contains issues relevant for nancial reporting, management control, and auditing, three of our most important subject areas in

he American Accounting Association. Among the risk management issues I would like to know more about, consider the following

uestions:

1. How can we measure or quantify risk? Measurement is about the past; even so-called leading indicators are measuring events that have already occurred. How can we quantify risk or develop risk indicators for an event that has not yet occurred and, we hope, may never occur? Quantifying risk exposure is a challenging measurement issue.

2. Will a single risk management system work for all types of risks, or do different types of risks require different types of risk management systems?

3. The optimal level of risk is not zero. How can companies introduce risk management without sacrificing their innovation and risk-taking activities? What is the appropriate balance between innovation and risk management, and how can this balance be main- tained?

he Accounting Review March 2011 American Accounting Association

O

D l s L W a m a

m m z c t d p b m s u

i e

374 Kaplan

T A

4. Much work is currently being done by regulators and standard-setters �COSO, Basel, rating agencies, stock exchanges, etc.� to promulgate rules and standards on companies’ risk management practices:

a. Is the practice of risk management sufficiently stable, mature, and understood that now is a good time to develop risk standards and regulations? Or is it better for companies to innovate and experiment with different risk management approaches before regu- lators standardize and codify practices?

b. How can academic scholars participate in and document such innovation and experi- mentation?

c. Does and should the optimal form of risk management vary by type of company, strategy, degree of innovation? How?

5. What is the relationship between a professional risk management officer and line man- agement?

6. What is the relationship between risk management and internal audit? 7. Other than perhaps in financial institutions, what do people mean when they talk about

“risk appetite”? When the Board is told it must determine the “risk appetite” for the company, how can risk appetite be quantified and made operational?

8. Can scholarship improve the reporting and disclosure of risk in company documents?

n this last question, I looked at a recent risk disclosure in a company’s 10-K:

In the event of changes in market conditions, such as interest or foreign exchange rates, equity, fixed income, commodity or real estate valuations, liquidity, availability of credit or volatility, our business could be adversely affected in many ways … Further declines in real estate values in the U.S. and continuing credit and liquidity concerns could further reduce our level of mortgage loan originations and increase our mortgage inventory while adversely affecting its value. �emphasis added�

oes this sound like the risk exposure of a huge financial institution that would file for bankruptcy ess than two months after this 10-K submission? Yet this was the “risk disclosure” in the 2008 econd quarter filing of Lehman Brothers, a financial institution born in the South in the 1850s. ehman survived the U.S. Civil War, World War I, the Great Depression of the 1930s, and World ar II. It built its capital during the great post-World War II global expansion, and somehow failed

fter a 5 percent decline in U.S. real estate prices. After 40� years of academic research on capital arkets and financial economics, is Lehman’s 10-K disclosure the best we can offer to quantify

nd disclose a company’s risk exposure? I hope not. At present, we do not know enough to address the risk management questions with research

ethods at the top of the “knowledge enterprise.” If we rely on traditional social science research ethods, we will provide sketchy answers and only with long lags. Risk management in organi-

ations is highly complex and context-specific. It involves many moving and interconnected omponents including leadership, organizational structure, incentives, finance, management con- rol, and … risk and uncertainty. We are going to learn little by sitting in our offices and analyzing ata that others produce. Similar to how we went about developing new approaches to the cost and erformance measurement issues in the 1980s, we should start by using research methods at the ase of the knowledge enterprise tree: identify, observe and describe the practice leaders, docu- ent and analyze the practice innovations, develop the associations of where risk management

eems to be working well and where it has failed, and start to formulate the general principles nderlying effective risk management practice.

While this research agenda will take years of work, I can describe several accounting topics n risk measurement that could be immediately addressed by creatively applying existing knowl- dge.

he Accounting Review March 2011 merican Accounting Association

M

d i p c c a

f d e p C t

s p l i i t p

fi b i c f

w m a t t a s h q p

Accounting Scholarship that Advances Professional Knowledge and Practice 375

T

easuring Pension Plan Risk

Pension accounting has been addressed by standard-setters for decades, yet little has been one on measuring and reporting the risk of pension plans. Measuring the risk of pension plans is mportant not only for companies with defined benefit pension plans, but also for states, munici- alities, and even countries. Consider an example, adapted from Merton �1995, 473�, of two ompanies, each with $100 million in net present value of pension liabilities �incidentally, the alculation of pension liabilities is a measurement that accountants have already out-sourced to ctuaries�:

Company A holds $90 million in default-free government bonds whose durations are matched perfectly with the liabilities in its pension plan. Company B holds $25 million in such duration- matched government bonds plus $75 million in a diversified portfolio of common stocks. Which pension plan is riskier?

Today’s accounting standards would report that A’s plan is riskier because B’s plan is “fully unded” while A’s is underfunded by $10 million. But Company B holds the equivalent of a giant erivative. It could have held $100 million in duration-matched risk-free assets, but it made the quivalent of a $75 million total return on equity-for-debt swap. Now, we all know the equity risk remium is about 800 basis points higher return per year than the return from risk-free bonds. ompounded over many years, a diversified portfolio of common stocks should return far more

han the risk-free portfolio. So how risky could B’s equity-for-debt swap be? Suppose the trustees of B’s plan conduct the following thought experiment. Even though

tocks are expected to out-perform bonds over long periods of time, how much would it cost to urchase insurance to guarantee that the return on the equity portfolio, 20 years from now, is not ower than what could be achieved in a zero-risk government bond portfolio? While I have not ndependently calculated the following, an expert in this field has estimated to me that such nsurance would cost $15 million or more. So it would cost plan B’s trustees at least $15 million o protect its asset holding from generating returns below the risk-free rate. In fact, Company B’s lan is significantly riskier than Company A’s because of its equity-for-debt swap.

Today, however, such quantification of this risk exposure appears nowhere in the company’s nancial statements even though the technology for assessing the risk of Company B’s plan has een around for four decades, ever since the introduction of the options pricing model. How do nvestors and creditors get information about the self-insurance risk that companies �and states and ountries� take on when their pension plans contain risky assets intended to pay the commitments rom their fixed pension liability structure?

An accountant attempting to calculate the risk of a pension plan’s assets would need to decide hich options model, among many alternatives, to use. Merton describes that an options pricing odel is like an autopilot on an airplane. Both process a great deal of information in a coherent

nd consistent way—perhaps better than human intuition and experience. But pilots understand he model and assumptions being used by the autopilot mechanism and are prepared to override hem based on judgment and experience. Which accounting scholars have acquired the experience nd judgment to know when and which options pricing models are valid and under what circum- tances to override the model? A model that is highly accurate for short periods of time may have igh errors for long-term valuation, while one that is approximately valid for long periods may be uite inaccurate for short time intervals. Merton posed the following challenge to the accounting rofession:

accounting does a good job at valuation but … it is totally inadequate to deal with risk allocation, which … is one of the critical issues today. Accounting must change in a major way to address this in the future. �Merton 1995, 470�

he Accounting Review March 2011 American Accounting Association

H a t

i T a i fi l b t w

S

c s d i c s t m

i l l r m o m I

r c q C

3

4

376 Kaplan

T A

is challenge, however, has gone unaddressed. The entire focus of the FASB and IASB, and lmost all accounting academics, has been to debate the valuation of assets and liabilities but not heir risk.

What would a risk balance sheet look like? I do not know; if I did, I would be writing about t already. Yet, this is exactly the problem of not having scholarship follow contemporary practice. he technology to develop a risk balance sheet has existed for decades. To get started learning bout this issue, I would attempt to learn from the most skilled practitioners I could find, probably n hedge funds, investment banks, and sophisticated diversified corporations whose financial of- cers must assess and manage every single day their risk exposure from mismatched asset and

iability positions. This is a great issue where you could establish a reputation as the world’s “risk alance sheet practice innovator and thought-leader.” If successful, your framework would be aught throughout this century to accounting and business students, and would also be applied idely in practice.

tudying the Role for Credit Default Swaps A credit default swap �CDS� represents another practice innovation that has profound impli-

ations for accounting and auditing practice. Introduced by J.P. Morgan in the mid-1990s, a CDS, imilar to a deep out-of-the-money put option, offers an insurance policy on the default of risky ebt. An increase in a CDS rate implies an increased likelihood of future asset price declines or an ncrease in the volatility of future cash flows. Accounting academics have conducted much ac- ounting research on assessing credit and bankruptcy risk, yet have paid little attention to this new ecurity that provides a pure signal on default likelihood. The CDS signal avoids the complica- ions from bond or credit-specific features—such as sinking funds, collateral, and covenants—that

ake inferring company risk from bond yields a complex calculation. Governments and legislators continue to debate the role of rating agencies, and how to

mprove their performance. The U.S. government just passed a 2,000-page bill to increase regu- ation of financial markets hoping to avoid future systemic risk from the failure of one or more arge financial institutions. Many of the details for its implementation have been left to various egulatory agencies. Who do you think would do a better job of assessing the default risk of a ajor financial institution? A government regulator, who is paid less than a new assistant professor

f accounting, or the aggregation of beliefs among sophisticated professionals betting hundreds of illions of dollars, and more, on understanding and pricing the risk of corporate default?3 �This is, hope, a rhetorical question.�

Accounting professionals can play an important role to improve the pricing of CDSs by their eporting and disclosure of information about all the debt, commitments, and contingencies in- urred by a company �or public sector entity such as a state and country�. Among the interesting uestions, about which we currently lack adequate knowledge on the quality and reliability of DS pricing, are the following:

1. Are CDS prices a valid estimate of default likelihood? Do they predict financial distress better than our current models of bankruptcy and credit risk? These questions must be answered affirmatively before we could consider shifting corporate risk measurement from bond-rating agencies and regulators to CDS securities markets.

2. What information is impounded in a CDS price? What new information causes a CDS price to change?4

I am indebted to Professor Scott Richard, of the Wharton School, for this provocative suggestion. I am encouraged to see an article in The Accounting Review �Callen et al. 2009� begin to address this topic, but this first CDS paper in an accounting journal appeared five years after they started appearing in finance journals, and 15 years after the original innovation itself.

he Accounting Review March 2011 merican Accounting Association

d b p a v t o

m i v m c e t a a d

w a m d W b l v s s h p t a

fi e t

V

w d

T p

Accounting Scholarship that Advances Professional Knowledge and Practice 377

T

3. What kinds of risk models do sophisticated participants in CDS markets use? For you to become the world’s thought-leader on the pricing and information content of credit

efault swaps, I do not recommend using only reduced-form statistical models of the correlations etween CDS prices and various accounting and financial ratios. CDS pricing is an emerging ractice area, and I would attempt to learn, through observation and description, about the models nd processes used by leading CDS-pricing practitioners. You could then move on to classify the arious approaches you encountered and conduct association studies to assess the performance of hese practice-based models against alternative default prediction models, or the human judgment f bond-rating personnel and government regulators.

V. FAIR VALUE MEASUREMENT A third opportunity for innovative and important faculty research and scholarship is the

easurement of fair values. Unlike research on pension plan risk and CDS pricing, many account- ng scholars are already doing correlational studies between fair values and security prices, moti- ated to a considerable degree by the FASB and IASB mandates for increased fair value measure- ent and disclosure. This research reinforces the benefits from epidemiological scholarship on

ontemporary practice issues. Yet all this scholarship takes the “fair values” as given and does not xplore how fair values actually get estimated. The accounting profession is in danger of poten- ially outsourcing this critical measurement, and its validation, to others. Do we have, today, cademic experts on how to do “fair value” analysis for complex assets? Who is our leading cademic “fair value” thought-leader, the go-to person when a company or audit firm has a ifficult question on a complex fair value calculation with a lot at stake?

Let us explore the state of the art today on determining fair values for financial instruments, hich are, by far, the simplest applications for fair value measurements. The traditional accounting

pproach of using contemporary market prices works well for assets that trade continually in thick arkets. For assets that are not actively traded, banks advocate and accounting educators teach the

iscounted cash flow approach, using the interest rate at the time the financial asset was issued. hile current accounting standards require that impairments in these assets get recognized, most

anks argue against recognizing impairments as long as debtors continue to make payments. This eads to nontraded or thinly traded financial assets being carried at historical cost �or terminal alue�. Wachovia, and other banks, resisted fair value reporting of their financial assets by clas- ifying them into their “held-to-maturity” portfolios, a classification that defies economic sub- tance except in a highly restricted case. Wachovia in July 2008, reported $75 billion in share- olders’ equity, even after taking “modest” impairments of more than $10 billion during the revious 12 months in its more than $300 billion loan portfolio �valued at historical cost�. Yet less han three months later, the bank had failed, and its acquirer, Wells Fargo, wrote down Wachovia’s sset loan position by an additional $74 billion.

This incident, and many others at the time, reveals a major shortcoming in the contemporary nancial reporting framework. The deterministic discounted cash flow model is not adequate for stimating the fair values of risky financial assets. And, sadly, the ability to estimate fair values of hinly traded financial assets has existed for decades.

aluing a Risky Financial Debt Instrument Again, I am in debt to my finance colleague Robert Merton, who proposed a simple �some

ould call it “elegant”� model, based on the following equation, to calculate the value of risky ebt �Merton 1974�:

Risk-free debt = Risky debt + insurance policy on the value of the underlying asset.

his equation states that a holder of risky debt can transform the asset into risk-free debt by urchasing an insurance policy that pays off should the debt default �the insurance could be a

he Accounting Review March 2011 American Accounting Association

d t

I e d

g t i

p m i V c

S

378 Kaplan

T A

efault-free CDS, helping to explain the popularity of this financial instrument�. After rearranging erms in the above equation, we obtain:

Risky debt = Risk-free debt

− insurance policy that the debtor ’ s assets exceed the loan amount.

n other words, a bank lending to a consumer, commercial customer, or homeowner holds the quivalent of a risk-free debt position less a short position on an insurance policy that the debtor efaults. The following example makes this insight clearer:

Consider a bank that makes a loan, L, which is secured by the assets of a company or homeowner currently having a value, V. The debtor now owns a put option for the face value of the loan because at the loan’s maturity:

If V � L, the company �homeowner� repays the loan and banker gets back L. If V � L, the company �homeowner� defaults and the bank loses L – V. Figure 2 shows the

raphical representation of the bank’s loss. The downward sloping line is the loss when the erminal value is less than the loan value. The horizontal portion shows that the loss will be zero f the asset value at loan maturity exceeds the loan value.

The hyperbola above this line in Figure 2 shows the value of the bank’s short position in the ut option that is embedded in this loan, for a time period before the loan’s maturity. When V is uch greater than L, the put option is deeply out of the money and there is essentially no decline

n the fair value of the loan. But as the asset value approaches the loan value from the right �when is only slightly greater than L�, the put option increases sharply in value. Any potential pur-

haser of the loan will be estimating the likelihood of further declines in the asset’s value and also

FIGURE 2 Graphical Representation of a Bank’s Potential Loss on a Secured Loan

$$

L/V > 1 L/V = 1

Amount of Loss

Value (V) of Asset L = Loan size

L/V << 1

Price of the “Put” before expiration

Loss on the “Put” option at expirationL

V Low Expected Loss

High Expected Loss

ource: Merton (2009).

he Accounting Review March 2011 merican Accounting Association

t t f i

w l t u r d h m p c m l m i

F

A t c n t e t w s p c

o a t a z w f B e s

F

b Y

Accounting Scholarship that Advances Professional Knowledge and Practice 379

T

he increased volatility of the asset’s price, which occurs as prices decline. In deciding how much o offer to purchase the risky loan, the potential purchaser will subtract the value of the put option rom the terminal loan value �L�, even though the loan may still be current �and therefore not mpaired at all, as argued by the bank�.

The large “bid-ask” spread, between what a rational bidder is prepared to pay for a risky loan ith a loan-to-value ratio of 1, or less, is interpreted �incorrectly� by the bank as a decrease in the

iquidity for its asset, not the actual decline in its underlying value. The bank rejects the markdown o fair value �the bid price� by claiming �1� it intends to hold the loan to maturity, and �2� the nderlying cash flows from the asset have yet to be impaired in any way. But these are not valid easons for failing to fair value a financial asset. When a company leverages its asset position with ebt that is shorter term than the underlying asset, it loses control over whether the asset will be eld to maturity; that right has been transferred to its creditors. They will make their own esti- ates of the asset’s fair value and will likely subtract an additional discount to compensate for the

ossibility �from lack of disclosure� that they are dealing with a “lemons problem.” Should the reditors’ estimates of the bank’s asset values be less than the bank’s current liability position, the arket for the bank’s credit will disappear precipitously. The bank will complain about a sudden

oss in liquidity, but the loss has been in the value of its asset holdings, not in the liquidity of arkets. It is in the bank’s self-interest to understand the fair value of its asset position better than

ts creditors, which many fail to do.

air Values are Often Not the Most Recent Market Price A secured loan, illustrated above, is just one example of a thinly traded financial asset.

ccountants could benefit from learning how to use contemporary market prices of similar assets o infer what price a particular asset would trade at should it trade at the next point of time. In the ase of mortgage-based assets in the second half of 2008, trading may have decreased, but it was ever zero. Nearly 1,000 securities traded each day, and these provided sufficient reference points o value all the asset holdings. A similar situation occurs for U.S.-based mutual funds holding quity assets that trade exclusively in overseas markets. The funds must, at the end of the U.S. rading day, calculate the fair values of these overseas equities for calculating the price at which it ill buy or sell fund shares. The funds use estimates of the correlations between overseas and U.S.

tock prices as well as the information in the closing prices of U.S.-traded equities to estimate the rice at which the overseas equities would have traded in a transaction executed at the U.S. losing time.

As another example, U.S.-based funds of fixed income assets hold individual securities, most f which do not trade continually or even daily. Yet, at the end of each trading day, they must mark ll their holdings to best estimates of their fair values. This calculation requires extensive statis- ical analysis of the correlations between the assets held by the funds and the assets that did trade t or near the end of the day. In effect, the funds use models similar to those used by the website illow.com to estimate home prices based on actual sales of homes in the same neighborhood. The ebsite has an algorithm that adjusts the actual transaction prices of homes in the neighborhood

or differences between the traded homes and the nontraded homes. Financial institutions, such as ear-Stearns, AIG, Merrill Lynch, and Lehman Brothers, that had not developed the capabilities to stimate the fair value of their asset holdings at the end of each trading day ended up being urprised by their insolvencies in 2007/08.

air Value of Compensation Packages Complex fair value calculations also arise in executive compensation contracts, as illustrated

y the “inducement portion” of the compensation plan offered to Carol Bartz to become CEO of ahoo:

he Accounting Review March 2011 American Accounting Association

S

A u C o s g c t

e t e b c fi c a p l

b p t i e

p c s i f m v i p

a

380 Kaplan

T A

Shares Granted Vesting Price % Increase over

Current Market Price

5,000,000 11.73 0% 1,666,667 17.60 50%

833,333 20.53 75% 833,333 23.46 100% 416,667 26.39 125% 416,667 29.33 150% 833,333 35.19 200%

ource: Yahoo Form 10-K for period ending 12/31/08. Note 17: 109.

ccounting scholars have studied compensation plans for decades. Has this research helped us to nderstand the properties of Bartz’s inducement grant? Could we advise a member of the Yahoo! ompensation committee about the parameters of such a grant, such as the number and distribution f price points and the quantity of shares vested at each price point? How many accounting cholars or our students could calculate the fair value of this equity grant? If not, how can we give uidance to our graduates who we hope will become the auditors, CFOs, and board members of ompanies that issue such compensation contracts? Given all our interest in CEO compensation, his seems like it could be a fruitful area for our research and teaching.

I conducted some field research while at the 2010 AAA Annual Meeting to ascertain how we xplain fair value measurements today to accounting students. I walked around the exhibition hall o examine Intermediate Accounting textbooks �a task that enabled me to skip my weight machine xercises in the hotel’s fitness center later that day�. In skimming the fair value chapter in each ook, I saw detailed instructions on how to record asset value changes to fair values as required by urrent standards and regulations, but no coverage of how to estimate the fair values of risky nancial assets. We have, implicitly, delegated to others the task of estimating fair values of omplex securities and compensation contracts. Our students learn only how to enter the debits nd credits, not the knowledge or expertise to estimate fair values. I doubt whether accounting rofessionals who have outsourced such a critical measurement can understand or audit the va- idity of the fair value estimates supplied to them.

I left the exhibition hall rather depressed. I imagined what a comparable situation would have een in physics in 1945, 40 years after the development of quantum mechanics. Would academic hysicists still be teaching only Newtonian mechanics, and not instructing their students that this heory was not even approximately valid in certain important situations? Would biology professors n 1995 be instructing students only about 19th century Linnaean classification and Darwinian volution, and ignoring a half-century of innovations in genetics and molecular biology?

I learned earlier this year that a major bank employs 500 accountants to mark its entire global ortfolio of securities to fair value each day. The chief accounting officer told me that the bank annot hire graduates from U.S. accounting departments for this task. The students do not know ufficient economics, mathematics, and statistics to perform the fair value calculations. This deficit s a direct result of accounting scholars not doing research on fair value measurement and there- ore not being able to teach our students how to perform such calculations. And, unfortunately, easuring the fair value and risk of financial instruments is easy compared to estimating the fair

alue and risk of a company’s real and intangible assets, including the capabilities and loyalty of ts employees, its network of global suppliers, its customer relationships, and its new product ipeline.

VI. IMPLICATIONS FOR ACCOUNTING SCHOLARSHIP To my young accounting scholar, I hope that you can begin to rectify this situation by

ttempting to increase the relevance and impact of your research, education, and teaching. Indi-

he Accounting Review March 2011 merican Accounting Association

v i s i

a a g c n p

I

s a b p e n e

d t m r p

fi n b M m a p c

i

Accounting Scholarship that Advances Professional Knowledge and Practice 381

T

idual faculty scholarship, however, is only one component for closing the gap between academics n professional schools and their audience of current and future practitioners. Academic scholar- hip is a system of interconnected parts. Reform requires changes in all the system’s elements ncluding:

1. Doctoral training �Polzer et al. 2009�. 2. Publication standards applied by journal editors and reviewers. 3. Faculty promotion criteria. 4. Textbook writing and educational curriculum. 5. School and program accreditation criteria. 6. University promotion standards that are sensitive to the different expectations between

scholars in academic versus professional schools. Over the past 40 years, these six elements have converged into a stable equilibrium of

ccounting scholarship characterized by a narrow set of acceptable research methods, all of which re clustered at the top of Roethlisberger’s �1977� Knowledge Enterprise diagram. It will require reat energy to break out of this equilibrium to allow and reward research methods that document, lassify, and create new knowledge for emerging problems and opportunities in practice. That is ot your job as a new Assistant Professor; it is a task that falls on tenured Full Professors in your rofession to take on.

nnovation in Accounting Education In my remaining time, let me address just component #4 in the system of academic scholar-

hip: the implications of my remarks for your teaching and educational activities. First, take dvantage of being in a professional school by choosing a portion of your teaching assignment to e in executive education programs. As you develop expertise in a particular field relevant to the rofessional practice of accounting and finance, you can test your ideas by teaching them to xperienced professionals. These executives are sacrificing much of their time and their compa- y’s money to attend and benefit from your teaching. By sharing your research with them, you get xcellent feedback along two dimensions:

1. Is the problem you are addressing relevant to their experience and practice? 2. Is the expertise you have developed for this problem relevant to their practice? Are you

teaching them something worthwhile about the problem that they can productively apply to their company’s situation?

If the answer to either of these two questions is “no,” then you will certainly hear about their isappointment. But if knowledgeable professionals value learning about your proposed solution o a problem they face, you will have accomplished an important and highly satisfying teaching

ilestone. And you will not have to rely only on large sample surveys to get feedback on the elevance, applicability, and innovation of your proposed solution to an important problem in ractice.

Second, mastering some of the new common body of knowledge for future accounting and nance professionals will require your students to receive extensive training in economics, fi- ance, higher mathematics �including stochastic calculus�, and multivariate statistics. This may be eyond what can be delivered in a core curriculum in undergraduate accounting or a two-year .B.A. program. But certainly, as an innovative accounting scholar, you can develop teaching aterials, and new courses for an elective track in valuing and measuring the risk for complex

ssets, strategies, and transactions. You will contribute to your profession by educating future ractitioners in innovative, contemporary solutions to problems they will encounter during their areers.

Third, try to codify your innovations by developing new classroom teaching materials and nnovative textbooks �or at least textbook chapters� that will allow your insights to be taught by

he Accounting Review March 2011 American Accounting Association

y e

n o p a b

b o s w o

A

B

B

B

C

E F

H H

J

K

—

K

—

K

K

382 Kaplan

T A

our accounting academic colleagues to a much wider audience than your own students. This will nable your contributions to have a wider impact on future accounting practice.

VII. CONCLUSION I hope you can now appreciate the many opportunities for productive scholarship during the

ext four or more decades of your academic career. Risk measurement and management is only ne example of the many interesting and important issues being faced by accounting and finance rofessionals today. I am sure that the next four decades will contain as many new developments nd challenges as the previous four have uncovered. I regret only that I cannot turn my own iological clock back 40� years so I could join with you on the wonderful journey ahead.

While on this journey, work on important problems whose solutions will expand the common ody of knowledge for the practice of accounting, finance, and management. Use research meth- ds that help you understand the problems professionals face and attempt to develop innovative olutions that they can apply. Teach your insights and solutions to your students, and share them idely with your global academic colleagues. Remain excited about the research and teaching pportunities in your chosen academic field. Help to improve the world of professional practice.

REFERENCES

AA Research Impact Task Force. 2009. The impact of academic accounting research on professional practice: An analysis by the AAA Research Impact Task Force. Accounting Horizons 23 �4�: 411–456.

lack, F., and M. Scholes. 1973. The pricing of options and corporate liabilities. The Journal of Political Economy 81 �3�: 637–654.

owen, K., A. Kazakis, A. Muir-Harmony, and B. Lapierre. 2005. The Langer Lab: Commercializing sci- ence. Harvard Business School Case #605-017. Cambridge, MA: Harvard University Press.

owen, K., and R. Gino. 2006. The Whitesides Lab. Harvard Business School Case #606-064. Cambridge, MA: Harvard University Press.

allen, J. L., J. Livnat, and D. Segal. 2009. The impact of earnings on the pricing of credit default swaps. The Accounting Review 84 �5�: 1363–1394.

instein, A. 1920. Relativity: The Special and General Theory. London, U.K.: Methuen & Co Ltd. eltham, G., and J. A. Ohlson. 1995. Valuation and clean surplus accounting for operating and financial

activities. Contemporary Accounting Research 11 �2�: 689–731. enderson, L. J. 1970. On the Social System. Chicago, IL: University of Chicago Press. unton, J. E., and A. Gold. 2010. A field experiment comparing the outcomes of three fraud brainstorming

procedures: Nominal group, round robin, and open discussion. The Accounting Review 85 �May�: 911–935.

ohnson, H. T., and R. S. Kaplan. 1986. Relevance Lost: The Rise and Fall of Management Accounting. Boston, MA: Harvard Business School Press.

achelmeier, S. J. 2009. Annual report and editorial commentary for The Accounting Review. The Account- ing Review 84 �November�: 2047–2075.

—–. 2010. Annual report and editorial commentary for The Accounting Review. The Accounting Review 85 �6�: 2173–2203.

aplan, R. S. 1986. The role for empirical research in management accounting. Accounting, Organizations and Society 11 �4–5�: 429–452.

—–. 1983. Measuring manufacturing performance: A new challenge for managerial accounting research. The Accounting Review 58 �4�: 686–705.

hurana, R. 2007. From Higher Aims to Hired Hands: The Social Transformation of American Business Schools and the Unfulfilled Promise of Management as a Profession. Princeton, NJ: Princeton Uni- versity Press.

rugman, P. 1994. The fall and rise of development economics. In Rethinking the Development Experience, edited by L. Rodwin, and D. Schön, 39–58. Washington, D.C.: The Brookings Institution.

he Accounting Review March 2011 merican Accounting Association

M

M

—

—

—

M

O

P

P

R S

W

Z

Accounting Scholarship that Advances Professional Knowledge and Practice 383

T

cCloskey, D. N. 1998. The rhetoric of significance tests. In The Rhetoric of Economics, 2nd edition, 112–138. Madison, WI: University of Wisconsin Press.

erton, R. C. 1973. Theory of rational option pricing. The Bell Journal of Economics and Management Science 4 �1�: 141–183.

—–. 1974. On the pricing of corporate debt: the risk structure of interest rates. The Journal of Finance 29 �2�: 449–470.

—–. 1995. Financial innovation and the management and regulation of financial institutions. Journal of Banking & Finance 19 �3–4�: 461–481.

—–. 2009. Observations on macro financial risk propagation, structural risks, and regulatory recommen- dations from the crisis. Harvard Business School Deep Dive Seminar Series. �January 23�. Cambridge, MA: Harvard University Press.

iller, D., R. Greenwood, and R. Prakash. 2009. What happened to organization theory? Journal of Man- agement Inquiry 18 �4�: 273–279.

hlson, J. A. 1995. Earnings, book values, and dividends in equity valuation. Contemporary Accounting Research 11 �2�: 661–687.

feffer, J. 2007. A modest proposal: How we might change the process and product of managerial research. Academy of Management Journal 50 �6�: 1334–1345.

olzer, J. T., R. Gulati, R. Khurana, and M. Tushman. 2009. Crossing boundaries to increase relevance in organizational research. Journal of Management Inquiry 18 �4�: 280–286.

oethlisberger, F. J. 1977. The Elusive Phenomena. Cambridge, MA: Harvard University Press. chön, D. A. 1992. The crisis of professional knowledge and the pursuit of an epistemology of practice.

Journal of Interprofessional Care 6 �1�: 49–63. est, J., A. Sagar, and M. Ashiya. 2005. Judah Folkman and the war on cancer. Harvard Business School

Case #604-091. Cambridge, MA: Harvard University Press. ucker, L., and M. Darby. 1996. Star scientists and institutional transformation: Patterns of invention and

innovation in the formation of the biotechnology industry. Proceedings of National Academy of Sci- ences USA 93 �23�: 12709–12716.

he Accounting Review March 2011 American Accounting Association