Recovery in sight for semiconductors—or, one hates to sound pessimistic ...

There has been a fair amount of discussion lately about the beginning of the economic recovery. Analysts have cited indicators from a number of sources indicating that the level of economic activity has at least stopped falling so rapidly, and in some cases has begun to rise. And in some corners of the semiconductor industry design starts and orders may actually be rising.

A number of parties have cited this data to declare the imminence of the recovery. That would be great, but that is only one way to interpret the data. Another, and it seems to me much more likely interpretation is that we are approaching a flat spot. Whether this flat spot is the critical-point for a V-shaped recession, the beginning of a long dead spell like Japan’s lost decade, or a shelf on the way down the cliff cannot be determined from this data. But I would offer some cautionary interpretation.

First, because the semiconductor industry happened to enter this catastrophe with almost historically low inventories, orders should be very sensitive to a slowing in the rate of freefall in end markets. OEMs working on very small finished-goods and work in progress have to respond quickly to any indication of increasing demand, and that response will ripple all the way back through an empty supply pipeline. This is true whether the OEMs actually have much visibility of future demand or not. So if in fact OEMs are starting to see orders for end-user goods that reflect the pent-up baseline demand, built up over the months when everyone was afraid to buy anything, that should send ripples of demand back up through the supply chain—just what we appear to be seeing now. But this increase is an artifact of inventory management, not a leading indicator.

A second point. What we are likely to see is a restructuring, not a recovery. It’s not just a matter of time until masses of North Americans and Western Europeans troop back to the store to resume buying huge-screen HDTVs, home-media networks, smart phones, and high-end sedans, putting the purchases on their credit cards and home-equity lines of credit. Nor are we likely to see stimulus money from governments in those areas recreating the great consumer binge. Those funds are going to other purposes, to fulfill political promises.

What we are likely to see is a very gradual return to replacement-level buying of durable goods, with much more conservative household and corporate spending and a lot more savings and debt-reduction. When something new and hot comes along, it will diffuse into the market the way new consumer goods did in the 70s or 80s, not with the explosive growth of an iPhone or an HDTV. There will of course be exceptions, but they will be exceptions. So the bottom line is not to expect a return to the kinds of revenue streams we saw in 2007.

Finally, some further caution. There are several large pieces of bad financial news out there that are yet to have an impact on the crisis. The rapid growth in delinquencies and defaults on unsecured consumer credit—credit cards—has yet to show up in the banks’ earnings projections, through a slight-of-hand apparently approved by the folks who brought you the Stress Tests. That bad news may force the banks to substantially increase their reserves yet again, putting stress on all the credit markets.

There is a similar problem with commercial credit—commercial real estate loans, operating loans to corporations, and corporate bonds. Companies that do not have substantial cash reserves may be clinging by their fingernails, and default rates are already rising. This, economists point out, will hit primarily not the big investment banks that are already in intensive care, but the regional and local banks that do the majority of the commercial lending. A crisis there could strangle commercial credit across such a wide front that there would be no easy way for the Fed to intervene. This problem may be most acute, by the way, not in the US but in Central Europe, where some West European banks have heavy exposure.

And there may well also be a sub-surface crisis in the insurance industry. The problems of AIG have been widely discussed, but AIG was rather special—it had a genius for exposing itself to horrendous risks. Much of the rest of the insurance industry, though, faces a more subtle problem. Insurance works by collecting premiums, investing them profitably, and using only a portion to pay claims. But with spreading unemployment and home losses, premium payments may be declining. It is likely that insurance companies have fared about as well on their investments as the rest of us, so they face shrinking reserves. And in areas of financial insurance in particular, the companies involved in these contracts face spiraling claims. There may be a shock coming there as well.

All of this of course doesn’t mention the further shock to employment as the US auto industry ratchets down to its new size, spewing order cancellations and job cuts through its supply chain. That one industry is big enough to have national impact, and the impact has been delayed so far by negotiation and procrastination, so we may get the hole bitter pill within a few quarters.

Finally, there is the lingering question of just how much the US Treasury can borrow before buyers decide that they have enough Treasury Bonds, and interest rates start to rise rapidly. I think no one really knows the answer to that question.

All of these shocks possibly can be absorbed by the system, bolstered as it is by historic levels of government spending. Or perhaps they can’t. I doubt that anyone has a sufficiently intact econometric model to make a sure statement on the question. If we can tolerate them all, at least spread out a bit, then we are likely to bounce along the bottom of this recession until the shocks have worked their way through the system. If we cannot, then we may be rolling across that aforementioned ledge. In any case, recovery is far from a sure bet this year, and even restructuring may be starting later and taking longer than we might like.

What does this mean for us in semiconductors? We should plan on continued base-level activity in the industry for the foreseeable future. Growth is likely to come only directly from Federal stimulus programs, not from growing demand or new markets. And until we are past a few known problems, the risks are on the downside, not the upside. It’s a time to be aware and agile, but not overconfident.