Destroying Lemons (or at least locking them up in a "Bad Bank")

In a previous post I built a toy model to explain how some assets going bad could reduce market liquidity and reduce the total value of assets by a multiple of the original loss in value. My model was of course based on Akerlof's famous "Market for Lemons", but I introduced a demand for liquidity into that model (the chance that people need to sell assets quickly).

When there is a demand for liquidity, the lemons problem does more than just make markets thinner, and reduce market prices (because people only trade lemons, unless they really have to). By making all assets less liquid, it makes them all less valuable, in a fundamental sense. Owning a bond not only pays you interest, it also gives you a means to raise cash quickly, if you discover you need to. Owning a car not only pays you transportation dividends, it also gives you a means to raise cash by selling it. But cars are not so liquid as bonds, and so the required rate of return on cars must be higher than the required rate of return on bonds, for that very reason. That higher required rate of return translates into a lower asset value. If cars were as liquid as bonds, they would be more valuable, and more people would want to own a car.

When some assets go bad (as has happened recently), and if many people can't easily distinguish good assets from lemons (which seems plausible), and if there is a demand for liquidity (I am confident about the existence of that demand, but less confident about its magnitude), then all assets become less liquid and less valuable. This might explain part, perhaps a large part, of the recent decline in asset values.

So what are the policy implications? (Anyone who has ever sat through an Economics PhD defence will recognise this as a stock question.)

I had another look at my toy model in that previous post, and came to a startling (to me) conclusion: the country would be better off if the lemons were destroyed. And that's true even though the lemons (though not as valuable as the good cars) have a positive fundamental value (maintenance costs for lemons are still only half the bus fare). The fundamental value of the 100 cars in the "lemons equilibrium" is $7,000. If the 20 lemons were destroyed, the remaining 80 good cars would now become perfectly liquid assets, and the stock of cars would now be worth $8,000.

If the government were to identify and destroy the lemons, that would be a good policy. An even better policy would be for the government to identify the lemons, indelibly brand them as such, and put them to use. The liquidity problem would be solved either way, and by putting indelibly branded lemons to some use, the total value of the 100 cars would rise to $9,000.

Of course, this assumes that the government can somehow identify lemons, when ordinary car buyers can't. But this problem can be solved, in one of two ways.

The first way, is for the government to announce that it will buy 20 cars, in a reverse auction, after the annual car market has closed. The equilibrium outcome is that all the lemons will be offered for sale, at a price of $95. To prove this is an equilibrium, note that with all lemons off the market, the remaining good cars will be worth $100 each, while a lemon-owner who deviated from equilibrium and held back his lemon would have a car he could sell for $100 at the next year's market, but would pay $5 maintenance expenses. Even if the government destroyed all 20 cars, the fundamental value of national wealth would rise by $1,000.

The second way is a more gradualist variant on the first way. The government announces that it will buy one car per year, in the annual car market, at the market price. The advantage is that it will cost the government less money to do this, since the price will not rise to $100 immediately. The disadvantages are: though the government would be buying disproportionately lemons, it would also be buying some good cars; and it would take longer to resolve the lemons problem.

Liquidity is impaired because the assets offered for sale on the market are disproportionately lemons (the percentage of lemons on the market is higher than the proportion of lemons in the population). By buying in the market, the government reduces the proportion of lemons in the population, reduces the proportion of lemons in subsequent markets, and thus increases liquidity, making all assets more valuable.

Depending on the numbers, the total value of assets might be increased even if the government destroyed the assets it bought. But better yet would be to put those assets in a "bad bank", and keep them off the market (the government rents out the cars, promising to pay any maintenance needed, but never sells them).

Perhaps there was a "rush to judgement" over the original version of TARP. (Understandable really, given how quickly events unfolded, and the absence of any clear explanation of what it was supposed to do). Sweden set up a "bad bank" for the bad assets. So did the US in the S&L crisis. Japan didn't, and that is instructive. It's less about improving bank capital ratios than about improving asset liquidity and fundamental values. Perhaps it's even time for the Canadian government and Bank of Canada to start buying garbage and less of the good stuff, even if they do overpay. By buying only the good stuff they lower the fundamental value of the remainder, both good and bad.

Comments

[Chris S]

Your example brings to mind Mazda and the 4703 potential "lemons"...

http://online.wsj.com/article/SB120942873506551291.html

They came to the same conclusion as you did.

[Don the libertarian Democrat]

The problem with the government buying CDOs was that there was never a way to give a convincing argument that the government wouldn't end up losing a lot of money on the deal, when all was said and done. There was no stomach for telling taxpayers that they have to overpay for poor investments, bailing out buffoons.

The only way the purchasing of CDOs could work would be if you could guarantee that the government was getting a good deal by severely underpaying for these assets.

Now, you might well argue that TARP as it stands now is a bust, and could end up costing the taxpayers a lot of money. But the money to the banks was sold as a purchase of equity. That seems, on a first look, a better investment than CDOs.

[Don Lloyd]

The lemon car story is a myth.

"Therefore, owners of good cars will not place their cars on the used car market"

This is absurd. It fails to take into account the fact that economic value is individual and subjective. Prices are not measures of value, certainly not subjective use-value.

If you buy a new car, the value of your previous car, and your willingness to sell it, is primarily the result of the law of diminishing marginal utility. The new car now provides the services that previously were the responsibility the old car. As a result, the value to you of the old car is greatly diminished no matter how 'good' you may think it is. Your old car may well be very valuable to someone else who doesn't have any car at all, but the difference between the importance of the old car to him and to you means that, in the end, the potential buyer will start with a substantial consumer surplus which can help insure against the risk of problems that neither the buyer or the seller can accurately forecast.

Regards, Don

[Phillip Huggan]

Driving cars over a cliff is way different from paying people to cut brake lines. If government subsidizes lemons people will produce more K-cars. This is literally what the big-3 execs want with their bailout request.
I think your advice is the same as are recommendations for accounting transparency with the added caveat that instead of decreeing new regulations, companies have the option of instead accepting a Crown buyout.
The entire investment industry is founded on secrecy. Employees make more money if losses are hidden, so expect these players to fight such regulations tooth and nail. An honest broker industry is 1/2 the size of an industry funded by commissions and options. I agree Nick, send them to Labor Ready. A bank is residential mortgages, business loans and financial services. The latter is poisoned in USA and small business loan rates are now unnecessarily high. A difficulty might be regulators might not be skilled enough to verify if "transparent" balance sheets are fudged by CFOs.

[Andrew]

Phillip, I'm not sure how much of a subsidy it is: presumably the reverse auction would not yield prices significantly different than the actual perceived value of the asset on the part of the seller. Since these assets will sell for less than they cost to produce, I'm not sure it leads to production of poor assets--it might dull the vigilance in ensuring the quality of production (whether cars or financial instruments). It's not a money pump, so firms will still try to minimize the production of lemons, although minimize given that the penalty of producing lemons may not be as severe.

[Nick Rowe]

Chris S: I had forgotten all about that news story. It illustrates my model exactly (even down to the asset being cars!). Presumably Mazda decided that the total value of the cars would be higher if it destroyed the lemons.

Don the LD: sometimes things are bad for the taxpayers but good for the citizens. Actually, if that weren't true, taxes would (or ought to be) zero! But yes, politically it might be a hard sell.

Don Lloyd: I'm not following you. In my model, the values were use-values (the amount a rational person would be willing to pay for a car, given a 10% rate of subjective time preference, the $10 bus fare, plus my assumption that they are subjectively indifferent between riding the bus and driving a car). I am assuming that everyone has the same use-values (preferences). And I (implicitly) assumed the marginal utility of a second car is zero. Fully Austro-compatible, I think. I just didn't use the terminology.

Phillip and Andrew: I dunno. I ducked the whole moral hazard question. Still trying to get my head around liquidity.