IBM’s (IBM) second quarter revenue miss yesterday (July 19)—the company reported revenue grow of just 2% instead of the 4% to 5% that Wall Street had expected—undermined the entire technology sector. Last week after Intel (INTC) reported earnings of 51 cents a share instead of 43 cents and revenue about $600 million above projections, and then raised guidance for the rest of 2010, Wall Street started to believe that technology could lead an economic and stock market recovery.
The fear today, after IBM’s results, is that technology revenue too is headed into a slump.
I think Wall Street is misreading the messages in both Intel’s and IBM’s quarterly earnings report. The technology recovery was never as widespread as some bulls hoped after Intel’s results. But IBM’s disappointing results don’t change the very real, but somewhat narrower, trend in Intel’s numbers. (For more on Intel’s results see my post Update Intel (INTC).)
Wall Street got caught up in some wishful thinking after Intel’s report. Corporate technology buyers were back, analysts and investors concluded. And most other technology companies would report equally robust numbers as corporate customers put in orders. That rising tide of orders, Wall Street concluded, would be based on a belief among those corporate buyers that the U.S. economy wasn’t slumping and that growth in the second half of the year looked solid.
That conclusion was built, in my opinion, on a misunderstanding of the relatively special selling proposition that Intel offered those corporate customers. After two years of not buying during the Great Recession, corporate buyers looked out of their bunkers to see a new generation of servers—those computers specialized to handle the vast amounts of data that flow over the Internet—that were faster more powerful, and, since they used less electricity, cheaper to run. Even if the economy wasn’t going much of anywhere, these machines, because of their efficiency, would generate a positive return on investment in 12 to 24 months or less. For companies sitting on cash, and many are because they’ve been spending as little as possible while the economy flirted with meltdown, buying a new server was a no brainer. The return on the investment was better than the extremely low return on cash even if the economy didn’t turn around. And if economic growth was stronger than expected, the corporate buyer would be sitting pretty with extra capacity and more powerful servers.
It was this specific selling proposition that drove Intel’s sales of chips for servers to 170% growth in the quarter from the second quarter of 2009.
IBM, whatever its weaknesses and strengths (and as one of the few companies to show earnings grow in every quarter over the last three and a half years those strengths are pretty substantial), doesn’t fit that selling proposition very well.
IBM has become a services (58% of 2009 sales) and software (23% of 2009 sales) company. That’s one reason that it rode through the downturn as well as it did. Companies can put off buying new hardware but service agreements tend to be multi-year contracts and it’s hard to argue that a company should go without the services that keep its computer systems running day to day—even if the volume of business that those systems is handling has dropped.
But the service and software business doesn’t produce the kind of leap in efficiency or savings that comes with a new generation of servers. In fact, even software that will improve productivity eventually usually comes with a steep learning and implementation cost. New software and services will eventually find their way to the bottom line in many cases but the return is less obvious and the payback period less predictable.
So in the same quarter that Intel reported a 170% increase in sale of server chips, IBM reported a 12% drop in newly signed service contracts.
And the two results don’t say much of anything about each other.
IBM’s companywide results were driven even further away from Intel’s by the timing of IBM’s next hardware product cycle. New mainframe computers and new servers will go on sale over the next few months. The server products at least, will have a chance of tapping into the same dynamic that powered Intel’s sales in the June quarter.
For investors, the two earnings reports put together argue that we’re not likely to see the kind of sector-wide revenue growth that will indicate that the economy’s current weakness is just a slight temporary dip. It’s unlikely, I’d conclude, that the technology sector across the board will provide the kind of strong earnings and revenue news that can lead the entire stock market upward.
That doesn’t mean that Intel’s report was a fluke, however. I expect the same kind of surprisingly strong revenue growth in hardware sales from Cisco Systems (CSCO) and EMC (EMC). We’ll know if I’m right very soon. Although Cisco Systems won’t report until August 11, EMC reports tomorrow, July 21, before the stock market opens.