Spero Consulting

Archive for May 8th, 2009

In the pre­vi­ous post we noted a WSJ edi­to­r­ial that stated, “The best that can be said about the stress tests is that they’re over.”

We wouldn’t go that far, but we’ve been highly crit­i­cal of the reg­u­la­tors for a num­ber of rea­sons. (See our var­i­ous posts about the stress tests, and you can search for our posts on reg­u­la­tors and cap­i­tal require­ments if you want.) Here’s a new crit­i­cism that we haven’t men­tioned before.

Within firms, it’s almost never the case that stress tests and sce­nario analy­ses are per­formed to argue for an increase in cap­i­tal allo­cated to an invest­ing or trad­ing activ­ity. (Yeah, yeah, we know about eco­nomic cap­i­tal, which is the­o­ret­i­cally fine but prac­ti­cally worth­less.) Occa­sion­ally, if the results of pes­simistic sce­nar­ios show large losses, the trader or invest­ment man­ager is required to take an action, e.g., sell the asset, end the trade, or hedge the trade to some extent. (We ital­i­cize Occa­sion­ally in the pre­vi­ous sen­tence because too often the results are dis­missed as deriv­ing from an “unlikely” or “improb­a­ble” or “unre­al­is­tic” sce­nario and noth­ing is done to mit­i­gate the poten­tial tail losses).

Now, we know that some firms are sell­ing bits and pieces of them­selves to raise cap­i­tal, but we haven’t heard of any banks sell­ing the bits and pieces of them­selves that lost the most in the stress tests or cur­tail­ing activ­i­ties that per­formed par­tic­u­larly poorly in the tests. Per­haps we don’t read enough, per­haps the banks are being secre­tive and other tac­tics are being rec­om­mended and applied, but it seems that the government’s pre­ferred response in every case is “raise cap­i­tal.” Only the degree changes.

While there is a very small chance that such a tac­tic is robust (globally-​optimal), it’s unlikely that share­hold­ers con­sider the fur­ther dilu­tion of their inter­ests to be the sin­gle best response to every sin­gle, stress-​induced defi­ciency. Reg­u­lar read­ers will recall that we think forced receiver­ship and nation­al­iza­tion have been under-​utilized tac­tics dur­ing the cri­sis; so, we’re not a shill for bank investors, but given the num­ber of con­trols and alter­na­tives avail­able, it doesn’t seem as if the Trea­sury Depart­ment and the Fed­eral Reserve have con­sid­ered those options.

In our mind, gov­ern­ment solu­tions tend to be sim­plis­tic and heavy-​handed; so, we can’t give it (them) the ben­e­fit of the doubt. For exam­ple, this week the chair­man went to pur­chase decon­ges­tant allergy pills at the local CVS. She was denied because she had already pur­chased 30 pills dur­ing the past month. We both take allergy pills; so, that’s 60 pills per month. That means we both have to make two trips per month to get the med­i­cine we need. That’s not environmentally-​friendly, and we doubt that pre­vents any­one from get­ting the ingre­di­ents they need to pro­duc­ing meth. Those restric­tions – like most air­port secu­rity pro­ce­dures – are sim­ply incon­ve­niences for law-​abiding citizens.

Such actions are showy, expen­sive, but inef­fec­tive. Now, why do they remind us of TARP and SCAP?

Is that Really the Worst-​Case?

Today’s front-​page arti­cle in the The Wall Street Jour­nal, Fed Sees Up to $599 Bil­lion in Bank Losses, is sub­ti­tled “Worst-​Case Cap­i­tal Short­fall of $75 Bil­lion at 10 Banks Is Less Than Many Feared; Some Shares Rise on Hopes Cri­sis Is Easing.”

While it is the worst of the two cases ana­lyzed by the reg­u­la­tors, it is not the worst, rea­son­able case that could be imag­ined. Per­haps that’s why the short­fall was “Less Than Many Feared.”

• It’s rea­son­ably pos­si­ble for the down­turn to be deeper and/​or longer than con­sid­ered. In fact, one could assign a rea­son­able, non-​trivial prob­a­bil­ity to it.
• It’s pos­si­ble – maybe even rea­son­ably prob­a­ble – that given the sce­nar­ios used, the banks (and the reg­u­la­tors) pos­i­tively biased the val­u­a­tion results. (Con­vert­ing macro-​economic assump­tions into asset val­ues is a highly spec­u­la­tive and sub­jec­tive busi­ness – regard­less of the num­ber of cal­cu­la­tions per­formed to gen­er­ate those values.)
• It’s quite pos­si­ble that a man-​made or nat­ural dis­as­ter could com­pound the economy’s prob­lems and shave sev­eral addi­tional per­cent­age points off of GDP dur­ing the next two years. For exam­ple, we’ve writ­ten sev­eral times in recent weeks about the need for sce­nar­ios that include the effects of swine-​flu pan­demics, but large earth­quakes in Cal­i­for­nia, severe drought in the Mid­west, and mas­sive hur­ri­canes and floods in the South­east could be just as dev­as­tat­ing. While none of them is likely, they’re all pos­si­ble – even rea­son­ably possible.

So, we ask, is that sec­ond case really the worst-​case?

The Journal’s Nar­ra­tive Fallacy

While we take issue with the sub­ti­tle of the above arti­cle, there’s another arti­cle in today’s The Wall Street Jour­nal, How the Stress Tests Stopped the Mar­ket Bleed­ing, that’s even more dubi­ous. In fact, the entire premise of it is flawed.

Does any­one other than the two cred­u­lous reporters believe the title to be the case? Geithner’s promise that none of the 19 would be allowed to fail may have had a big­ger effect, but the stress tests them­selves? The evi­dence for that is shame­fully inad­e­quate. So we ask: who are they try­ing to fool?

From our per­spec­tive, there’s noth­ing in the arti­cle to jus­tify the title’s con­clu­sion. In fact, our ref­er­ence to it as a “Nar­ra­tive Fal­lacy” is too gen­er­ous, because if it were, there would be a sequence or set of facts that could be con­cocted to tell such a tale, but there doesn’t seem to be such a set, here. Instead, it seems more like a coin­ci­dence, i.e., it seems that one could eas­ily argue that North Car­olina win­ning the NCAA tour­na­ment had more of an effect or that the mar­ket decreased as tem­per­a­tures cooled in the north­ern hemi­sphere and has begun to rise with the arrival of Spring’s warmer weather. Both of those “expla­na­tions” seem just as compelling.

As Nas­sim Nicholas Taleb has pointed out many, many times, such fal­la­cious story-​telling is all too com­mon in the busi­ness press, where reporters con­stantly ascribe causes and rea­sons to daily (and hourly) changes in prices and indices.

For­tu­nately, the Paper Is Schizophrenic

By that we mean that unlike the report­ing staff, the edi­to­r­ial page writ­ers are a bit more skep­ti­cal about the ben­e­fits of the stress tests, in par­tic­u­lar, and gov­ern­ment involve­ment in the finan­cial sec­tor, in gen­eral. In their top Review and Out­look col­umn today, Stress for Suc­cess?, they con­clude: “The best that can be said about the stress tests is that they’re over.”

We think that’s a bit too harsh, but not by much. See our var­i­ous posts about the stress tests. The next post lists a new complaint.