tag:blogger.com,1999:blog-2499715909956774229.post3419716590159916038..comments2024-03-22T22:37:02.639-07:00Comments on Stephen Williamson: New Monetarist Economics: Gary Gorton InterviewStephen Williamsonhttp://www.blogger.com/profile/01434465858419028592noreply@blogger.comBlogger5125tag:blogger.com,1999:blog-2499715909956774229.post-50874244369407228122010-12-31T08:00:46.800-08:002010-12-31T08:00:46.800-08:00anon1:
Thanks. The point from your last paragraph...anon1:<br /><br />Thanks. The point from your last paragraph is particularly interesting.Stephen Williamsonhttps://www.blogger.com/profile/01434465858419028592noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-13832400398128745882010-12-30T22:37:57.214-08:002010-12-30T22:37:57.214-08:00Nick Rowe argues that "a big enough shock&quo...Nick Rowe argues that "a big enough shock" can cause the "testing" equilibrium to occur. This overlooks a point Gorton makes in the interview: it was not a big shock at all, but the flattening out of home prices, that resulted in shadow bank liabilities becoming information-sensitive. IMO, Gorton is saying that the "amplification" construct is not useful when one is talking about a very small shock resulting in a large liquidity event. <br /><br />Reading Gorton's work, it is clear he focuses on the structure of shadow bank liabilities: it was the dominance of AAA-rated ABS collateral backing shadow bank "deposits" (ABCP and repo's) that ultimately led to a run on those deposits. This unique initial condition is what made the system vulnerable to a system-wide run, not an "amplification" mechanism ("testing") potentially generic to any recession. <br /><br />One implication of Gorton's work is that Fed AD stabilization can result in lower and lower requirements for collateral backing liquid deposits. The more aggressively stabilized the system, the less liquidity insurance taken out by actors (margin in the form of collateral haircuts), and the more vulnerable the system ultimately becomes to even a small shock.Anon1noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-19894797248488766332010-12-30T07:52:10.476-08:002010-12-30T07:52:10.476-08:00Nick,
1. Gary seems to think he is talking about ...Nick,<br /><br />1. Gary seems to think he is talking about something out of the realm of conventional theory, but it certainly sounds like a sunspot equilibrium, which we know a lot about. I think that is what you are sketching out as well.<br /><br />2. We do have models of debt, based on Townsend's (1979) costly state verification setup. What that gives you, given some assumptions about commitment and pure strategies, is non-contingent debt. The fixed payments are real, though (i.e. think of them as indexed). No one has a model of nominal debt contracts, but Townsend's idea is quite simple and compelling. Non-contingent payments allow you to efficiently economize on verification costs, which can be broadly interpreted as costs of extracting payment.Stephen Williamsonhttps://www.blogger.com/profile/01434465858419028592noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-14251929867164178872010-12-30T04:22:52.918-08:002010-12-30T04:22:52.918-08:00GG: "That’s why I said before that you need a...GG: "That’s why I said before that you need a theory of debt; you need to explain why there’s this debt and what is the purpose of having this debt."<br /><br />That's what's so hard. Why would two people agree to a loan which isn't indexed to anything? Not even the CPI.<br /><br />And simply assuming there's a "cost to indexation" just doesn't cut it for me. But I can't think of anything better.<br /><br />It's easier to talk about the externalities of debt than to explain why debt exists.Nick Rowehttps://www.blogger.com/profile/04982579343160429422noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-14873190285423763762010-12-30T04:09:07.597-08:002010-12-30T04:09:07.597-08:00GG: "This notion of a kind of regime switch, ...GG: "This notion of a kind of regime switch, which happens when you go from debt that is information-insensitive to information-sensitive is different conceptually than an amplification mechanism."<br /><br />A sketch of a model:<br /><br />Some fraction of the assets are lemons. It is costly to test which one is or is not a lemon. Your incentive to test depends on whether the person selling you the asset tested, and whether the person to whom you will sell it will test. (You don't want to get stuck buying a lemon at the average price and having to sell it at the lemon price).<br /><br />So you get two equilibria: in the first equilibrium nobody tests; in the second equilibrium everybody tests. A big enough shock, even if temporary, could flip the economy from one equilibrium to the other. And the asset is much less liquid in the second equilibrium (since there's a fixed cost of testing before you buy).<br /><br />I think that's an amplification mechanism, because of the strategic complementarity in testing.Nick Rowehttps://www.blogger.com/profile/04982579343160429422noreply@blogger.com