Tuesday, October 6, 2009

Negative Feedback

It's not what you think.

If I said I was getting "positive feedback" you'd probably figure I heard from people who like my posts. If I said "negative feedback" you'd figure they didn't like 'em. Fair enough. That's probably what I'd mean. But y'know, if that was all there was to it, I wouldn't be writing this post.

Tack the word "loop" on to those feedback phrases, and the meanings are totally different. A positive feedback loop is a self-reinforcing loop. A negative feedback loop is self-negating. These are not at all the meanings we ordinarily use. We might think of a bad situation making itself worse as a negative thing. But it's a positive feedback loop.

Here's an old (July 2008) post from Paul McCulley of the investment management firm PIMCO. Here's a piece of it:

Once the double bubbles in housing valuation and housing debt burst a little over a year ago, everybody, and in particular, every levered financial institution – banks and shadow banks alike – decided individually that it was time to delever their balance sheets. At the individual level, that made perfect sense.

At the collective level, however, it has given us the paradox of deleveraging: when we all try to do it at the same time, we actually do less of it, because we collectively create deflation in the assets from which leverage is being removed. Put differently, not all levered lenders can shed assets and the associated debt at the same time without driving down asset prices, which has the paradoxical impact of increasing leverage by driving down lenders’ net worth.

This process is sometimes called, especially by Fed officials, a negative feedback loop. And it is, though I prefer calling it the paradox of deleveraging....

If it was a negative feedback loop, the problem would have results that tend to solve the problem. Mr. McCulley and the Fed officials must really mean to call it a positive feedback loop, a problem that makes itself worse.

First of all, if we want to avoid this kind of problem in the future, all we have to do is avoid excessive leverage. If leverage never increases beyond a reasonable level, the panic that leads to collective deleveraging need never occur.

Leverage is debt. "The use of debt to supplement investment," Wikipedia says. If debt never increases beyond a reasonable level, a sudden mass movement to reduce debt need never occur. (And if it does occur, the effect on the economy is less.)

And not to be picky, but people don't use debt. People use credit. Debt is just a way to keep track of credit-in-use. Is the distinction important? Well, yes, if we want to reduce our debt it is very important. Economic policy encourages the use of credit. Almost nobody objects to that. But we'll never successfully reduce our debt until we reduce our reliance on credit.

That's my standard argument. For me it always comes down to the reliance on credit. But that's not the reason for this post.

Articles like the one by Paul McCulley are not light reading. They demand attention. It's tough enough to make sense of them when they don't come with gross errors of meaning. Nobody's perfect. Still, there's no reason a writer should put less effort into an article than his readers put into it. But that's not the reason for this post, either.

I recently used the phrase positive feedback loop. But now I see that even smart people misuse the phrase. So I dwell on it. That's the reason for this post. That, and I think McCulley's observations are important and well said.

No comments: