Spero Consulting

Archive for May 5th, 2009

Or, Is It the Naïve Government’s Stress Test­ing Exercise?

More Lack of Plan­ning and Insight from Our Reg­u­la­tors and Gov­ern­ment Officials

About one month ago – on April 7, to be pre­cise – we asked, Where Will the Bank Stress Test­ing Exer­cise Lead?

In that post, we wrote that the tests could be designed one of three ways: (1) with a pos­i­tive bias to ensure that all or almost all of the banks could pass the tests, (2) with no bias to get a hon­est — though not nec­es­sar­ily accu­rate — assess­ment of each bank’s finan­cial con­di­tion (with accu­racy con­strained by the implicit and explicit assump­tions built into the exer­cise), or (3) with a neg­a­tive bias to ensure that most or all banks fail the test.

Given the var­i­ous news reports that four­teen of the 19 banks may have “failed” the tests and that the four­teen have since been nego­ti­ated down to ten that may “require cap­i­tal,” it doesn’t seem that the tests were designed or biased to gen­er­ate pos­i­tive results. In ret­ro­spect, it doesn’t seem that the eco­nomic assump­tions were par­tic­u­larly neg­a­tive – see We Can’t Sub­si­dize the Banks For­ever in today’s edi­tion of The Wall Street Jour­nal for evi­dence that first quar­ter eco­nomic activ­ity and sta­tis­tics were worse than pro­jected in the exer­cise. Note, how­ever, that if they were designed with a pos­i­tive or opti­mistic bias, then the reg­u­la­tors who designed the Super­vi­sory Cap­i­tal Assess­ment Pro­gram (SCAP) wre/​are hor­ren­dously clue­less and incom­pe­tent, and that’s not out­side the realm of possibility.

As we wrote last month, we can’t imag­ine any­one design­ing a neg­a­tive bias into the tests; so, that means that, most likely, the gov­ern­ment sought an “hon­est” though not nec­es­sar­ily accu­rate assess­ment of each bank’s abil­ity to absorb addi­tional losses.

That was and is prob­lem­atic given the amount of pub­lic­ity gen­er­ated about the pro­gram. It would have been much bet­ter to per­form the tests in total secrecy – in what appeared to be a dis­jointed, dis­or­ga­nized, ad hoc, and unsys­tem­atic man­ner to belie any sense that a thor­ough inves­ti­ga­tion or com­pre­hen­sive and national approach was being under­taken. (They should have been stan­dard­ized but secret tests with no pub­lic­ity or acknowl­edge­ments of their existence.)

The three-​day delay in announc­ing their find­ings shows that the reg­u­la­tors – the Fed, the OCC, etc – were unpre­pared for the results. As we wrote back then, there was no sce­nario analy­ses of the stress test out­comes. For exam­ples, what will we do if four­teen banks require more cap­i­tal, all nine­teen, what about two giant ones, etc?

It’s another exam­ple of gov­ern­ment offi­cials being too rash and not thought­ful enough for their own – and the economy’s – sake. That’s why the road to hell is paved with good intentions.

When we find the time, we’ll expand this post later today or tomor­row, but the events of this week show that the government’s response to the Liq­uid­ity Cri­sis, which is, in fact, a cri­sis in con­fi­dence in finan­cial inter­me­di­aries, is no more thought­ful than its reac­tion to the Mort­gage Débâ­cle, and that pan­icked and over-​publicized response cre­ated the Liq­uid­ity Cri­sis in the first place.

Please, folks, first “do no harm,” which means that you have to think before act­ing or cal­cu­lat­ing. Now where have you seen that before?

Last week we wrote two related posts: A New Influenza Stress Test and Influenza Pan­demic Stress Test, Part II. Both posts dis­cuss the need for banks to per­form stress tests/​scenario analy­ses that incor­po­rate the pos­si­ble neg­a­tive eco­nomic effects of a flu pan­demic in addi­tional to con­sid­er­a­tion of pos­si­ble addi­tional struc­tural weak­nesses (and shrink­age) in the economy.

In the sec­ond post, we men­tioned a gov­ern­ment study from a few years ago that esti­mated a five per­cent con­trac­tion in GDP if the USA faced a severe pan­demic. (In our best Jack Nicholson/​A Few Good Men courtroom-​scene imper­son­ation, we ask: is there any other kind of pan­demic, Danny?)

In today’s edi­tion of The Wall Street Jour­nal, Robert J. Barro and Jose F. Ursula pro­vide addi­tional evi­dence of the pos­si­ble neg­a­tive effects of a pan­demic in Pan­demics and Depres­sions. In it they pro­vide esti­mates of the his­tor­i­cal costs of such out­breaks. Well worth reading.

We’ll have more to say about the stress tests in our next post. The past week’s events pro­vide evi­dence to con­firm one of our hypothe­ses from a post one month ago when we asked Where Will the Bank Stress Test­ing Exer­cise Lead?