On Not Having an Advance Abyssal Plan

“Even though he could fore­see the prob­lem then, we can see it equally well now. There­fore, if he could fore­see the solu­tion then, we should be able to see it now. After all, Sel­don was not a ma­gi­cian. There are no trick meth­ods of es­cap­ing a dilemma that he can see and we can’t.”
-- Salvor Hardin

Years ago at the Sin­gu­lar­ity In­sti­tute, the Board was en­ter­tain­ing a pro­posal to ex­pand some­what. I wasn’t sure our fund­ing was able to sup­port the ex­pan­sion, so I in­sisted that—if we started run­ning out of money—we de­cide in ad­vance who got fired and what got shut down, in what or­der.

Even over the elec­tronic aether, you could hear the un­com­fortable silence.

“Why can’t we de­cide that at the time, if the worst hap­pens?” they said, or some­thing along those lines.

“For the same rea­son that when you’re buy­ing a stock you think will go up, you de­cide how far it has to de­cline be­fore it means you were wrong,” I said, or some­thing along those lines; this be­ing far back enough in time that I would still have used stock-trad­ing in a ra­tio­nal­ity ex­am­ple. “If we can make that de­ci­sion dur­ing a crisis, we ought to be able to make it now. And if I can’t trust that we can make this de­ci­sion in a crisis, I can’t trust this to go for­ward.”

Peo­ple are re­ally, re­ally re­luc­tant to plan in ad­vance for the abyss. But what good rea­son is there not to? How can you be worse off from know­ing in ad­vance what you’ll do in the worse cases?

I have been try­ing fairly hard to keep my mouth shut about the cur­rent eco­nomic crisis. But still -

Why didn’t var­i­ous gov­ern­ments cre­ate and pub­lish a plan for what they would do in the event of var­i­ous forms of fi­nan­cial col­lapse, be­fore it ac­tu­ally hap­pened?

Never mind hind­sight on the real-es­tate bub­ble—there are lots of things that could po­ten­tially trig­ger fi­nan­cial catas­tro­phes. I’m will­ing to bet the Amer­i­can gov­ern­ment knows what it will do in terms of im­me­di­ate res­cue op­er­a­tions if an atomic bomb goes off in San Fran­cisco. But if the US gov­ern­ment had any ad­vance idea of un­der which cir­cum­stances it would na­tion­al­ize Fan­nie Mae or guaran­tee Bear Stearns’s coun­ter­par­ties, this plan was not very much in ev­i­dence as var­i­ous gov­ern­ment offi­cials gave ev­ery ap­pear­ance of try­ing to figure ev­ery­thing out on the fly.

A pub­lished, be­liev­able ad­vance plan for the worst case—one that you could ac­tu­ally be­lieve the gov­ern­ment would carry out, in­stead of junk­ing the plan to try to keep the top spin­ning a lit­tle longer—would have made the mar­kets that much less un­cer­tain.

If you don’t pub­lish a plan for catas­tro­phe—or can’t pub­lish a be­liev­able plan—then the mar­ket just tries to guess what a re­al­is­tic, be­liev­able plan would look like. If that re­al­is­tic, be­liev­able plan in­volves fran­ti­cally at­tempt­ing to bail out the large fi­nan­cial en­tities in or­der to keep the whole sys­tem from melt­ing down fur­ther, you have moral haz­ard. If they ac­tu­ally do it, that’s lemon so­cial­ism (pri­va­tized up­side, pub­lic down­side).

If that’s what hap­pens in the abyssal case—then not pub­lish­ing that fact, doesn’t pre­vent any­one from fore­see­ing it. If you pub­lish that plan, maybe it will start a de­bate about whether to break up Bear Stearns into smaller en­tities, or change the plan to give coun­ter­par­ties a pre­dictable 10% hair­cut, or claw back ex­ec­u­tive bonuses no mat­ter what their con­tracts read (be­cause you re­ally aren’t sup­posed to screw up so badly that the gov­ern­ment has to get in­volved)...

But if you can’t pub­lish a re­al­is­tic, be­liev­able ad­vance abyssal plan that doesn’t call for res­cu­ing the huge en­tities—then who are you even kid­ding?

Govern­men­tal agen­cies failing to stare into the abyss in ad­vance gives us a dou­ble prob­lem: moral haz­ard as coun­ter­par­ties and in­vestors try to guess what the gov­ern­ment will re­al­is­ti­cally do; and fear and un­cer­tainty in the mar­ket when the worst does hap­pen.

It’s ques­tion­able whether the gov­ern­ment should be in the po­si­tion of try­ing to fore­cast the abyss—to put a prob­a­bil­ity on fi­nan­cial melt­down in any given year due to any given cause. But ad­vance abyssal plan­ning isn’t about the prob­a­bil­ity, as it would be in in­vest­ing. It’s about the pos­si­bil­ity. If you can re­al­is­ti­cally imag­ine global fi­nan­cial melt­downs of var­i­ous types be­ing pos­si­ble, there’s no ex­cuse for not war-gam­ing them. If your brain doesn’t liter­ally cease to ex­ist upon fac­ing sys­temic melt­downs at the time, you ought to be able to imag­ine plau­si­ble sys­temic melt­downs in ad­vance.

Sure, you might have to make some mod­ifi­ca­tions on-the-fly be­cause you didn’t get the ex­act causes and cir­cum­stances right. But it shouldn’t be ob­vi­ous and pre­dictable that the mod­ifi­ca­tions will con­sist of “Oh dear it’s more awful than we planned for and the sys­temic haz­ard is worse and now we re­ally do have to bail out ev­ery­one even though we said we wouldn’t.” Then the plan is not be­liev­able.

So long as the plan is not wrong in the stupidly ob­vi­ous di­rec­tions, it’s hard to see how we’d be worse off if the gov­er­nors of the Fed­eral Re­serve had taken a week once per year to play through sce­nar­ios more night­mar­ish than this one in their minds, de­cid­ing in ad­vance what to do about it, re­al­is­ti­cally.

I sup­pose the main ar­gu­ment against pub­lish­ing the plan would be that the un­in­formed pub­lic (i.e. Congress) would re­volt against the emer­gency plans, de­mand­ing that un­be­liev­able plans be sub­sti­tuted (let the banks burn! don’t bail out GM!) and then chang­ing their tune as soon as the worst ac­tu­ally hap­pened.

But at least hav­ing the Fed­eral Re­serve pri­vately vi­su­al­iz­ing all sorts of hideous pos­si­bil­ities in ad­vance, war-gam­ing them with the Board of Gover­nors, and plan­ning for them re­al­is­ti­cally—so that when the worst starts hap­pen­ing, you don’t have ev­ery­one run­ning around be­ing vague and visi­bly un­pre­pared and re­fus­ing to talk about what hap­pens if things get even worse—in­stead you just take out folder #37-B and figure out what needs tweak­ing—for the lack of that pre­pared­ness, there seems to me to be very lit­tle ex­cuse. The Fed­eral Re­serve should not be in the busi­ness of fore­cast­ing prob­a­bil­ities—they’ve already demon­strated that they can’t, and they’re not in­vestors. They should just be always star­ing into the abyss.

Of course the Fed­eral Re­serve doesn’t read this blog, so far as I know. But it’s the sort of thing that doesn’t re­quire a ma­jor­ity vote for in­di­vi­d­u­als to use in their per­sonal lives.