Stress tests. May 6

This whole thing about the stress test is becoming a complete and totally useless joke. With all the leaks it is also quickly becoming irrelevant and most definitely does not deserve the attention that it has been given. Most stress test pretend to forecast what would happen to a system/ situation if certain attributes, credit ,interest rates, volume , counter-parties etc. etc. are tweaked a little this way or that, or, if done more professionally, by using statistics or other sophisticated methodologies to gauge such outcomes. Having actually run the treasury of two different foreign banks back in the early/mid eighties, I can give some interesting numbers without naming names.

The balance sheet total of one of these banks,(actually not yet a bank as the Bank Act had not yet passed) was about 2 bln. For the non-accounting types. that is, both sides of the balance sheet. Total capital was $400.000 making leverage 5000x  (the  BIS norm is somewhere in the neighborhood of 11 to 13X). At 5000x leverage only a single hair on your head needs to be combed the wrong way for the whole thing to collapse.

Then there is liquidity; all money was raised by commercial paper issuance for which we had the appropriate offering memorandum in which it was stated that we had adequate back-up lines ( a requirement for CP). In fact we did , all of 2 mln. backing up 2 bln. or 1/1000.

Then there is counter-party risk. Our largest counter party was actually Dominion Securities with whom most of the commercial paper was issued. On a typical day, because I was always long US $$ and they were always short we would start of with with 100mln tom/next, spot/next in the FX forward market,usually as a simple roll-over. There never was any netting as this was deemed too confusing so everything was paid in full. The net effect, on a typical day, not including “other” business of which there was a lot, would be that about $200 mln a day was paid in US$ Fed Funds or Clearing House Funds and the counter-value received in “balances” in Canadian $$. The problem was that you made the payment without knowing if you received, or would receive the counter value. The counter-party risk on these transactions was in the order of 500X  our capital for DS and, given that, at the time DS had a capital of around 16/20mln $$, about 10/15x their capital. Just a single failure, would have killed you and just to put it in perspective both A. E . Ames as well as Wood Gundy actually did hit the wall in 1987  due to general mismanagement and the British Petroleum deal respectively. DS was not far behind and even  a cameleon  would be envious seeing how fast DS changed convictions from newly minted stock company to bank owned “subsidiary”.

Then there were accountants, as a bank we had to rotate but invariable it would be Anderson , KPMG or one of the other 4 or 5. In many cases we would do the issuance of US$ commercial paper with the FX spot sale and forward repurchase as a single deal. Amazingly separate  accounting for the issuance (one counter-party) and the FX ( a different counter –party) was NOT required. Today people simple would not believe this.

There are quite a few other types of risk, too many to get into and some very technical, however,  the point is that the then, and I believe now, operates to a very large degree on faith. The Madoff affaire , if nothing else proves that, stress tests are there for internal use, little more.

By the way, these tests are done on the basis of a “ more adverse” case, not a worst case. Fill in the blanks, what if drops another 10-20 % would the banks  need another 2 trillion?? Perhaps we should get rid of the FDIC and let the depositors decide, let market forces work.