Understanding the Mark to Model Concept

All the recent buzz about the reintroduction of “mark-to-market” rules is a result of almost a decade of the use of a different (and arguably fraudulent) method called “mark-to-model”.

The first of two underlying principles of a “mark-to-X” system is that investors have a right to know the value of any asset they purchase, be it a house, bond, or derivative. The second is that the federal government needs to be able to keep track of whether a bank is “liquid” or “illiquid”, and must know the values of bank assets to do so. The question asked by the mark-to-X system is “how much is the asset worth right now?”. Under mark-to-market, the answer is simply the current value at which the asset could be sold. Mark-to-model answers with the predicted value that the asset might have at some future time, or with an estimate of the current value that is not based on market prices.

Silly as it may seem by contrast to mark-to-market, mark-to-model is the only viable way to evaluate assets like derivatives and other complex financial instruments. Generally, there is no market that these can be immediately sold on, so their value would be $0 under mark-to-market. By contrast, mark-to-model would do a better job of realistically approximating the actual value of the asset in question. Thus, the $0 might suddenly become a far more realistic value such as $100,000.

The argument for using mark-to-model is that any asset for which there is no market is undervalued by mark-to-market, and that this creates an unfair burden on banks and other financial institutions which hold these assets. Banks are required to maintain roughly 10% of all deposits in “liquid” form (cash, or anything that can be turned into cash easily), and bankers claim that by shuffling derivatives around – even though there is no market – allows them to turn these into cash so quickly that they are effectively liquid. Counting the entire derivative as part of that 10% makes it much easier for them to meet the requirement.

The potential for abuse of the mark-to-model concept is clear: anyone who develops a fraudulent model can claim to own valuable assets, while in fact being a deadbeat with nothing of value. The fraudulent model does not even need to be fraudulent by intent: as we have learned in the past two years, a model of the predicted value of subprime-mortgage-derivatives can be wildly off the mark, leading to widespread bank failures, and even global economic collapse.