Letter From Prison

In addition to A Letter from Prison, please read the Preface of Eugene's book, Paper to Practice: Why Compliance Programs Fail and How to Build Programs That Work. You will receive this book at the beginning of the program.

The Professor's Questions

1. How serious were Stephen Richards’s actions?
2. If Computer Associates achieved the same financial results through GAAP flexibility, does your answer to Question 1 change?
3. Suppose you were placed in Stephen Richards’s position at Computer Associates and were under pressure to extend the fiscal quarter. How would you have handled the situation differently? What would be the expected consequences of those actions?

My Answers

Q1. How serious were Stephen Richards’s actions?

His actions were very serious. The fact that contracts and cash were genuine does not make the timing immaterial. Investors were told that business had been completed in a quarter when the recognition conditions had not been met. In FY2000 Q2, improperly accelerated revenue was about $558 million, roughly 53% of correctly recorded revenue. Reported EPS of $0.60 beat the $0.59 estimate; without the improper revenue it was $0.05. This changed the apparent operating result, not just the presentation of a minor item.

As global sales head, Richards could influence deadlines and what finance was told about contracts. He was not a passive recipient of an accounting choice. The repeated extension of quarters also taught teams that hitting the number could override the rule. The investigation adds a further concern: the case describes misleading responses by senior executives and an internal counsel who later admitted coaching such behavior. Concealment prevents governance from correcting the original conduct.

I would distinguish explanation, responsibility, and harm. Analyst pressure, compensation, and fear of losing status help explain the behavior. They do not remove responsibility. The case supports serious distortion and damage to trust, but it does not provide a defensible total dollar measure of investor loss. The $2.115 billion sum is accelerated revenue across six quarters, not stolen cash. Nor does the seriousness depend on whether the company later went bankrupt.

I can take his experience and suffering seriously while rejecting the claim that a few days made the conduct unimportant. His letter is especially useful because an intelligent manager can describe the pressures clearly and still understate his own choices.

Q2. If Computer Associates achieved the same financial results through GAAP flexibility, does your answer to Question 1 change?

Yes, my assessment can change, but the answer depends on what “GAAP flexibility” means. If the company uses a genuinely permitted, supportable, consistently applied estimate, with appropriate disclosure and no false contract dates, that is different from deliberate misstatement. The same reported earnings do not make two processes equally wrong. I would not call a legitimate accounting judgment fraud merely because it helps meet a target.

However, an accounting choice still needs evidence and economic meaning. I would ask whether the estimate reflects new information, whether assumptions are applied consistently in good and bad quarters, and whether material effects are transparent. A technically arguable treatment used to obscure deterioration raises a different ethical concern from a neutral estimate. The case alone cannot resolve every hypothetical GAAP treatment, and auditors' acceptance is not a substitute for accurate underlying facts.

Real business actions also deserve separate scrutiny. Delaying useful R&D or an economically sound project may lawfully increase current earnings while destroying long-term value. That can be a poor or harmful management decision without being the same legal category as falsifying the reporting period. Conversely, a sound business decision that improves earnings is not suspicious merely because it improves earnings.

My test has three parts: are the facts and method valid; does the report allow a fair understanding; and does the decision serve sustainable business value? Outcome, intent, process, and disclosure all matter. I would reduce the severity judgment for a genuinely legitimate choice, while still examining any deception or value destruction.

Q3. Suppose you were placed in Stephen Richards’s position at Computer Associates and were under pressure to extend the fiscal quarter. How would you have handled the situation differently? What would be the expected consequences of those actions?

I would refuse to change the reporting period and make that position operational. I would tell the team: keep selling after quarter end, but preserve actual signature and delivery dates and book each deal in the period supported by the evidence. I would ensure finance has the original contract history rather than a target-driven summary. A shortfall should be reported as a shortfall.

Before escalation, I would clarify the facts and the proposed treatment with the controller or CFO, state my objection in a factual record, and request an independent review. If senior finance or internal counsel is implicated, I would go to the audit committee or an independent board channel, with appropriate independent professional advice. Preserve records and do not coach employees on a common story. Disclosure or external reporting obligations would need qualified advice; the classroom plan is not a claim about today's legal procedure.

I would also offer a business response: revise the forecast, explain late contracts honestly, examine excessive end-quarter concessions, and reduce incentives that make a single cutoff decisive. Separate bookings, recognized revenue, cash collection, and customer value in management reporting. Reward accurate forecasting and verified contracts, and protect people who flag concerns. Do not simply add another annual training slide.

Expected consequences are not painless. Earnings could miss expectations, the share price could fall, bonuses could decline, and I might lose influence or my job. The board could initially resist. The 42% share-price reaction in the case shows why the pressure felt real, but it does not forecast the exact reaction to my alternative. The benefit is avoiding further distortion and allowing the company to confront its actual business problem. If the organization insists on continuing and I cannot stop participation, I would be prepared to leave after seeking independent advice and addressing any reporting duties. Leaving alone would not erase what I already knew or did.

Case Summary

Computer Associates (CA) was founded in 1976 in enterprise software and grew into a large software company through acquisitions and global expansion. Stephen Richards joined the Sydney office in 1988, rose quickly, and became head of worldwide sales in 2000. The sales organization revolved around quarterly targets: strong results brought large bonuses and stock options, while missing targets brought pressure.

Customers typically bought three- to ten-year licences. Under the accounting treatment of the time as described in the case, once a contract was signed, the software delivered and collection reasonably assured, the present value of the licence fee could be recognised in that quarter; usually at least 80% of the fee was allocated to the licence. So the signing date of a few large contracts could significantly change a quarter's revenue. Customers knew they had negotiating leverage at quarter-end and often pushed signing to the last minute.

Some managers effectively extended the fiscal quarter, booking contracts signed after quarter-end into the previous quarter. The products, contracts and cash could be real, but the reporting period was not. After an earnings warning in July 2000, the share price fell 42%, showing the capital-market pressure the company faced; that does not establish the pressure as sufficient justification for illegal conduct.

After media scrutiny and regulatory investigations emerged, the initial internal investigation did not uncover the problem. In 2003 the board asked an independent law firm to investigate further, which found 23 boxes of previously missed documents and more electronic evidence. The investigation involved finance executives and ultimately implicated Kumar, Richards and others. The case describes executives misleading the investigation, and the head of internal legal later admitted to directing executives to mislead it. "We already asked the company's lawyers" cannot be treated as sufficient for an effective independent review.

Richards pleaded guilty in 2006 and was sentenced to seven years in prison. In 2008 he wrote from prison to Eugene Soltes, then a student. He explained how analyst expectations fed into internal targets, and how targets shaped teams, organizational structure and incentives; he also discussed how self-identity and recognition for performance can be stronger drivers than money, and stressed that managers often operate in grey areas. The letter preserves his explanations and self-understanding; it is not an independent conclusion about his conduct.

The professor asks us to assess how serious the conduct was, then to judge whether that changes if the same numbers were achieved through flexibility GAAP allows, and finally to put ourselves in his shoes and propose a different course of action and its consequences. The focus of the class is not only recognising the wrongdoing, but whether you can adopt a workable arrangement at the moment of greatest pressure.

Key Data

Fact/Calculation	Content and Basis	Source
Contracts and recognition	Licences of 3–10 years, usually at least 80% allocated to the licence fee; under the rules at the time per the case, required signing, delivery and reasonable assurance of collection	p.2; not guidance on current accounting policy
Executive incentives	Richards's base salary was US$500,000 a year plus options; some high-performing salespeople earned over US$1 million	pp.2–3
Earnings warning	Share price fell 42% after the July 2000 warning	p.3; cannot be equated with investor losses caused entirely by the conduct in this case
FY2000 Q2	Correct revenue $1,047,256,904; prematurely recognised $557,743,096; inflated by about 53.26% relative to correct revenue	Exhibit 4, p.16; the 53% is not the share of premature revenue in total reported revenue
FY2000 Q2 EPS	Analyst expectation $0.59, reported $0.60, $0.05 after removing improper recognition	Exhibit 4
FY2000 Q3 EPS	Expected $0.90, reported $0.91, $0.31 after removing improper recognition	Exhibit 4
Total premature recognition over six quarters	244,834,719+557,743,096+572,097,259+378,868,969+142,400,000+218,960,000=$2,114,904,043, about US$2.115 billion	Derived from Exhibit 4; a total of period misstatements, not fictitious sales or cash losses
The four FY2000 quarters	Reported EPS met or beat expectations in all four; adjusted EPS fell short in all four	Exhibit 4
Evidence and investigation	The 2003 independent investigation received 23 boxes of previously missed documents; in 2004 more emails revealed executives' involvement	p.4
Conviction and letter	Guilty plea in 2006, seven-year sentence; letter dated 14 February 2008	pp.1, 5, 7
The letter's explanations	External expectations → internal targets → team pressure; recognition and self-identity; absence of a control framework; grey areas	Exhibit 2 pp.7–13; the account of the person involved
Earnings management research	In a survey of 312 CFOs/finance executives, 59% said they might forgo a positive-NPV project to avoid missing analyst expectations, and 80% would cut discretionary spending	Exhibit 3 pp.14–15; sample and hypothetical questions, cannot be extrapolated to the actual share of behaviour across all companies

Key People

Person/Institution	Role	Responsibility or Information Position
Stephen Richards	Head of worldwide sales	Influenced signing pressure, sales instructions, and information on cross-quarter contracts; responsibility can't simply be pushed onto accounting
Sanjay Kumar	Company president/senior leader	Influenced targets, organizational culture and explanations to the board
Ira Zar	CFO	The investigation found him linked to backdated contracts; financial gatekeeping failed
Steven Woghin	Company general counsel	The case says he later admitted to directing the misleading of the investigation, showing that internal escalation channels may themselves be compromised
Board, audit committee and independent investigating lawyers	Governance and oversight	Need complete original evidence and independent sources of information
Analysts, investors, customers, employees	Holders of external expectations and those affected	Timing attribution affects judgements, trading prices, incentives and trust

Core Tensions

1. The choice between admitting a missed quarterly target and changing the reporting period to preserve an image of success. The short-term cost is real and cannot be eliminated by rewriting the facts.
2. A real transaction and truthful reporting are requirements at two different levels.
3. Legitimate room for judgement and intent to mislead must be distinguished item by item; "everyone does it" cannot substitute for analysis.
4. Understanding the pressure and holding people accountable can both hold true; explaining the mechanism does not absolve individual choice.
5. There is a gap between control responsibilities on paper and whether they are carried out in the face of real incentives.