No Reason for Nasdaq Exuberance

About a month ago, I was reading a "man on the street" article in The Wall Street Journal that quoted one investor as saying, "I'm always interested in stocks if they are moving, and the tech sector has been moving."

If that quote doesn't smack of a return to late-1990s la-la land, I don't know what does. While there's no question that the underlying business prospects for most tech companies have improved somewhat, the 60% surge in the Nasdaq over the past year implies that investors are assuming things have improved enormously. Unfortunately, the data simply don't support this assumption.

Consider that a recent Goldman Sachs survey on information-technology spending showed that IT budgets for the companies surveyed are likely to be roughly flat with 2002 budgets. Projections for 2004 came in at a whopping 4% increase in technology capital spending over 2003.

We’re hearing similar things from the 85 or so tech and telecom companies that we cover.  Verizon Communications (VZ) cut its capital spending projections for 2003 by $1 billion just yesterday, and  SBC Communications   slashed its capital expenditure budget for 2004 a few weeks ago. Sure, the percentage declines are getting smaller, which means a bottom is likely near, but whatever increase eventually occurs will be modest and nowhere near the heady days of yore.

In software, license revenue has been declining at double-digit rates for virtually every company we cover, which means that firms are simply buying less new software this year than they were last year. And since license revenue leads to service revenue, most large software firms are going to have weak growth in service revenue next year no matter how many new licenses they sell. Lest you think this is all in the past and that the Nasdaq's forecast of a rosier future is on target, let me note that the outlook for spending on software is still pretty grim--software CEOs are a lot less optimistic than their shareholders, it seems. 

Finally, contract manufacturers--the large outsourcing firms that assemble everything from routers to servers to MP3 players--are by and large running at around 50% capacity, and that level has been pretty constant over the past year. The only sign of any strength in the hardware arena is in consumer electronics, and given the fickleness of consumer taste and the speed of price compression for these kinds of products, it's not an area that I’d bet the ranch on.

So, while the fundamental outlook isn't getting any worse, it's not getting much better. But you wouldn’t know that to look at the valuations of most tech stocks. Using Wall Street consensus estimates for 2003, the average price/earnings ratio of the tech stocks in the Nasdaq 100 is 45. (Interestingly, only about half the Nasdaq 100 is tech and telecom.) That's not a stratospheric level, but it's hardly one that screams "bargain" to me. Moreover, it’s been proven over and over again in academic studies that Wall Street estimates are consistently too optimistic, which means that P/E of 45 could be closer to 50 or 60 if the "E" comes down.

The funny thing is that a lot of technology CEOs seem to agree that their shares are overvalued. My colleague Fritz Kaegi recently penned a fascinating study of firms that Morningstar covers that have issued convertible debt since last December. Those that issued convertibles were three times more likely to be trading above our fair value estimates than below them, which says to me that a lot of firms are taking advantage of their overpriced equity while they can. I'd also note that one of the firms that recently issued convertibles,  Juniper Networks (JNPR), last did so in March 2000. Pretty savvy timing on Juniper's part, wouldn't you say?

So, a high P/E and a slew of convertible offerings are two signs that tech stocks are overvalued. Here's another: The average price/fair value ratio of the 85 tech and telecom stocks that we cover is 1.46. In other words, the prices of the tech stocks we cover are, on average, almost 50% higher than our cash-flow-based estimates of their intrinsic value. For most of these firms, we're assuming some pretty robust double-digit sales growth over the next several years--and they're still trading at prices way above our fair value estimates.

Add it all up, and I'd say that you’d be playing some pretty poor odds if you bet on the Nasdaq right now.

Etc.
From the "selective amnesia" department: A perceptive colleague forwarded me a recent interview with  Sun Microsystems  CEO Scott McNealy. In it, McNealy scoffed at anyone foolish enough to buy his firm's shares at their peak valuations. When asked about the people who bought Sun stock in 2000, he responded:  "At 10 times revenues? Do the math. Do the math at 10 times revenues. There is no way to justify anything. Two times revenues implies 15 percent compound annual growth rate forever. Jack Welch did that for 20 years and went down in the hall of fame as the greatest CEO ever. So what does 10 times require? Do the math."

The irony is that Sun itself aggressively bought back its own stock at bubble-era valuations to offset the dilutive effects of its massive stock-option grants. As McNealy recommended, we did the math, and it turns out that Sun was paying about seven times revenue in 2000, and around four times revenue in 2001, for its own shares. By McNealy's own admission, that means he was expecting the firm to grow at between 30% and 45% "forever." At a minimum, the firm must have thought the shares were a decent value. Why else spend money buying them back?

It seems a bit odd for McNealy to use his 20/20 hindsight to chide investors for buying Sun's shares at sky-high valuations a few years back, when he was wasting shareholder capital doing the exact same thing.