Editor's Note: Dan Englander, formerly a Managing Director at Allen & Company, is currently Managing Partner at Ursula Capital Partners, which invests in US equities. Mr. Englander is also a director of Copart, Inc. and America's Car-Mart and serves as head of the compensation committee for both companies.

Much has been written about how excessive CEO compensation has been for such crash luminaries as Richard Fuld, Jimmy Cayne, Angelo Mozilo, Stan O’Neal, Chuck Prince, the entire senior management of AIG (AIG), even Lloyd Blankfein and John Mack.

These executives all presided over firms who used excessive leverage to make enormous financial commitments, and were able to take multi-generational wealth off the table for themselves - even though the loans ultimately went very bad. So bad, in fact, that shareholders (and in some cases bondholders) received next to nothing. What went wrong?

The problem is not the amount of compensation. In our economic system, executives should be able to earn virtually unlimited sums. It's right that talented people who create value earn just rewards - but the criteria we currently employ to calculate value are wrong.

Simply put, these executives -- and, for that matter, every executive who ever worked in the lending/mortgage/credit industry -- were keeping track of the wrong things. If the criteria were correct, their ultimate compensation would be dramatically lower.

To understand this, one must first understand the limits of GAAP accounting. GAAP works very well when someone manufactures a product and sells it with no contingencies. The manufacturer/merchant has a cost to acquire the product, and his profits will be whatever his sales are less his fixed costs. No assumptions are required to calculate these profits. So this discussion doesn't apply to industrial companies that sold a product without a financing component, or those instances where the customer paid for the product without financial assistance from the merchant.

Now consider financial company earnings. And not just banks and mortgage lenders: Consider also automobile companies (“zero percent down” financing), insurance companies, furniture companies, even bargain retailers such as Target (TGT). When a loan is made -- or credit of any kind is extended -- GAAP earnings require certain assumptions be used to calculate current period earnings.

If a finance company loans $1 million, with an interest rate of 7%, the GAAP math works out as follows:
 

• Principal Amount of Loan: $1,000,000
• Interest Rate: 7.0%
• Assumed Interest: $700,000
• Term of Loan: 5 years
• Assumed Default Rate: 5.0%

GAAP earnings would show an immediate profit on this loan, because the assumed default rate is only 5%. However, the actual cash collections of the loan may be far different. Ultimately, if the default rate is less than the assumed loss rate (5%), the loan will perform better than GAAP; if the default rate is higher, the loan will perform worse; and if the default rate is precisely equal to the assumed GAAP loss rate, the loan performance over time will be equal to GAAP.

The problem that GAAP accounting presents is that it involves assumptions, and assumptions -- no matter how much data supports them, or how honest or well-intentioned the management -- create problems of their own.

It's quite easy to see how finance company earnings can be abused. If you were an executive of the company described above, advancing the million dollar loan at a 5% default rate, you could in fact be paid based on GAAP earnings: Your current income would be based on your assumption of how good the loan you just made was. What possible motivation do you have to acknowledge the loan you made could go bad? The better you make the loan look, the more money you make today.

This is, in part, how millions of garbage subprime loans were rated AAA (as if that term has any current meaning) - and how so many people got rich because of them. History shows very clearly that executives at Bear Stearns, Lehman Brothers and Countrywide all maintained unrealistic “marks” on their portfolios as financial armageddon approached in 2005 and 2006. Can it be a coincidence that by doing so, they inflated their own compensation?