The Uncertainty Principle in Accounting

The debate rages on over how to do proper accounting for financial institutions. After each major debacle, there is a stampede to some other method. Maybe the chronic problem is the ongoing assumption of modern accounting that every security has an instantaneous value based on its expected revenue and its risk.

The present “value” of a future receivable is the subject of speculation. Laws that constrain how much banks may lend should not presuppose any relation between the two. The debate of “mark to market” accounting versus “mark to model” accounting, each of which makes such a presupposition, misses this point.

Take, as an example, a simple toy model: Two mortgage banks each lend $300,000 to two home buyers, in return for $700,000, $300 K principal and $400 K interest, in monthly payments over the next 30 years. They can each make $400 K profit over that rather long time scale. How much is the right to collect the $400 K in the future worth at the present time? Who knows? Banking laws should not be formulated in such a way that their enforcement requires knowing it.

Now suppose each bank sells the other an identical “security” backed by the mortgages (MBS), for $500 K, soon after the loans are made. The exchange of identical securities is obviously an empty exercise in terms of meaningful wealth, yet, on paper, each bank has lent out $300K and received $500K shortly thereafter. Sounds like a $200K profit, right? True, they spent $500K buying each other’s MBS, but in return they each got an MBS worth (for the time being) $500K, so they broke even on that; you can hardly consider it a loss. Where, then, did the $200K profit come from? It came from the future, and was declared a profit in the present. With that profit, the banks can immediately bloat their operating costs by paying its employees bonuses, larger salaries, and hiring more employees.

Now the paper trail in the mortgage industry is of course far more complicated than the toy model, but I bet that underneath all the complicated mortgage derivative paper, the same principle was at work: anticipated future wealth was cashed in and distributed in the present. For example, the obligation in a credit default swap or insurance policy on a mortgage may be activated as soon as the default is declared, even though the default was on a future debt repayment. Creditors that collect their insurance on a defaulted debt enjoy a payment in the present on a debt that would have been collected in the future.

Debating how much anticipated future revenues are worth at present does not change a basic fact of life: A promise is worth less than what is promised. The risks that the promise might not be kept are a form of negative `wealth. Mortgage-backed securities simultaneously create mutually canceling positive wealth (short term profits) and negative wealth (future risk) out of thin air, the way positive and negative charge are created by a spark. In this manner, they allow present wealth to flow, like electric current through a short circuit, from the depositors to the bank administration. In the above example, the promise of $400 K in future interest payments is worth only $200 K at present. The risk that it might not be paid, which is worth minus $200 K, is the “hole” left by the profit grabbed by the bank. There is only one acceptable limit to the amount of negative wealth that mortgage banks should be allowed slip to its depositors without their consent: zero. (Would you debate how many short circuits should be acceptable in a car battery?)

The banking laws should therefore insulate the future from the present to the full extent possible. A proper evaluation of a bank’s solvency should respect the dimension of time; it should both keep track of the bank’s books and regulate their activities accordingly on a year-by-year basis well into the future. The amount of interest they promise to pay out in 2015 should be tied to the amount they expect to collect in 2015. The maximum amount of debt to which they are liable at present should be tied to their present reserves, not to anticipated future revenues. Confusing future receivables with present assets is asking for trouble, no matter by what mathematical modeling procedure you compare the two.

Uncertainty in how much anticipated future interest payments are worth at present should remain the risk of speculators, not depositors. The depositors, whose money was loaned to the home buyers, are already assuming a fair share of the risk merely by parting with cash in exchange for the collateral, whose value may yet go down, and granting occupancy to the home buyer. Anyone’s rights to a share in the future interest payments should be predicated on their waiting patiently for them.

Physicists used to think that a particle had a particular energy at any instant, but finally, in formulating quantum mechanics nearly a century ago, they realized that it doesn’t. The shorter the time interval over which you try to guess the energy of a particle, the greater the uncertainty in your guess. Only after physicists humbly (and therefore with great difficulty) accepted this uncertainty did they proceed to the dramatic advances of the 20th century in understanding the microscopic structure of matter. Maybe it is time for economists and accountants, now puzzling over the microscopic details of failed financial instruments, to make a similar transformation, and accept the fact that future receivables simply don’t have a well-definable present value.

Short term gain and long term consequences go together like love and marriage. The current difficulties of lending institutions is what we folks get from their recent whoopie. The future has just arrived, and there is plenty more of it.

David Eichler

June 16, 2009