Wall Street bets its chips on fantasy

By Bill Fleckenstein

Let's begin with a perfect example of what passes for knowledge and information on Wall Street: A dead fish from Morgan Stanley (MS, news, msgs) recently penned a "research" report that's right up there with specimens from the dot-com days, when folks were talking about valuing eyeballs.

Despite his acknowledgment -- "It's difficult to quantify the exact (semiconductor) equipment exposure, given the complexity of the wafer allocation and technological mix at the foundries where most of the iPhone chips are being manufactured" -- he proceeded to recommend a handful of equipment stocks as "plays" on Apple's (AAPL, news, msgs) new iPhone.

Specifically, he said, "Applied Materials (AMAT, news, msgs), ASML Holding (ASML, news, msgs) and Lam Research (LRCX, news, msgs) are best exposed to iPhone semiconductor content on a relative basis."

Beyond that caveat, he failed to address the size and scope of the iPhone: Maybe 1 million units will be produced this quarter, and Apple says it will sell around 10 million in the next 18 months. In that time, there will probably be sales of 1.5 billion cell phones and 400 million personal computers. So it's laughable to think that the addition of a puny 10 million iPhones will move the needle at any foundry anywhere -- but again, those details don't make it into Wall Street's version of "analysis."

Bottom callers batting zero

Nor do the facts stop semiconductor dead fish (and the managements of these companies) from issuing optimistic forecasts. For a year now, they've been predicting margin expansion and a new up cycle, when all we continue to see is inventory building and demand sputtering.

It's certainly hard to support that view given recent tepid guidance from Texas Instruments (TXN, news, msgs), Fairchild Semiconductor International (FCS, news, msgs), Analog Devices (ADI, news, msgs), Altera (ALTR, news, msgs), Xilinx (XLNX, news, msgs), Molex (MOLX, news, msgs) and Microchip Technology (MCHP, news, msgs). To me, this suggests that the so-called, and continuously proclaimed, bottom is really a prolonged top.

The fact that Microchip was trading at an all-time high just a few days before it announced lowered guidance illustrates a point I have made many times recently: The stock market currently discounts nothing. Of course, had the analysts been researching the fundamentals, perhaps the companies' problems would not have come as news to them.

Now let's turn to another arena that has caught the Wall Street community by surprise: the escalating troubles in mortgage-related assets. A case in point is Bear Stearns' (BSC, news, msgs) HGSCSEL fund (High-Grade Structured Credit Strategies Enhanced Leverage), which is enduring a rather painful unraveling. (Editor's note: At the end of March, according to The Wall Street Journal, the fund had piled more than $6 billion in borrowing on top of $638 million in investor capital.)

The onion is peeled ...

Though rescue attempts for the fund were planned, they appear to have been abandoned, as the would-be knights have realized that many of the fund's collateralized debt obligations (CDOs) -- really, CDOs squared -- are not worth what was thought. That's because of the silly way of valuing them known as mark to model, which potentially overstates the value of slices of the mortgage market.

In any case, the abandonment of that pricing model is important because, if reality intrudes and changes the mark-to-model world, it will likely precipitate further selling. And that will feed on itself.

Meanwhile, there continues to be plenty of behind-the-scenes deterioration in mortgage pools, as noted by a friend who last fall and winter was so accurate about the subprime-business debacle. And my friend in the dark-matter universe says there is pressure on the commercial version of the ABX stack, a set of assets backed primarily by residential-real-estate loans.

It seems the financial termites are working full throttle. That, combined with a weakened economy, means it's just a matter of time before all this bursts into the spotlight. As Pimco Vice President Mark Kiesel told Bloomberg News: "We're talking about a two- to three-year downturn that will take a whole host of characters with it, from job creation to consumer confidence. Eventually, it will take the stock market and corporate profit."

... and all is revealed

The housing bubble and the housing ATM were built on fantasy. They were driven by Wall Street's ability to take any and all mortgages, package them up and turn them essentially into high-grade securities.

It was like the medieval notion of spinning straw into gold: Wall Street alchemists wanted volume, and they got it -- assembling trillions of dollars' worth of mortgage paper that probably should never have been created, certainly not under the terms as structured. One day when we look back on this period, I believe we'll shake our heads and say it was all so obvious. Then we'll ask: What was the stock market "thinking" when it viewed itself as an entity that reality couldn't touch?