Economists, mathematics, and where the semiconductor industry is headed this year

A snide person might observe that if circuit designers were equipped with the same formidable mathematical skills as economists, we would all be using single-ended triode amplifiers that oscillated when we turned the gain up. (As it is, only the very rich get to use these, but that’s another rant altogether.)

In my local newspaper this morning was an article reporting that the housing market was in "a tailspin" with no hope in sight. I suppose that the implication was that if you don’t own your home outright, you’d just as well pack up and move to an apartment now and waste no more money on mortgage payments.

Unfortunately for the reporter’s premise, he or she published along with the article a graph of housing prices vs. time. It showed a slightly-damped cosine wave—the sort of time-domain response you would expect if you hit a slightly under-damped system with a downward step function. Yep, housing prices continue to fall.

The thing that caught my attention was that the inflection point on the cosine wave occurred in about April. According to the data—as opposed to the text—the rate of descent has actually been decreasing for almost a full quarter. That suggests we could see a bottom to the housing-price decline, at least in terms of national averages, sometime in the third or early fourth quarter. This minor detail was completely lost on the author of the news story. It made me once again wish that economists, reporters, and others who directly influence public sentiment about the economy were required to pass at least one course in linear systems before they were allowed access to a keyboard.

This is not to say we are out of the woods yet, by any means. Other data suggest that, as one would expect given that consumer spending behavior tends to lag changes in disposable income by about two quarters, consumer spending in the US is dropping off rather sharply. There are some indications that it is doing so in Europe as well, and that the more affluent Chinese consumers may be starting into the same descent. But just how the new Chinese middle class will react to the evaporation of their stock market investments and the step function in petroleum and grain prices is hard to predict. These folks are fairly new to a market economy, their previous experiences with downturns having mostly been national catastrophes triggered by mad government policies. See, for instance, the Great Leap Forward.

This drop in spending will pretty much have to percolate back through the system to trigger adjustments in inventories and production levels in the electronics and semiconductor areas. But there is no sign of that yet. Analyst Malcom Penn’s most recent forecast at Future Horizons suggests that, if you take out the crippled DRAM sector of the semiconductor industry, we could actually see quite strong growth through the rest of the year. That could mean that we are simply not yet seeing the leading edge of the reaction to reduced consumer spending and industrial investment. Or it could mean that there are enough things still going well—including strategic investments in industries such as network infrastructure, defense, and health care—to propel the semiconductor industry through the trough. We just don’t have enough data, or comprehensive enough models. (Note to economists—regression models based on the previous ten years are not predictive tools. Please sit down.)

Certainly there are still a lot of negative issues. Capital restrictions are cutting into industrial spending because even companies simply can’t get affordable loans. Consumers, encouraged by the panic-stricken press, are spending conservatively. The balance sheets of the Western nations who have to intervene in some of these problems are in no shape for largess—particularly in the USA, where the current administration has devastated the books, and created an unprecedented—and unpredictable—link between US public debt and Chinese government policy.

But there are positive signs as well. As mentioned, the housing prices that underlie a lot of the credit-market problems may be near a local minimum. Stable housing prices would give both home-owners—even those with ludicrous mortgages—and lenders a lot more room to reduce defaults, and to minimize the impact of the inevitable defaults on the collateralized securities into which the mortgages have been pureed. That in turn should help the process of reestablishing markets for those creatures.

And the banks that have been most heavily invested in the mortgage-backed and loan-back instruments are well along in their process of clearing the things off their books. This will stabilize the banks, making them more willing to invest their capital in loans rather than hoarding it for their own survival. Finally, as the market for the creative securities solidifies, some hedge funds and other investors are going to make huge fortunes as the contracts they bought at five cents on the dollar (I’m not making that up, by the way) regain some semblance of a market value.

So are we out of the woods? Not yet. But are we diving into a black hole? Probably not, unless the press succeeds in creating general panic, or politicians find a way to turn a serious cyclic correction into a collapse. You can’t count either of those possibilities out yet, of course.