Arnold Kling  

Confidence and the Wrong Map

PRINT
Does Gravity Kill?... EconLog Book Club: For a Ne...

Over at econtalk, Russ Roberts interviews Daron Acemoglu. Self-recommending, as Tyler would say.

Also, here is a video featuring Daniel Kahneman and Nassim Taleb. Taleb, like me, wants to get rid of risk-taking by banks, and leave non-insured institutions free to take whatever risks they want, as long as they are not creating risks for others. His solution is to nationalize banks. (me: why would this mean that they would not take risks? Suppose that Freddie Mac and Fannie Mae had been fully nationalized as of three years ago. Would they have taken more risk or less risk?)

Kahneman tells a story of Swiss army men who got lost in blizzard in the Alps. When they finally find their way back, they are asked, "How did you find your way?" They say, "We had a map." But the map is of the Pyrenees! Kahneman's point is that people have a lot more confidence if they have maps, even if those maps are wrong.

Taleb does not explicitly dispute this point, but he clearly does not like the notion. It would imply that you want a doctor to think he is correct, even if he is wrong, because the doctor is giving you a map. Taleb says that religion succeeded because when people had faith, they stayed away from doctors, who were wrong. (He makes those sorts of comments a lot, as you know.) He wants business schools to stop giving students the "map" of financial modeling, because he thinks those models are wrong.

In my Lost History of Macroeconometrics, I say that macroeconometrics tries to satisfy the need for a map, but it does not provide a reliable map. But I am in the same position as Taleb. People want to believe in a map, so telling them they have the wrong map is not going to get me very far.

Finally, a reminder that on Tuesday, February 10, I will be speaking here. If they don't record it, I'll try to remember what I said. (I try not to use notes or a text. Instead, I prepare by practicing while I walk. These days, if you're talking to yourself while walking, people just assume you are on the phone, so they don't even look at you funny.)


Comments and Sharing





COMMENTS (11 to date)
Don the libertarian Democrat writes:

"Taleb, like me, wants to get rid of risk-taking by banks, and leave non-insured institutions free to take whatever risks they want, as long as they are not creating risks for others. His solution is to nationalize banks. (me: why would this mean that they would not take risks? Suppose that Freddie Mac and Fannie Mae had been fully nationalized as of three years ago. Would they have taken more risk or less risk?)"

Although I'd like to nationalize a few banks in this mess, I agree with you. We don't need to run them, especially if we have narrow/limited purpose banks. I didn't like this idea at first, but if it allows the existence of risk-taking non-insured institutions, then I'd be for it.

happyjuggler0 writes:

I'm a pretty hardcore minarchist libertarian. However I've gotten to the point in this massive fiasco that I think that from the perspective of the politically possible, that nationalizing the banks that are in a quagmire is the least bad option.

Just take over them as if in a bankruptcy. Strip out everything but the operating asset which has value as an ongoing concern, fire the boneheads at the top who screwed up, replace them with folks who seem bright enough, create a similar board of bright enough people, and take it public (i.e. issue an IPO).

Keep the liabilities at first. Hire an investment bank(er) to dispose of (i.e. sell off) the damn "assets", both "toxic" and nontoxic. Pay off the creditors with the proceeds, and if there is anything left after that, then the shareholders get some pennies back!

Quite simply, we need a functional banking system again, one that everyone has confidence in, and which is well capitalized. We aren't going to get that if we commit the Japan blunder or the FDR blunder.

Speaking of which, anyone who thinks that Friedman et al was wrong, and that the depression didn't end for monetary reasons but rather because of WWII, is conceding that FDR's programs sucked and failed to end the Depression.

spencer writes:

How can you have a bank that does not take risk?

That is what they do.

What would it do?

Take deposits -- no that has risk.

Make loans -- no that has risk.

Invest for its own account -- not likely.

Grant Gould writes:

spencer has it very right: The very act of loaning money, or of making long term commitments from money from demand deposits, entails risk. And it's not clear that there's a bit qualitative difference between the risk of loaning money for a mortgage versus loaning money for a business, or trading stocks on margin, or what-have you.

Politically, the "risk-free bank" is going to make a compelling political case that loaning lots of money at low interest rates is a social goal of the banking system, and its regulators are going to listen. If (per Kling) these banks' executives stick to the spirit of the law, they will be replaced by their political masters with executives who will run to the edge of the law in risk-taking and then plead "only following orders" when it blows up.

A "risk-free bank" gets that way by guarantees. Those guarantees are a risk taken by the guarantor, who must price it correctly. If it is underpriced, the non-insured banking sector is outcompeted and vanishes; if it is overpriced, business fleeing to the non-insured sector will prompt political intervention, crisis, or both. It's a bad mix.

JoshK writes:

IMHO, we should just get rid of the FDIC.

I think people would then use fee-based cash management accounts for day-to-day money handling and invest in funds for yield. The funds can have structures to meet consumer demand and offer different risk profiles and minimum holding periods.

JoshK writes:

And one more point. How come Shelia Blair comes out of this a hero? If she had differentiated between banks and charging risk-based fees we would not be here. She basically whored out the FDIC and gave away the FDIC insurance for free.

Yancey Ward writes:

There is just one problem with Taleb's suggestion-it is stupid and naive. Ok, that is two problems.

Nationalized banks will still be making loans and will, thus, be risky. They will be making these loans for different reasons (politically connected borrowers etc.), and these loans will eventually go bad for different reasons, but go bad they will.

The unnationalized system will still find politicians willing to subsidize and insure risk for the "social good".

Walt French writes:

IMHO, we should just get rid of the FDIC.

Weren't those really FUN and exciting years when the first whiff of rotting meat brought out all the depositors to make a run on the bank, crashing the institution, the depositors' assets, AND the little community in which the bank operated! Bring 'em on, as the experts say! Who cares about setting up an environment where it's impossible to honor contracts?

What? You never read about that in your economic history classes?

I like going the other way: require that any money market fund, which (even implicitly) guarantees not to break the buck, be regulated as a bank, never to stuff their portfolios full of Auction Rate Notes and their ilk, unless insured by the Lender of Last Resort. Absolutely no complaint about "short term paper" funds that buy the stuff, but NO GUARANTEES plastered all over 'em unless they're FDIC-regulated.

Economists know about the necessary and sufficient conditions to prevent runs. Eliminating the FDIC goes 180 degrees from sensible.

That scary story about MM depositors making a run on the banks (e.g., State Street) last September, taking out $550 bil (can that really be true?), nearly collapsing the US banking system, and the world's with it another few hours later, is a roller coaster that I'll happily strap JoshK onto, but for me, no thanks.

Tyler L writes:

I'm gonna agree with Spencer and say that developing a risk free banking system is impossible; Taleb and Kahneman are barking up the wrong tree. If banking institutions become non-insured and allowed to take whatever risks they want people are gonna invest in their mattresses again. And as far as using a "wrong map" for false confidence that sounds like a foundation built for crumbling.

Mike Rulle writes:

"People want to believe in a map, so telling them they have the wrong map is not going to get me very far."

We are far more "self aware" as a species or a society than we were just a few years ago. One of the recurring themes of this economic downturn is "psychology" and "confidence". People may want a map or believe in placebos, but not if they know the map you are giving them is wrong to begin with. Today every one believes that we are being handed maps the makers know are false.

Ed writes:

It's more likely that a Federally controlled bank would be pressured to make loans to anyone who asks, regardless of ability to repay. Mortgages for all. Yipppieeee.

Comments for this entry have been closed
Return to top