Mannerisms

"All our share prices are below where they were ten years ago", says Brian Halla, CEO of National, "P/E ratios are in the teens when historically they are 30-50."

In days of yore, whenever a semiconductor company felt its share price was too low it would announce an important breakthrough in its R&D department.

The great thing about this was that no one expected an R&D advance to translate into sales anytime soon, while the analysts who influence share prices couldn't understand the significance of the announcement.

But analysts, like economists, hunt as a pack, and if one could be persuaded that the announced 'breakthrough' was a big deal, then the others dropped into line. The shares got recommended. Their prices went up.

Unfortunately, things have changed. For one thing, the analysts nowadays know too much about semiconductor companies. In the days when making ICs was a black art, chip companies could easily pull the wool over the eyes of those who analysed them.

Now, however, like a tired old stripper, the industry has revealed too much of itself to be able to intrigue its watchers.

Another thing is, the big companies have stopped doing basic R&D so they don't have any breakthroughs to announce. If the only R&D breakthroughs that are going to happen are coming from universities and government-funded laboratories, then no one company will be able to claim that they have the sole right to exploit them.

While they have no new technology to call their own which they can exploit.

So the analysts threaten to ruin the industry. Having persuaded CEOs that all that matters is revenue growth, margins and ROI, the CEOs have obediently cut off the one thing that can keep their product lines refreshed, interesting and super-competitive - a flow of proprietary technology breakthroughs from vibrant R&D departments.

So c'mon you CEOs, before it's too late, start doing some R&D to add value to your boring commoditised CMOS structures.

And then these lousy share prices and p/e ratios might start perking up.