collateral constraints

In the past this blog focused on my experience in Israel and on Israeli politics. Yonah has wittily and prolifically taken up that thread, so I leave it to him.

I turn now to economics. I'm a fan of the econ blog Marginal Revolution, but that cannot be my model if I hope to finish my PhD on time. Instead I'll use the blog as a commitment device.

Macro and public finance are my fields. As I read articles, I'll distill them into a brief nontechnical summary. I won't do this for all the articles I read, but blogging on a schedule will hopefully force me to keep up a brisk pace and be a good exercise in translating jargon into English. My plan is three summaries per week, posted Monday Wednesday and Friday at noon. Hat tip to Yonah for the schedule structure.

The first paper* is about borrowing limits. Think of a farmer who must borrow money to buy seed and machinery. The bank lending money to the farmer worries that the farmer will not have enough money to repay the loan if there is a bad crop. The bank requires the farmer to pledge farmland as collateral, which the bank will own if the farmer cannot repay. The limit on the farmer's borrowing is the value of the land he has available to pledge as collateral.

The paper considers the case where if one farmer has a bad crop, they all have a bad crop. Think of a bank lending to farmers in one region that is sometimes hit by a flood or drought. Then the bank will own all of the farmland pledged by all of the farmers, because none were able to repay. Things get complicated because the price of the farmland depends on whether there is a good crop or bad crop. In particular, following a bad crop, the price of farmland is low because the bank would like to sell lots of the land it now owns. Knowing that the value of the collateral will drop if it is ever collected, the bank is unwilling to lend very much to the farmers.

Krishnamurthy observes that the farmers could mitigate this problem by buying insurance. If there is no flood, they give some of the bumper crop to the bank. If there is a flood, the bank not only allows them to keep the farmland but also gives them enough money for new equipment and seed.

That seems reasonable enough. But remember that when one crop goes bad they all go bad. That means that when there is a flood, the bank will have to shell out insurance payments to everybody at the same time. How do the farmers know the bank will be able to pay? The same way the bank guaranteed the farmers could pay: by demanding collateral. And now the tables have turned, because the bank might not have enough cash on hand to provide a sufficient amount of insurance to all the farmers in the event of a flood.

Why don't firms fully insure against risk? Krishnamurthy shows that one possible answer to this question is that insurers don't have enough collateral. Krishnamurthy notes that if insurance markets are failing for this reason, then it could be a justification for the government to be more active in providing insurance.

*Krishnamurthy JET 2003 Collateral constraints and the amplification mechanism