Financial Crisis: How to Value Toxic Assets?

20 February 2009

Auction sign on foreclosed house in USANeil McCulloch - 16 February 2009

The key to solving the financial crisis, which is affecting both developed and developing countries, is restoring faith in the financial system. Yet that faith is undermined by the existence of billions of dollars of toxic assets on bank balance sheets across the world. But the US Treasury Secretary’s plan to tackle the problem is faltering because nobody knows how to value the toxic assets held by banks.

The market for 'lemons'

Joe Stiglitz, George Akerlof, and Michael Spence must have a wry smile on their face these days. Their Nobel Prize winning work on asymmetric information and adverse selection is suddenly at the centre of the argument about how to solve the current global financial crisis. We are witnessing the largest 'market for lemons' that the world has ever seen. 

Akerloff’s The Market for Lemons paper is the world’s most famous explanation of the problem of 'adverse selection'. He considers the market for second hand cars. Say there are two types of cars, good ones and bad ones (‘lemons’); the good ones are worth $1000, the bad ones are worthless. Now say the seller knows whether the car is good or bad, but the buyer does not. In this case the buyer will offer less than $1000, because of the risk that the car might be a ’lemon’. But the sellers of good cars won’t take less than $1000, so they will withdraw from the market, leaving only the bad cars. Knowing that the market will only consist of lemons, the buyers won’t buy at all and the market collapses entirely. 

Exactly the same thing may be happening in the market for toxic assets. Banks want to offload their toxic assets. But they have better information than the market does about how toxic their assets are. The markets therefore discount heavily what the banks offer for sale, so heavily that the Banks don't offer good assets for sale. But knowing this, and not wanting to buy worthless assets, the market withdraws. It is therefore not surprising that Tim Geithner is finding it hard to get people to buy the assets that the banks are offering.

How to signal quality

The traditional solution to adverse selection is 'signalling'. Michael Spence got his share of the Nobel prize for pointing out that sellers often go to great lengths to ‘signal’ their quality to buyers to get around the adverse selection problem. For example, second hand car dealers sometimes offer warranties as a signal that the car is not a lemon. Unfortunately that solution is harder to apply with toxic assets because it is much harder for the buyer to know, even after purchase, whether they got a lemon or not. If you buy a bad car, you know pretty quickly. But very few buyers keep financial assets to maturity to see whether they really pay off. This means that, for practical purposes, the value of financial assets are determined in the market, the same market that doesn’t know what they are worth. 

If the asset is sufficiently uniform such that nearly identical assets are traded routinely in the market, then this need not be a major problem. Banks who wish to sell assets can offer them with market price based insurance. For example, they can offer the assets for a certain price, but with the option to sell it back to the bank (or the underwriter) should the market price fall below a specified level. By setting a relatively high option price, the banks can signal the quality of the assets. 

Unfortunately, many of the complex securities that banks need to sell are sufficiently unique that there is no market price for them. Providing insurance is impossible in these cases because there is no market price for the asset’s value to fall below. Yet nobody will want to buy these assets because they don’t know their true value and they know that they will therefore be difficult to sell. In this situation government action is needed, both to solve the adverse selection problem and to provide insurance for buyers.

Could government action help?

The adverse selection problem can be solved by not allowing banks to determine which assets they sell. If the banks choose they will deliberately sell the assets that look better, but which they know to be of worse quality. But governments could force banks to accept independent audits of their asset quality. Such audits would rank assets by quality, based upon observable characteristics and then force banks to sell a significant share of the lower ranked assets. By ensuring that banks cannot stack the sale with good looking but poor quality assets, buyers will have more confidence in purchasing the assets. 

Solving adverse selection may not be enough. Buyers are also, understandably, risk averse. Yet neither the banks nor government can provide insurance in the traditional sense, since the final value of the asset cannot be observed. However, government can provide a risk premium to buyers. To minimise the cost and to identify the level of premium necessary to stimulate purchase, the government can simultaneously auction both the asset and the premium. Bidders that offer a high price for the asset along with a low premium from government will win the asset (the government can specify the indifference curves between the purchase price and the premium to reflect its own risk aversion for high premiums). In this way the government can induce buyers to take poor quality assets off the banks for the true market valuation of the asset, whilst minimising the cost to the government and ensuring that the buyers receive only the premium necessary to induce the purchase.

A first step to restoring confidence

The key downside of this approach is the need to independently audit the asset quality of the banks. But this may not be necessary. If, as some bankers have claimed, the banks do not actually have any additional knowledge about the true valuation of their assets, then there is no adverse selection problem and the market is merely failing because of the risk aversion of the buyers. In this case, the banks themselves can put together the ranking of their assets with the government only having to facilitate the simultaneous price/premium auctions. Indeed the extent to which the market failure results from adverse selection or purchaser risk aversion can be assessed by doing both kinds of auction and seeing whether demand is higher when the quality of the assets being auctioned has been independently assessed. Either way, the price/premium auctions could provide an important first step in restoring confidence in the banking sector. 

Neil McCulloch is a Fellow in the Globalisation Team at Institute of Development Studies, Sussex, UK

Image: Piotr Malecki / Panos