Journal of Accounting and Public Policy l Assignment

Fair-value accounting: A cautionary
tale from Enron
George J. Benston*
Goizueta Business School, Emory University, Atlanta, GA 30322, United States
Abstract
The FASB’s 2004 Exposure Draft, Fair-Value Measurements, would have companie
determine fair values by reference to market prices on the same assets (level 1), sim
assets (level2) and,where these prices are not available or appropriate,present value
and other internally generated estimated values (level3). Enron extensively used level
three estimates and,in some instances,level2 estimates,for its externaland internal
reporting. A description of it’s use and misuse of fair-value accounting should provide
some insights into the problems that auditors and financial statement users might fa
when companies use level 2 and, more importantly, level 3 fair valuations. Enron firs
used level 3 fair-value accounting for energy contracts, then for trading activities gen
ally and undertakings designated as ‘‘merchant’’ investments. Simultaneously, these
values were used to evaluate and compensate senior employees.Enron’s accountants
(with Andersen’s approval) used accounting devices to report cash flow from operati
rather than financing and to otherwise cover up fair-value overstatements and losses
projects undertaken by managers whose compensation was based on fair values. Ba
on a chronologically ordered analysis of its activities and investments,I believe that
Enron’s use of fair-value accounting is substantially responsible for its demise.
Ó 2006 Elsevier Inc. All rights reserved.
0278-4254/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.jaccpubpol.2006.05.003
* Tel.: [phone]; fax: [phone].
E-mail address: [e-mail]
Journal of Accounting and Public Policy 25 (2006) 465–484
www.elsevier.com/locate/jaccpubpol

Keywords:Fair-value accounting; Enron
1. Introduction
The US and International Financial Accounting Standards Boards (FASB
and IASB) have been moving towards replacing historical-cost with fair-value
accounting. In general, fair values have been limited to financial assets and l
bilities,at least in the financialstatements proper.1 A Proposed Statement of
FinancialAccounting Standards,Fair-Value Measurements (FASB,2005,p.
5), specifies a ‘‘fair-value hierarchy.’’Level 1 bases fair values on ‘‘quoted
prices for identicalassets and liabilities in active reference markets whenever
that information is available.’’If such prices are not available,level2 would
prevail, for which ‘‘quoted prices on similar assets and liabilities in active ma
kets,adjusted as appropriate for differences’’would be used (FASB,2005,p.
6). Level 3 estimates ‘‘require judgment in the selection and application of va
uation techniques and relevant inputs.’’ The exposure draft discusses measu
mentproblems that complicate application of allthree levels.For example,
with respect to levels 1 and 2,how should prices that vary by quantity pur-
chased or sold be applied and, where transactions costs are significant, shou
entry or exit prices be used? As difficult as are these problems,at least many
independent public accountants and auditors have dealt with them extensive
and are aware ofmeasurementand verification pitfalls.However,company
accountants and external auditors have had less experience with the third le
(at least for external reporting), which use estimates based on discounted ca
flows and other valuation techniques produced by company managers rather
than by reference to market prices.
Indeed, there are few situations that have revealed the problems encounte
when companies use third levelestimates for their public financialreports.
Instances in which transaction-based historical-based numbers have been mi
leadingly and/or fraudulently reported abound,such as companies reporting
revenue before it is earned (and sometimes not ever earned), inventories mi
ported and mispriced, and expenditures capitalized rather than expensed. M
ford and Comiskey (2002) and Schilit (2002) provide many illustrations of suc
‘‘schenanigans’’ (as Schilit characterizes them). But they (and to my knowled
few,if any,others) do not describe how fair-value numbers not grounded on
actual market prices have been misused and abused. Enron’s bankruptcy an
the subsequent investigations and public revelations of how their managers u
level3 fair-value estimates for both internaland externalaccounting and the
1 The exceptions include goodwillimpairment and,in Europe,appraisals of other assets under
specified conditions.
466 G.J. Benston / Journal of Accounting and Public Policy 25 (2006) 465–484

effect of those measurements on their operations and performance should pr
vide some usefulinsights into the problems thatauditors are likely to face
should the proposed SFAS Fair-Value Measurements be adopted.
Although Enron’s failure in December 2001 had many causes,2 both imme-
diate (admissions of massive accounting misstatements) and proximate (mor
complicated, as described below), there is strong reason to believe that Enro
early and continuing use of level 3 fair-value accounting played an important
role in its demise.It appears that Enron initially used level3 fair-value esti-
mates (predominantly present value estimates) without any intent to mislead
investors, but rather to motivate and reward managers for the economic ben
efits they achieved for shareholders.Enron first revalued energy contracts,
reflecting an innovation in how these contractswere structured,with the
increase in value reported as current period earnings. Then level 3 revaluatio
were applied to otherassets,particularly whatEnron termed ‘‘merchant’’
investments.Increasingly,as Enron’s operations were not as profitable as its
managers predicted to the stock market, these upward revaluations were us
opportunistically to inflate reported netincome.This tendency was exacer-
bated by Enron’s basing managers’ compensation on the estimated fair-valu
of their merchant investment projects. This gave those managers strong ince
tives to over-invest resources in often costly, poorly devised, and poorly impl
mented projectsthat could garnera high ‘‘fair’’ valuation.Initially, some
contracts and merchant investments may have had value beyond their costs
But, contrary to the way fair-value accounting should be used,reductions in
value rarely were recognized and recorded because they either were ignored
or were assumed to be temporary. Market prices, specified as level 2 estimat
in Fair-Value Measurements(FASB, 2005),were used by Enron to value
restricted stock, although in most instances they were not adjusted to accou
for differences in value between Enron’s holdings and publicly traded stock, a
specified by the FASB.Market prices were also used by Enron’s traders in
models to value their positions. In almost all of these applications, the numbe
used tended to overstate the value of Enron’s assets and reported net incom
As the following largely chronologicaldescription of Enron’s adoption of
level3 fair-value accounting shows,its abuse by Enron’s managers occurred
gradually until it dominated their decisions, reports to the public, and accoun
ing procedures. Although, technically, fair-value accounting under GAAP was
limited to financialassets,Enron’s accountants were able to get around this
restriction and record present-value-estimates of other assets using procedu
that were accepted and possibly designed by itsexternalauditor,Arthur
Andersen.
2 See Partnoy (2002), who blames Enron’s use of derivatives, and Coffee (2002),who points to
inadequate ‘‘gatekeepers,’’ particularly external auditors and attorneys.
G.J. Benston / Journal of Accounting and Public Policy 25 (2006) 465–484467

In the following section, I describe Enron’s initial and then widespread use
and abuse of level 3 fair-value in its accounting for energy and other commo
ity-trading contracts, energy production facilities, ‘‘merchant’’ investments, i
major international projects and energy management contracts, investments
broadband (including particularly egregious accounting for its Braveheart pro
ject with Blockbuster), and derivatives trading.3 This is followed by a descrip-
tion of the incentives from basing managementcompensation on fair-value
estimates. I then show how Enron, by structuring transactions so as to report
cash flows from operations, ‘‘validated’’ the profits it reported. Finally, I con-
sider why Enron’s internalcontrolsystem and its externalauditor,Arthur
Andersen, did not prevent Enron’s using fair-value estimates to produce mis-
leading financial statements.
2. Enron’s adoption and use of fair-value accounting4
Enron’s initial substantial success and later failure was the result of a succ
sion of decisions. Fair-value accounting played an important role in these dec
sions because it affected indicators of success and managerial incentives. Th
led to accounting cover-ups and, I believe, to Enron’s subsequent bankruptcy
presentthese developments essentially in chronologicalorder,which shows
how Enron’s initial‘‘reasonable’’use of fair-value accounting evolved and
eventually dominated its accounting and corrupted its operations and report
ing to shareholders.
2.1. Energy contracts
Enron developed from the merger of several pipeline companies that made
the largest natural gas distribution system in the United States. In 1990, Jeffr
Skilling joined Enron after having been a McKinsey consultant to the com-
pany. He had developed a method of trading natural gas contracts called the
Gas Bank.Enron’s CEO, Kenneth Lay, persuaded him to join the company.
Skilling became chairman and CEO of a new division,Enron Finance,with
the mandate to make the Gas Bank work, for which he would be richly com-
pensated with ‘‘phantom’’equity (wherein he received additionalpay in pro-
portion to increases in the market price of Enron stock). Enron Finance sold
long-term contracts for gas to utilities and manufacturers. Skilling’s innovatio
3 Although Enron used the term ‘‘mark-to-market’’ accounting, it rarely based the valuations on
actualmarket prices. Hereafter, ‘‘fair value’’refers to level3 valuations,those based on present-
value and other estimates that are not taken from market prices.
4 This description is largely derived from and documented in McLean and Elkind (2003) and (to
much less extent) Bryce (2002) and Eichenwald (2005), as well as other sources, as noted.
468 G.J. Benston / Journal of Accounting and Public Policy 25 (2006) 465–484

was to give natural gas producers up-front cash payments, which induced th
to sign long-term supply contracts.He insisted on use of ‘‘mark-to-market’’
(actually ‘‘fair-value,’’as there was no market for the contracts) accounting
to measure his division’s net profit. In 1991 Enron’s board of directors, audit
committee and its external auditor, Arthur Andersen, approved the use of thi
‘‘mark-to-market’’accounting.In January 1992 the SEC approved it for gas
contracts beginning that year. Enron, though, used mark-to-market accountin
for its not-as-yet-filed 1991 statements (withoutobjection by the SEC)and
booked $242 million in earnings. Thereafter, Enron recorded gains (earnings)
when gas contracts were signed, based on its estimates of gas prices project
over many (e.g., 10 and 20) years.
In 1991,Enron created a new division thatmerged Enron Finance with
Enron Gas Marketing (which sold naturalgas to wholesale customers)and
Houston Pipeline to form Enron Capitaland Trade Resources (ECT),all of
which were managed by Skilling.He adopted fair-value accounting for ECT
and he compensated the division’s managers with percentages of internally g
erated estimates of the fair values of contracts they developed. An early (199
example was a 20-year contract to supply naturalgas to the developer of a
large electric generating plant under construction, Sithe Energies. ECT imme
diately recorded the estimated netpresentvalue ofthat contractas current
earnings. During the 1990s, as changes in energy prices indicated that the co
tract was more valuable,additionalgains resulting from revaluations to fair
value were recorded,which allowed Enron to meet its internaland external
quarterly net income projections. By the late 1990s, Sithe owed Enron $1.5 b
lion. However,even though Enron’s internalRisk Assessmentand Control
(RAC) group estimated that Sithe’s only asset (worth just over $400 million)
was inadequate to pay its obligation,the fair value of the contract was not
reduced and,consequently,a loss was not recorded.In fact, the loss was not
recorded until after Enron declared bankruptcy.
2.2. Energy production facilities
Enron International (EI), another major division of Enron, developed and
constructed naturalgas power plants and other projects around the world.
Enron’s developer, John Wing, had previously (in 1987) developed a gas-fired
electricity plant in Texas City, Texas. It was financed almost entirely with hig
yield, high-return (junk) bonds and was very profitable as measured by histor
ical cost accounting.In 1990,Wing completed a dealfor a giant plant in
Teesside (UK) that could produce 4% of the United Kingdom’s entire energy
demands. Enron put up almost no cash and still owned half the plant valued
$1.6 billion in exchange for its role in conceiving and constructing the plant.
the early 1990s, Enron booked more than $100 million in profits from develo
ment and construction fees. Wing received at least $18 million in Enron stock
G.J. Benston / Journal of Accounting and Public Policy 25 (2006) 465–484469

plus several million dollars in salary and bonuses for his work on the project,
establishing performance-based compensation. Such rewards were later agg
sively demanded by and used to reward Enron’s senior managers, but with o
major difference – unlike Wing’s compensation that was based on his project
profitable completion and operation as measured by historicalcost numbers,
their rewards were based on the projected future benefits from (or fair values
of) projects.
To provide gas for the Teesside plant, Enron signed a long-term ‘‘take-or-
pay’’contract in 1993 for North Sea J-Block gas.After gas prices decreased
instead of increasing as Enron had expected, the contract became increasing
costly. Nevertheless, the contract was not marked down to fair value to refle
the losses until 1997. That year, Enron was able to change the contract to on
where the price of gas floated with the market, at which point it had to recor
$675 million pre-tax loss.
2.3. ‘‘Merchant’’ investments
Enron also used fair-value accounting for its ‘‘merchant’’ investments–part
nership interests and stock in untraded or thinly traded companies it started
in which it invested. As was the situation for the energy contracts, the fair va
ues were not based on actual market prices, because no market prices existe
for the merchant investments. Although the SEC and FASB require fair-value
accounting for energy contracts,FAS 115 limits revaluations of securities to
those traded on a recognized exchange and for which there were reliable sha
prices, and valuation increases in non-financial assets are not permitted. Enr
(and possibly other corporations) used the following procedure to avoid these
limitations.Enron incorporated major projects into subsidiaries,the stock of
which it designated as ‘‘merchant’’investments,and declared that it was in
the investment company business,for which the AICPA’s Investment Com-
pany Guide applies. This Guide requires these companies to revalue financial
assets held (presumably)for trading to fair values,even when these values
are not determined from arm’s-length market transactions. In such instances
the values may be determined by discounted expected cash flow models,as
are level3 fair values.5 The models allowed Enron’s managers to manipulate
net income by making ‘‘reasonable’’assumptions thatwould give them the
gains they wanted to record. (Some notable examples are provided below.)
Enron chiefaccounting officer,Rick Causey,used revaluations ofthese
investments to meetthe earnings goals announced by Skilling and Enron’s
5 ‘‘Real’’ investment companies, which often are limited partnerships, tend to value investment
conservatively and values are notchanged untila materialeventoccurs to change the value
(National Venture Capital Association, undated and unpaginated, under valuation), apparently to
limit the amount that would be paid out to investors who take out their investments.
470 G.J. Benston / Journal of Accounting and Public Policy 25 (2006) 465–484

CEO and Chairman of the Board, Kenneth Lay. ‘‘By the end of the decade,’’
McLean and Elkind (2003, p. 127) report, ‘‘some 35 percent of Enron’s assets
were being given mark-to-market treatment.’’ When additional earnings were
required,contracts were revisited and reinterpreted,if increases in their fair
values could be recorded. However, recording of losses was delayed if any po
sibility existed that the investment might turn around.
An example is Mariner Energy,a privately owned Houston oil-and-gas
company that did deepwater exploration in which Enron invested and which
it boughtout for $185 million in 1996.Enron’s accountantsperiodically
marked-up its investment as needed to report increases in earnings until,by
the second quarter 2001,it was on the books for $367.4 million.Analyses in
the second and third quarters of 2001 by Enron’s Risk Assessment and Contr
department (RAC) that valued the investment at between $47 and $196 milli
did not result in accounting revaluations. After Enron’s bankruptcy, Mariner
Energy was written down to $110.5 million.
Enron’s investmentin RhythmsNet Connections(Rhythms)is another
example, except that the valuation was based on an actual market price. Enr
April 1999 investment of $10 million before Rhythms went public in April 199
had an estimated market value of $90 million following its initial public offeri
(IPO). However, Enron could not realize this gain because it had signed a lock
up agreement that precluded sale of the stock until November. Not surprising
Enron could not purchase a reasonably priced hedge.With the approvalof
Enron’s board of directors,a specialpurpose entity (SPE),LJM1, controlled
by Enron’s Chief FinancialOfficer,Andrew Fastow,was created to provide
the hedge through an SPE itcreated,LJM Swap Sub. 6 LJM1 was almost
entirely funded with Enron’s own stock purchased with a promissory note at
a 39% discount (because it was restricted). LJM Swap Sub was funded with ha
the stock and a promissory note from LJM1. In effect, Enron was writing the
option on itself, for a substantial fee paid to Fastow. It was not really hedging
against a possible economic loss on the Rhythms stock, but against having to
recognize the loss in its accounts. Thus, Enron’s accounting ‘‘shenanigans’’ w
not limited to booking income from increasing estimated fair values.
2.4. Dabhol, other Enron International projects, and Azurix
In 1996 Rebecca Mark became CEO of Enron International, having previ-
ously been head of Enron Development.She developed projects around the
world at a frenetic pace.She and her managers were given bonuses for each
project they developed of about 9% of the present value of its expected net c
6 The transaction is very complicated.See Benston and Hartgraves (2002,pp. 109–110) for a
much more complete description.
G.J. Benston / Journal of Accounting and Public Policy 25 (2006) 465–484471

flows, one half paid at the financial close and the other half when the project
became operational. The costs of projects that did not come on line but were
not officially declared to be dead were recorded as assets, if the amounts we
under $200 million. Mark’s largest project, begun in 1992 when she managed
Enron Development,was a giantelectricity power plantin Dabhol, India.
To be economically viable, the government of the Indian state of Maharashtr
would have to purchase a fixed amount of electricity at a high price, despite
fact that it was unable to collect for electricity sold at lower prices. The proje
was severely criticized in India, and the contract was renegotiated several tim
Eventually,Enron invested and lost about $900 million in the project,which
now stands idle.Nevertheless,Mark and her team received $20 million in
bonuses for the project, based on their estimates of its present value.
Relatively few of Enron International projects actually became operational
and few were profitable according to traditional accounting standards after t
became operational. Some pre-bankruptcy evidence is available as a result o
dispute between Skilling and Mark. In 1998, Skilling had an in-house accoun-
tant value Enron International’sprojects.He calculated thatthe division
returned only a 2% return on equity, excluding Enron’s substantial contingen
liability for project debts it guaranteed. Mark’s accountant, though, estimated
that her division returned an average of 12% on equity. After Enron declared
bankruptcy, few of the division’s projects were found to have any value.
In May 1998 Skilling forced Mark out of her post as CEO of Enron Inter-
national. (In December 1996, Lay had appointed Skilling, rather than Mark, a
Enron’s President and Chief Operating Officer.) Perhaps as a consolation prize
in July 1998 the board of directors (with Skilling’s blessing) allowed Mark to
establish a new Enron subsidiary,Azurix, which would develop water-supply
projects around the world.The business began with a $2.4 billion purchase
of a British water utility, Wessex Water Services, for a 28% premium.7 The pur-
chase was largely financed with debt sold by an off-balance-sheet partnershi
Marlin, which itself was financed with debt that Enron guaranteed.Enron’s
obligation on that debt was not reported on its financial statements. Most of
the balance came from public sales of Azurix shares for $700 million. Azurix
then successfully bid for a Buenos Aires, Argentina water utility that was priv
atized,paying three times more than the nexthighestbidder.Azurix later
learned that the dealdid not include the utility’s headquarters and records,
making it difficult (often impossible) to collect past-due and present account
balances.8 By year-end 2000, $402 million had to be written off on the project.
Other disastrous projects were undertaken,Azurix stock declined to $3.50 a
share and Enron had to repurchase the publicly held shares for $8.375. Mark
7 Eichenwald (2005, p. 191) reports the purchase as costing $2.25 billion.
8 Eichenwald (2005, p. 231) reports that Azurix paid twice the next bid and that computers and
records were trashed by the former employees.
472 G.J. Benston / Journal of Accounting and Public Policy 25 (2006) 465–484

resigned, receiving the balance of her $710,000 Azurix contract, based on th
projected (mark to fair value) benefits to Enron of the enterprise.
2.5. Energy management contracts
In December 2000,after Skilling became president of Enron,he created a
separate business,Enron Energy Services (EES),with Lou Pai as its CEO.
EES expected to sellpower to retailcustomers,based on assumptions that
the market would be deregulated and that the existing utilities could be unde
sold. Enron sold 7% of EES to institutional investors for $130 million. Based
on this sale (which might have qualified as a level 2 estimate), Enron valued
company at $1.9 billion, which allowed it to record a $61 million profit. How-
ever, EES’s efforts were unsuccessful, in part because retail energy was gene
ally not deregulated.Losses on the retail operationswere not reported
separately, but were combined with the wholesale operations.
Pai then concentrated on selling contracts to companies and institutions to
provide them with energy over long periods with guaranteed savings over th
present costs. Customers often were given up-front cash payments in advanc
the promised savings.These contracts were accounted for on a mark-to-fair-
value basis as of the date the contracts were signed. Sales personnel and ma
ers (especially Pai) were paid bonuses based on those values. Not surprisingl
this compensation scheme generated a lot of bad contracts. A particularly co
(to Enron) contract was signed in February 2001 with Eli Lilly to make improv
ments in its energy supply and use over 15 years. Discounting these amount
8.25–8.50%, Enron valued the contract at $1.3 billion and recorded a $38 mi
gain.9 Within two years, this contract was considered to be worthless.
In 2001, after Pai left EES and Enron, a long-time in-house Enron accoun-
tant, Wanda Curry, was asked to evaluate the EES contracts. Her group exam
ined 13 (of 90) contracts that comprised 80% of the business. Each of them h
been recorded as profitable.Nevertheless,Curry found that the 13 contracts
had a totalnegative value of at least $500 million.For example,a dealfor
which the company had booked $20 million in profits, actually was $70 millio
under water. Although, accordingto mark-to-fair-valueaccountingthe
decrease in value documented by Curry should have been recorded,no such
entry was made and Curry was reassigned.
2.6. Enron broadband services
Enron Broadband Services (EBS)was another major portion ofEnron’s
business. Skilling established it in April 1999 to develop a fiber-optic network
9 See Batson (2003a, pp. 32–33) for details.
G.J. Benston / Journal of Accounting and Public Policy 25 (2006) 465–484473

and trade capacity in its and other firms’ networks. Skilling announced the ne
venture to stock analysts on January 20,2000,together with Scott McNealy,
CEO of Sun Microsystems,who said that Enron would purchase 18,000 of
Sun’s best servers for use in its network. By day end, Enron’s stock increased
by 26%. Enron, though, did not then or ever have software that could provide
bandwidth on demand by and for alternative networks. Rather, Enron’s busi-
ness involved swapping the right to use surplus (dark) fiber on its own netwo
for the right to use surplus on other networks.Overall,Enron invested more
than $1 billion on broadband and reported revenue of $408 million in 2000,
much ofit from sales to Fastow-controlled SPEs.For example,in the first
quarter 2000, EBS recorded a mark-to-fair-value-determined gain of $58 mil-
lion from revaluing and then swapping dark fiber,which was designated a
‘‘sale.’’In the second quarter 2000 EBS revalued and ‘‘sold’’that assetto
LJM2, a SPE controlled by Fastow,and recorded another $53 million pre-
tax gain.Based on mark-to-fair-value accounting,EBS booked a $110.9 mil-
lion profit in the fourth quarter 2000 and first quarter 2001.
In the third quarter 2000 EBS recorded a $150 million fair-value gain on its
$15 million investmentin a tech start-up (AviciSystems)that wentpublic,
using the public IPO price as the basis for the transaction even though Enron
stock could not be sold for 180 days. Enron ‘‘locked in’’ the gain with a hedge
provided by another SPE (Talon),even though Talon would not have been
able to meet its obligation if the stock price declined. Before year-end, the st
price declined by 90%. Talon and other similar SPEs (collectively called ‘‘Pro-
ject Raptor’’) could not cover this loss and other losses amounting to $500 m
lion. Nevertheless,the losses were notrecorded,based on Enron’s (invalid)
assertion thatthe SPEs’obligations could be cross-collateralized with other
SPEs that were claimed (incorrectly) to have sufficient assets. Those assets w
Enron shares and rights to shares obtained from Enron for which the Raptors
had not paid. Consequently, for Enron the assets did not exist, because if the
SPEs had to pay Enron their obligations for the hedges by selling the shares,
they would be unable to pay their other debt to Enron.Andersen’s partner-
in-charge, David Duncan, agreed to this procedure despite an objection from
Carl Bass,a member of the firm’s ProfessionalStandards Group who previ-
ously was on the audit team. At Enron’s request, Bass was excluded from com
menting on issues related to Enron.Andersen was paid $1.3 million for its
Raptor-related work. When the Raptors were terminated in 2001 a $710 mil-
lion pre-tax loss was booked.
2.7. Braveheart partnership with Blockbuster
In the fourth quarter 2000 EBS announced a 20-year project (Braveheart)
with Blockbuster to broadcast movies on demand to television viewers. How-
ever, Enron did not have the technology to deliver the movies and Blockbust
474 G.J. Benston / Journal of Accounting and Public Policy 25 (2006) 465–484

did not have the rights to the movies to be broadcast.Nevertheless,as of
December31, 2000,Enron assigned a fair value of $125 million to its
Braveheart investment and a profit of $53 million from increasing the invest-
ment to its fair value,even though no sales had been made.Enron recorded
additionalrevenue of$53 million from the venture in the firstquarterof
2001,although Blockbuster did not record any income from the venture and
dissolved the partnership in March 2001.In October 2001 Enron had to
announce publicly that it reversed the $110.9 million in profit it had earlier
claimed,which contributedto its loss of public trust and subsequent
bankruptcy.
How could Enron have so massively misestimated the fair value of its Brav
heart investment, and how could Andersen have allowed Enron to report thes
values and their increases as profits? Indeed, the Examiner in Bankruptcy (Ba
son, 2003a, pp. 30–31) finds that Andersen prepared the appraisal of the pro
ject’svalue.Andersen assumed thefollowing:(1) the businesswould be
established in 10 major metro areas within 12 months;(2) eightnew areas
would be added per year until 2010 and these would each grow at 1% a year
(3) digitalsubscriber lines (DSLs)would be used by 5% of the households,
increasing to 32% by 2010,and these would increase in speed sufficientto
accept the broadcasts;and (4) Braveheart would garner 50% of this market.
After determining (somehow) a net cash flow from each of these households
and discounting by 31–34%, the project was assigned a fair value. I suggest t
this calculation illustrates an essentialweakness of level3 fair-value calcula-
tions that necessarily are not grounded on actualmarket transactions.How
can one determine whether or not such assumptions about a ‘‘first-time’’ pro
ject are ‘‘reasonable’’?
2.8. Derivatives trading
Enron’s (derivatives)trading activities expanded beyond naturalgas and
power contracts to contracts in metals, paper, credit derivatives, and commo
ities.Much of this trading was done over an Internetsystem itdeveloped,
Enron On Line (EOL), which enabled Enron to dominate severalmarkets.
Enron often established the prices on those markets, prices that were used to
mark its trades to fair values. Thus, Enron’s traders could establish the prices
at which positions would be valued and, as described below, their compensa-
tion was determined.10
10 As the dominantplayer, Enron took large positions,which made maintaining market
participants’perception that its credit standing was solid absolutely necessary,which is a major
reason that Enron’s accountants went to great lengths to hide its debt obligations, as documente
by Batson (2003a, Appendix Q).
G.J. Benston / Journal of Accounting and Public Policy 25 (2006) 465–484475

Trading may have been the only really profitable portion of Enron’s busi-
ness.Partnoy (2002) analyzed Enron’s 1998,1999,and 2000 financialstate-
ments.He determined that‘‘other revenue’’meant income (loss) from
derivatives trading, and produced the following table:
Enron Corp. and Subsidiaries2000 ConsolidatedIncome Statement
(in $millions)
2000 1999 1998
Non-derivatives revenues 93,557 34,774 27,215
Non-derivatives expenses 94,517 34,761 26,381
Non-derivatives gross margin (960) 13 834
Other revenue (income from derivatives) 7232 5338 4045
Other expenses (4319) (4549) (3501)
Operating income 1953 802 1378
This table shows that the other aspects of Enron’s business, described earlier
were operated at a net loss or smallgain,even considering that the reported
numbers were inflated by mark-to-fair-value accounting and other accountin
procedures.
However, Partnoy (2002) found that the reported gains from trading deriv-
atives were not reported accurately.Note, though,that these reported gains
and losses were based on the traders’estimated present(fair) values ofthe
derivatives contracts, which often covered many years. Most of these contrac
were not actively traded, or traded at all. The traders, therefore, valued them
with models and estimates of forward price curves, both of which could be ea
ily manipulated. As Partnoy (2002, p. 327) puts it: ‘‘because Enron’s natural-
gas traders were compensated based on their profits, traders had an incentiv
to hide losses by mismarking forward curves.. . .In some instances,a trader
would simply manually input a forward curve that was different from the mar
ket. . . .For more complex trades, a trader would tweak the assumptions in th
computer model used to value the trades, in order to make them appear mor
valuable.’’ Traders also understated their profits or deferred reporting the pro
its with ‘‘prudency reserves.’’This practice allowed them to shift income to
future periods when they had already attained their maximum bonuses and
to offset losses.Partnoy (2002,p. 327) concludes: ‘‘The extent of mismarking
at Enron remains unclear, although several traders said the inaccuracies wer
more than a billion dollars.’’ Thus, the extent to which Enron’s traders actual
benefited Enron’s shareholders is unclear. Indeed, Bryce (2002, p. 336) repor
that, in 2001, when Enron was negotiating a merger with Dynergy, whose pe
sonnel examined Enron’s operations, ‘‘it became increasingly obvious to them
[Dynergy] that despite the enormous volumes that were being generated on
[EOL] web site, Enron actually was losing money on its trading business.’’
476 G.J. Benston / Journal of Accounting and Public Policy 25 (2006) 465–484

3. Management compensation, expenses, and fair-value accounting
Enron promised investors that its earnings would grow by 15% a year,a
goal that Enron employeescould meetwith mark-to-fair-value valuations
and revaluations of their projects. In addition, in granting options to its senio
executives, Enron specified that a third would vest each year if Enron’s earn-
ings grew by at least 15% (McLean and Elkind, 2003, pp. 92–93). Senior exec
utives additionally were motivated to inflate the values of projects with bonu
and stock options based on a percentage of the mark-to-fair-values of the pro
jects and deals they developed,as determined by their present-value calcula-
tions. Furthermore,Enron used a ‘‘rank and yank’’system ofevaluating
employees, whereby those who did not meet their targets had reason to beli
they would be dismissed.11
Enron’s senior managers exercised little controlover costs,in large part
because their compensation was based substantially on estimates of the pre
value of the projects they generated, rather than on the profitability of the pr
jects as substantiated by actual performance measured with historical costs
actualmarket prices,with the cost of developing the projects matched to the
revenue generated therefrom. Enron’s managers and traders were given a pe
centage (generally 9%) of the estimated present value of their deals and trad
which imparted a strong incentive for them to expend whatever resources w
necessary to develop and close a deal or trade. The deal makers were paid s
stantialamounts for projects when the financing was complete,even before
construction began. But, as McLean and Elkind (2003, p. 76) put it: ‘‘no one
felt responsible for managing the projects once they were up and running.’’
McLean and Elkind offer many examples of uncontrolled expenditures. They
report (McLean and Elkind, 2003, p. 119):
people began spending as if every day were Christmas. Expenses soared
for items large and small. . . .There was no requirement to use a particu
vendor; if you didn’t want to wait for something, you could just pick up a
phone and order it yourself. Anyone with a half-baked idea to launch a
business in Europe could hop a plane and fly to London.Hundreds of
deal makers made a habit of flying first class and staying in deluxe hote
no one seemed to care. . . .The corporate administrative types gave up
ing to keep a lid on things.
The expenditures were not limited to the deal makers, but extended throu
out Enron.With respect to trading,McLean and Elkind (2003,p. 226) say:
‘‘Overhead was obscene;one executive estimated thatthe North American
trading operation alone spent between $650 and $700 million a year just in
11 See Bryce (2002, pp. 126–131) for a description.
G.J. Benston / Journal of Accounting and Public Policy 25 (2006) 465–484477

overhead.Expense accounts were over the top,but nobody dared rein them
in.’’ But the largestexpense was compensation.McLean and Elkind (2003,
p. 241) citing a study by the Joint Committee on Taxation, report:
in 1998,Enron’s 200 most highly compensated employees took home a
total of $193 million in salaries,bonuses,and various forms of stock.
In 1999, that leaped to $402 million; in 2000, they took home $1.4 billio
. . .[In that year] each of the top 200 employees made over $1 million; 2
executives made over $10 million.In 2001,the year Enron went bank-
rupt, at least 15 employees made over $10 million.
4. Cash flow and fair-value-determined net income
Although mark-to-fair-value accounting allowed Enron to record substan-
tial profits,it did not provide cash flow.Enron had to dealwith analysts
who were suspicious of accounting net income and looked to cash flow as a
superior measure ofperformance.(‘‘Net profit is what accountants wantit
to be,cash is real.’’) Enron attempted to bring these alternative performance
measures into balance,as wellas obtain cash for its operations and projects
without having to sell stock or report debt, primarily with four schemes: pre-
pays, ‘‘sales’’ to SPEs, share trusts, and investment subsidiaries.12These repre-
sented ‘‘real’’ rather than only accounting manipulations, as they required th
development of legal structures for which Enron paid very substantial fees to
investment and commercial bankers, lawyers, and accountants.
Prepays are arrangements that allowed Enron to record what actually were
loans as cash inflows from operations. Enron agreed to deliver oil or gas to an
offshore entity established by a bank (Citigroup or Chase-Morgan).The off-
shore entity (Delta or Mahonia) paid Enron in advance with funds obtained
from the bank for the same amountof oil or gas,but for future delivery.
The bank, in turn, agreed to deliver the same amount of the oil or gas to Enro
in exchange for a fixed price plus an additional amount equal to what interes
on a loan would be. The bank did not actually take a price risk, as an Enron
subsidiary wrote the bank a guarantee. The bank’s only risk was similar to th
risk of a loan,that Enron could not meet its obligation (which,when Enron
declared bankruptcy,turned out to be the case).At the due date,the bank
‘‘delivered’’ the oil or gas to Enron and received payment, Enron ‘‘delivered’’
the oilor gas to the offshore entity and the entity ‘‘delivered’’the oilor gas
to the bank. Thus, Enron really justborrowed fundsfrom the bank,but
recorded the cash inflow as coming from operations rather than from financ-
12 The schemes were complex. See Batson (2003a, pp. 58–66 and Appendix E) for descriptions.
478 G.J. Benston / Journal of Accounting and Public Policy 25 (2006) 465–484

ing. The debt was recorded as ‘‘price risk liability,’’ which to the unaware fina
cial-statement reader appeared to offset or explain some marked-to-fair-valu
‘‘price risk’’ assets and the related ‘‘earnings from price-risk activities.’’ Bats
(2003a,Appendix Q) calculatesthat in its year 2000 financialstatements,
Enron overstated cash flow from operations and understated cash flow from
financing (and debt) by $1.53 billion.Indeed,32% of Enron’s reported cash
flow from operations in 2000 and almost all in 1999 were due to prepays (Ba
son, 2003a, Appendix E, p. 6). Over the years 1997–2001, when prepays wer
used, Batson estimates that Enron overstated its reported cash flow from ope
ations by at least $8.6 billion.
‘‘Sales’’ to SPEs allowed Enron to ‘‘validate’’ profits derived from mark-to-
fair-value increases in its subsidiaries’ stock (and the assets represented by t
stock).The proceduresused are complicated (see Benston and Hartgraves
(2002), Batson (2002, 2003a, Appendix M, pp. 1–23) for descriptions of these
‘‘FAS 140’’ transactions).13 One example, in simplified form, provides a good
illustration. Enron created a subsidiary to hold a major asset it was developin
receiving in return Class A voting and Class B non-voting stock that was enti-
tled to almostall of the subsidiary’s economic interests.At the same time
Enron created an SPE in which it held no equity interest. The SPE was finance
with funds borrowed from a bank, a subsidiary of which held the equity, whic
was equal to 3% of the assets (the other 97% was the bank loan). Enron then
‘‘sold’’ its subsidiary’s Class B common stock to the SPE in exchange for all it
assets(cash),but kept the ClassA voting-rightsstock.If Enron had not
already written up the stock (a financial asset held as a ‘‘merchant’’ investm
to fair value, thereby recording gains as current income, it now recorded a ga
from the ‘‘sale’’of the Class B stock to the SPEs.At the same time Enron
entered into a ‘‘totalreturn swap,’’wherein it assumed the SPE’s obligation
to pay the bank debt for all the cash flow from the Class B stock less amount
necessary to repay the 3-percent-equity holder’s investmentplus a specified
(usually 15% per annum) return. The swap was accounted for as a derivative
and, pursuantto FAS 133, marked to fair-value,which enabled Enron to
record the change in the present value of expected cash flows from the unde
lying asset as additional current profit (or loss).
Thus, Enron continued to control and operate the asset, recorded as a gain
(and,later,a loss) its calculated increase in the present value of expected net
cash flows, and did not report as a liability its obligation to repay the bank de
incurred by the SPE. This procedure allowed Enron to keep $1.4 billion of deb
off its December31, 2000 balance sheetand reportover $541 million of
13 Enron’s accounting for SPEs,as such,was notthe problem,as explained by Benston and
Hartgraves (2002). Rather, Enron failed to follow FAS 5 and report its contingent liability for the
SPEs’ debt and reported ‘‘sales’’ to the SPEs that were not valid, as explained and concluded by
Batson (2002).
G.J. Benston / Journal of Accounting and Public Policy 25 (2006) 465–484479

additional net income before taxes ($352 million after taxes) on its 2000 inco
statement (Batson, 2003a, Appendix Q, p. 1) by recording the transaction as
sale and the cash received as cash flow from operations rather than cash flow
from financing (loans).
Share trusts (Marlin and Whitewing) allowed Enron to refinance its pur-
chase of water systems for Azurix without reporting the debt on its financial
statements (Marlin),and to ‘‘sell’’assets to SPEs through Whitewing.These
share trusts obtained the necessary funds from banks and other creditors tha
Enron indirectly guaranteed.The trusts were not consolidated with Enron’s
financial statements as a result of Andersen’s accepting Enron’s contention t
it did not control the trusts because outside parties had the right to appoint h
the directors (a right that was never exercised, because the outside parties r
on Enron for repayment of their investments).14 Batson (2003a, Appendix Q)
calculates that this scheme allowed Enron in its year-2000 financial statemen
to overstate cash flow from operations by $0.42 billion and understate cash
flow from financing (and debt) by $1.67 billion.
Investment subsidiaries (also called minority-interest transactions) also we
used to provide ‘‘cash flow from operations.’’ In simplified form, this is how it
worked.(See Batson,2003a,Appendix I particularly Annex 3 for detailed
descriptions.) In 1999, for example, Enron created an SPE with outside equity
of $15 million, which borrowed $485 million from a bank;the equity usually
was provided by the bank’s subsidiary.Because Enron had no equity in the
SPE and the SPE had equity held by independent investors equalto 3% of
its assets, it was not consolidated with Enron. The SPE purchased $500 millio
in Treasury notes and used them to purchase a minority interest in an Enron-
controlled subsidiary whose businessincluded buying and selling Treasury
notes. Shortly thereafter, in December, the subsidiary sold the notes, recordi
a cash inflow from operations. When the subsidiary was consolidated, Enron’
financialstatements reported a minority equity interest (rather than debt) of
$500 million and cash flow from operations (rather than cash flow from finan
ing) of $500 million.
5. Auditors’ validation of Enron’s fair-value accounting
It is or should be obvious that managers’ whose career success and compe
sation are based on and derived from estimated fair-value amounts have stro
incentives to inflate those amounts, either because of a tendency towards ov
optimism,opportunism,or both.This bias could be constrained by internal
14 Batson’s (2003a, Section IV) analysis shows that Enron actually had more than a 50% equity
interest of Whitewing through subsidiaries that it owned or controlled.
480 G.J. Benston / Journal of Accounting and Public Policy 25 (2006) 465–484

controls and reviews and at least should be revealed and criticized by interna
or external auditors. In fact, at Enron it was not.
Skilling established and publicized a Risk Assessment and Control depart-
ment(RAC) that was charged with reviewing and,presumably,changing,
approving, or disapproving the numbers submitted by managers for deals th
were promoting and on which they were compensated.Proposals involving
more than $500,000 had to be supported by a deal-approval sheet (known a
DASH) that included an approval or disapproval by RAC. McLean and Elk-
ind (2003,p. 115) report that Skilling ‘‘made RAC a centerpiece of manage-
ment presentationsto Wall Street analysts,investors,and credit-rating
agencies.’’It had a $30 million budgetand a staff of150 professionals.It
was headed by Rick Buy, who McLean and Elkind describe (McLean and Elk-
ind, 2003,p. 116) as ‘‘uncomfortable with confrontation.When his analysts
raised issues with a deal, Buy would dutifully take them up the chain of com-
mand.But in a head-to-head with the company’s senior traders and origina-
tors, it was no contest,as those on both sidesof the table recognized.’’
Another factor that apparently blunted Buy’s unwillingness to allow his group
to challenge dealmakers’‘‘absurdly optimistic assumptions for the complex
models thatspatout the likelihood ofvarious outcomes for a transaction’’
(McLean and Elkind, 2003, p. 117) was fear that he would lose his compensa
tion, including stock options thathad not yet vested.McLean and Elkind
(2003, p. 118) report that he was paid $400,000 in 1999 and sold Enron shar
for $4.3 million in 2001.Other RAC employees also were reluctant to resist
pressures to approve deals, because this could hurt their careers within Enro
as well as affect their compensation. Bryce (2002, p. 131) reports that ‘‘[p]eo
who questioned deals ‘would get attacked by the business units because the
weren’t as cooperative on a deal as the developers wanted,’’’ which would ha
strong negative effect on their evaluation by Enron’s Personal Review Comm
tee.In addition,Bryce (2002,p. 131) quotes an Enron employee as saying:
‘‘You’d have a hard time finding another rotation if you were too hard on cer-
tain deals.’’
Internalauditors mighthave questioned the validity offair-value assess-
ments.However,as noted earlier,when in-house accountantWanda Curry
reviewed and questioned the validity ofpresumed profitability ofEES con-
tracts, she was reassigned and no changes were made. I am not aware of an
other form of or results from internal audits.
Considering Enron’scompensation structure,Enron’s externalauditor,
Arthur Andersen, had reason to be skeptical of management’s fair-values est
mates,as is required by generally accepted auditing standards (GAAS) stan-
dards of fieldwork 2 and 3. In particular, SAS 57 (‘‘Performanceand
Reporting Guidelines Related to Fair Value Disclosures,’’February 1993),
AU.9342.01-10, states that ‘‘the auditor should collect sufficient competent e
dential matter to reasonably assure that (1) valuation methods are acceptab
G.J. Benston / Journal of Accounting and Public Policy 25 (2006) 465–484481

and (2) estimation methods and significant assumptions are disclosed.’’ Othe
than Batson’s (2003a, pp. 30–31) description of Andersen’s appraisal of the f
value of the ill-fated Braveheart project, I am not aware of a publicly availabl
analysis(or privately conducted examination)of the proceduresAndersen
employed to audit Enron’s fair-value estimates.
It seems clear that Andersen had strong monetary incentives to keep its cl
ent, Enron, happy. Indeed, Zeff (2003) and Wyatt (2004), among many other
ascribe Andersen’s bad behavior to the corrosive influence of consulting fees
and consulting partners.15 The evidence (documented in Benston,2004,pp.
309–311), though, does not support their conclusion with respect to consultin
fees. Nevertheless, prescriptions against CPA firms providing many consultin
services to audit clients have been instituted in the Sarbanes-Oxley Act of 20
although consulting on tax matters, for which CPA firms necessarily adopt a
pro-management advocacy position,is still permitted.Furthermore,the fees
paid by companies for audit services and the importance of those fees to the
individualpartner in charge of the audit,often are sufficiently substantialto
give the auditor reason to be nice to clients.Thus, if current and prospective
monetary rewards were the force driving Andersen to compromise its integri
there stillshould be concern for future Enrons,particularly considering the
inherent difficulty of auditors challenging managers’ fair-value estimates.16
6. Conclusions
The Enron experience should give the FASB, IASB, and others who would
permit (indeed,mandate) level3 fair-value accounting,wherein the numbers
reported are not well grounded in relevant market prices, reason to be cautio
Once Enron was permitted to use fair values for energy contracts, it extende
revaluationsto a wide and increasing rangeof assets,both for external
15 Wyatt (2004)also blames consulting for having introduced into CPA firms a disregard or
insufficient regard for ‘‘accounting professionalism,’’ since consultants do not have the benefits o
auditing coursesthat ‘‘focused on professionalresponsibilitiesand the importance ofethical
behavior.’’However,it should be noted that Andersen not only was controlled by CPAs,rather
than by consultants, but took a very strong public stand on the importance of professional ethics
including preparing cases and training materials.
16 Andersen’s organizational structure and culture also appears to have played an important role
Andersen vested power in the partner-in-charge of an audit, discouraged more junior auditors fro
questioning decisions by their seniors, and emphasized increasing revenues to the firm, as descr
by Toffler (2003). Thus, although the firm would have (and did) pay a heavy price for compromis
and other derelictions of individual senior auditors, those individuals had strong incentives to do
what was necessary to keep from losing the account. See Benston (2003) for data on past failure
the SEC and other authorities to discipline individual independent public accountants who attest
to seriously misleading and fraudulent financial statements.
482 G.J. Benston / Journal of Accounting and Public Policy 25 (2006) 465–484

reporting and internalpersonnelevaluationsand compensation.As docu-
mented above,the result was overstatement of revenue and net income,and
structuring transactions to present cash flows from operations rather than fro
financing.Basing compensation on fair values gave employees strong incen-
tives to develop and overvalue projects,resulting in high operating expenses
and rarely successfulprojects.The lossesincurred gave rise to additional
accounting subterfuges, until the entire enterprise collapsed.
Enron’s internalcontrolsystem was incapable of controlling the misstate-
ments.Enron’s auditors,Arthur Andersen,did not appear to have criticized
the fair-value estimates,either to Enron’s accountants or to Enron’s Board
of Directors (US Senate, 2002, pp. 14–24). Nor did the Board’s Audit Commit-
tee ask Andersen’s auditors to review and comment on the fair-value calcula
tions, despite their having been told by the auditors that this was a ‘‘high risk
accounting procedure (Batson, 2003b, Appendix B, Section IV-I).
Of course, over-optimistic, opportunistic, and dishonest managers have mi
used historical-cost-based accounting to overstate revenueand assetsand
understate liabilities and expenses. Fair-value numbers derived from compan
created presentvalue models and other necessarily notreadily verified esti-
mates provide such people with additional opportunities to misinform and mi
lead investors and other users offinancialstatements.The Enron example
provides some evidence that this concern may not be misplaced or overstate
References
Batson,N., 2002.First interim report of NealBatson,court-appointed examiner.United States
Bankruptcy Court,Southern District of New York.In: re: Chapter 11 Enron Corp.,et al.,
Debtors, Case No. 01-16034 (AJG), Jointly Administered, September 22.
Batson, N., 2003a. Second interim report of Neal Batson, court-appointed examiner. United State
Bankruptcy Court,Southern District of New York.In: re: Chapter 11 Enron Corp.,et al.,
Debtors, Case No. 01-16034 (AJG), Jointly Administered, January 21.
Batson, N., 2003b.Final report of Neal Batson, court-appointed examiner,United States
Bankruptcy Court,Southern District of New York.In: re: Chapter 11 Enron Corp.,et al.,
Debtors, Case No. 01-16034 (AJG), Jointly Administered, November 21.
Benston, G.J., 2003. The regulation of accountants and public accounting before and after Enron.
Emory Law Journal 52 (Summer), 1325–1351.
Benston,G.J., 2004.The role and limitations of financialaccounting and auditing for financial
marketdiscipline.In: Curt Hunter, W. (Ed.), Market Discipline:The EvidenceAcross
Countries and Industries. MIT Press, Cambridge, MA, pp. 303–321.
Benston,G.J., Hartgraves,A.L., 2002.Enron: What happened and what we can learn from it.
Journal of Accounting and Public Policy 21, 105–127.
Bryce, R., 2002. Pipe Dreams: Greed, Ego, and the Death of Enron. Public Affairs (Perseus Books
Group), New York.
Coffee Jr.,J.C., 2002.Understanding Enron:‘It’s about the Gatekeepers,Stupid’.Business Law
Journal 57, 1403–1410.
Eichenwald, K., 2005. Conspiracy of Fools: A True Story. Broadway Books, New York.
G.J. Benston / Journal of Accounting and Public Policy 25 (2006) 465–484483

Financial Accounting Standards Board, 2005. Project Updates, Fair Value Measurements. May 2.
Available from: <http://www.fasb.org/project/fv_measurement.shtml>.
McLean, B., Elkind, P., 2003. The Smartest Guys in the Room: The Amazing Rise and Scandalous
Fall of Enron. Penguin Group, New York.
Mulford, C.W., Comiskey,E.E., 2002.The Financial NumbersGame: Detecting Creating
Accounting Practices. Wiley, New York.
NationalVenture CapitalAssociation,undated.The Venture CapitalIndustry – An Overview.
Downloaded on May 18, 2005. Available from: <http://www.nvca.org/def.html>.
Partnoy, F., 2002. Testimony of Frank Partnoy Professor of Law, University of San Diego School
of Law. Hearings before the United States Senate Committee on Governmental Affairs, Januar
24.
Schilit,H., 2002.FinancialShenanigans:How to Detect Accounting Gimics Fraud in Financial
Reports, second ed. McGraw-Hill, New York.
Toffler,B.L., Reingold,J., 2003.Final Accounting:Ambition,Greed,and the Fall of Arthur
Andersen. Broadway Books, New York.
US Senate, 2002. The role of the board of directors in Enron’s collapse. Report Prepared by the
Permanent Subcommittee on Investigations of the Committee on Governmental Affairs, Unite
States Senate, 107th Congress 2nd Session, Report 107-70, July 8.
Wyatt, A.R., 2004. Accounting professionalism – They just don’t get it. Accounting Horizons 18
(March), 45–54.
Zeff, S., 2003. How the US accounting profession got where it is today: Parts I and II. Accounting
Horizons 17 (September/December), 189–205, 267–301.
484 G.J. Benston / Journal of Accounting and Public Policy 25 (2006) 465–484