Friday, October 10, 2008

Fun with feedback

The last three weeks of financial panic have been traumatic for many who have investments in stocks as stock prices have lost about a quarter of their value on top of significant losses earlier in the year. During those three weeks, Obama has taken a clear lead over McCain in the national polls and on intrade. The common interpretation of that change is that the fear caused by the financial panic has made people more willing to toss out the incumbent party. Somewhere I saw a poll that said that McCain was losing support among older votes, where previously he had been stronger than Obama, and that result seems to me to consistent with a cause-effect relationship from stock market to political preference.

The folks at National Review has suggested another hypothesis, that the realization that Obama is increasingly likely to win has caused people to reassess the future of the economy and that reassessment has led to weakness in the stock market. I suppose the poll that shows 74% of CEOs saying that "they fear that an Obama presidency would be disastrous for the country" would support that position.

Both hypotheses could be correct. I think the idea of feedback is one of the most neglected ideas in economics. With feedback, causation does not just run in one way, but it runs both ways. Feedback can be either stabilizing or destabilizing. In this case I hope that the feedback hypothesis is not correct because it leads to predictions about my retirement portfolio that I do not want to contemplate. It would say that a decline in the stock market makes the election of Obama more likely. As the election of Obama becomes more likely, people pull out of stock, causing the market to fall further, and the cycle continues. I am hoping that in the next few days we will see this and the second hypothesis above refuted by the market.

There are some interesting feedback loops that are playing a role in the current financial panic. One involves expectations. As the market drops, people become more pessimistic about the future, and get out of the market. That in turn causes the market to drop further and causes even more pessimism. Another loop involves leverage, which is buying stocks with borrowed funds. When the stock market drops, some firms must sell because they get margin calls. Their selling depresses the market further, causing further margin calls. According to CNBC, this loop was active today.

There is also a stabilizing loop involved. As the market drops, the stocks become more and more attractive based on their fundamental value as investments. Hence, value buyers are more and more likely to buy them as their price declines. Eventually this loop will overwhelm the destabilizing loops mentioned above and will cause the stock prices to bottom out and begin to rise again. Greed will finally overwhelm fear.

Finally, government policy can be part of a feedback loop that will either stabilize or destabilize a crisis. Monetary policy during the 1929-1933 period has been criticized as being destabilizing because of something called the "real-bills doctrine." In recent years some economists have argued that many of the policies of the New Deal era prevented the stabilizing feedback of markets from working. The depressed spending should have lowered prices and wages, and the lower prices and wages should have gotten the economy back to normal employment. Those policies of the New Deal that attacked symptoms (low prices) rather than causes (insufficient nominal spending) interfered with this normal stabilizing feedback. Until at least the 1960s, the view that government policy was the cure and not a cause of the Depression was considered so obvious that it needed no justification. Today in economics anyone who suggests that government policy was the cure and not a primary reason for the Great Depression is considered fringe. (However, just about all monetary economists think that Roosevelt's gold policy was the key to stopping the fall and allowing the economy to turn around; not all policy was counterproductive.)

We cannot yet evaluate to what extent government policies have or will stabilize or destabilize the current situation. I am sure there will be economists researching the topic for many years in the future.

No comments: