One of the things that has intrigued me about the financial crisis is that it is pretty clear that looting took place at the high end of the financial services industry, yet few have called it by that name, in part because it has been difficult to identify the mechanisms by which it occurred.
Looting, as described by George Akerlof and Paul Romer, occurs when business owners go broke at society’s expense (loot) rather than go broke as the unfortunate result of having gambled on success and lost. And looking at Wall Street, the fact set strongly suggests that looting took place, In the old days of private partnerships, firms might blow themselves up on an individual basis, and you’d occasionally see serious industry downdrafts thinning the herd (the back office crisis of the late 1960s, the one-two punch of the end of fixed equity commissions plus the down leg of the 1970s bear market). But having the industry run of the cliff en masse was previously limited to commercial banking.
Now when William Black, a senior bank regulator during the S&L crisis, has talked about this phenomenon, he frames it as control fraud, meaning orchestrated in a deliberate fashion at the top level of a firm. While that probably contributed, particularly if an organization was desperate, but I suspect more complicated mechanisms played a big role.
To give a very simple example: back in the ages of stone knives and bearskins, when I was young, Citibank engaged in double and sometimes triple counting to encourage different groups internally to play nicely together. Thus if a deal yielded $1 million in profit center earnings, each operation that worked on it booked $1 million in profits.
The problem that then resulted was that individual units met their targets, but at higher levels, where the double/triple counting was netted out, the band didn’t.
My suspicion is we have had a massive version of an analogous version of the same problem, in which the internal (and probably even the reported) accounting did not accurately reflect the true economics of the business. And the way this turned out to be looting (as opposed to mere stupidity of the sort practiced traditionally by commercial banks) is that on Wall Street, a very big percentage of revenues is paid out as comp. So exaggerated earnings translate very directly into bigger pay packages.
Bruce Krasting, via Institutional Risk Analytics, sets forth one example:
Simplistic example. You are a bank. You have two customers. Southwest Airlines (NYSE:LUV) and Chevron (NYSE:CVX). SW calls and says, “Lets do a deal. Every 6 months for five years we will swap money based on 1 mm barrels of oil at $60. If the then cash price is less than $60 I pay you. If it is more than $60 you pay me.”
You say “fine” and then call Chevron. You enter into the exact opposite of the first swap trade. You make money doing this. You have a matched book. As a result of your purchases and sales you will have $50,000 in profit every six months for the next five years or a total locked in gain of $500,000.
At the end of the day your boss asks, “How did you do today?” You answer, “we made $500,000”. He says, “Great. Here is your bonus”. If you went to three market makers who did exactly the same as you did on that day the profit and loss results might be significantly different. For example:
– One could report a gain of $500,000. That would be justified. But would overstate the economic results
– Another might just record a gain of $100K (the current year portion that is “locked in”). The balance of the income is realized annually for five years. This would understate income. No one like that.
– Yet another might book just $400K of gains currently. This would represent the NPV of the $500k over the five-year period based on an internal capital rate set by the banks management. This is close to economic reality and is the form that most banks use to value this type of business. While it is the most appropriate and reasonable approach, it too is badly flawed.
The next day your bank announces quarterly earnings and says, “We made $400K. We are paying a dividend based on that and we are paying bonuses based on that. Net-net, retained earnings will go up by $250,000. It was a good quarter”
In this very simple example there is no cash from this profit. The cash will come to you over five years. So you have to borrow to pay the dividend and the bonus and all the other current expenses relating to this profit.
What you have done is ‘borrow’ retained earnings from future cash flow. You look like you have retained earnings to support your tier one capital ratios. But when there is a real ‘cash call’ on your equity (2008) you do not have the cash equity to survive. So you have a TARP party.
The OTC reform proposal plan does not give the banks the ability to do this. And that is why they do not like it. This same income recognition issue exists in the OTC swap market for currencies with maturities greater than one year.
Since the big financial intermediaries claim to run netted CDS books (and we found at least one big stuffee, AIG, who was eating a ton of risk), CDS alone could have played a major role in future profits being accelerated and treated to a large degree as cash flow earnings.
Do readers have any sense of how prevalent this was? I would assume it was close to universal for CDS. And what other products would have been subject to it besides CDS?
If you are not set up to comment, please e-mail me at firstname.lastname@example.org. Thanks!