When Meg Whitman was named chief executive of Hewlett-Packard in 2011, she inherited a real mess. HP’s share price had been slaughtered by the ouster of Mark Hurd and the reign of Léo Apotheker—an 11-month period that I like to refer to as the Grim Bumbling. Trying to revive hope in the company and its stock, Whitman vowed to invest in what had once made HP great. She would pour money into research and development, HP would invent wonderful things, and investors would eventually rejoice.
While this plan seemed sound enough in theory, a new study has suggested that it might have been exactly the wrong thing to do.
The data sleuths over at Bernstein Research have just issued a sizzling report with the title “Do High R&D Spenders in Tech Generate Stock Outperformance?” Even if it’s not the sexiest title, the conclusion is titillating. Bernstein examined technology companies since 1977 to measure R&D spending as a percentage of the company’s sales. The researchers found that over 1-, 3-, 5-, and 10-year periods, the companies with the lowest spending on R&D tended to perform the best on Wall Street. It really is just like Thomas Edison said: “Invention is 10 percent inspiration and 90 percent a waste of everyone’s time.”
Bernstein divided the technology companies into High, Medium, and Low R&D spenders. Roughly speaking, the high group tended to spend 18 percent to 35 percent of revenue on R&D; the medium group spent 11 percent to 17 percent; and the low group 10 percent or less. Members of the high-spending group tended to be heavy on hardware companies with names such as Nvidia and Intel, the middle range encompassed big names like Google and Microsoft, and the low-end group included HP, IBM, and Apple. Some companies changed groups depending on which time period Bernstein examined.
If we focus on results from the past five years, the biggest R&D spenders underperformed the other technology companies by 15 percent on average when it came to gains in share price. The middle R&D spenders outperformed the rest of the companies by 23 percent on average, while the lowest spenders outperformed the rest by 19 percent. And the majority of the low spenders—63 percent—outperformed their peers; from the middle and high spenders, only 43 percent and 40 percent, respectively, bested their peers.
The top-performing low spenders: Salesforce.com, Apple, and Akamai. VMware was the best-performing high spender.
Many of the companies that make Bernstein’s list as big-time R&D spenders come from the chip industry. These are companies like Marvell, Broadcom, and Nvidia that all spend about 25 percent to 33 percent of what they make on R&D. The chip companies suffer from the unfortunate fate of being locked in a constant battle with the laws of physics to build tinier, faster, cheaper things, and it’s only natural that you have to pay good money to keep pace with the competition.
These smaller chip companies also have to go up against a pair of industry heavyweights in Intel and Qualcomm that spend a lot on R&D, which makes life more difficult. As Bernstein notes, many of the high R&D spenders “lacked scale,” or, in other words, were underdogs trying to find an edge to help them play above their weight.
The research firm’s other big realization is that large companies can spend less but make more of their R&D investments through what it calls high R&D productivity:
“Within our coverage universe, Apple and IBM have relatively low R&D to sales ratios (2.9 percent and 6.2 percent respectively) and their stocks have been significant long-term outperformers. While they both benefit from scale, we also believe that have strong R&D productivity. For example, we note that Apple and Dell both spent the same dollar amount on R&D through 2004/2005 (roughly $500 million per year), yet Apple at the time owned its own operating system (OSX), had developed a digital music ecosystem (iTunes) and was generally considered to have superior and more innovative product design.”
So did Whitman really make a mistake by upping R&D spending? The answer might actually be: no. It’s not like HP has unleashed a fire hose of research spending. It’s making a couple of targeted bets around new types of servers and some 3D printing technology. HP actually remains one of the low R&D spenders that could benefit from using its market reach to make the most of its investments.
The Bernstein report is interesting and counterintuitive. Companies that conduct a lot of research are typically thought of as innovators and the producers of high-profit products. They’re the proper “technology” companies and should be the ones that investors celebrate over the long term because of their ability to fend off competitors and keep adding value for which customers are willing to pay.
But the report is also unsatisfying. R&D is a tricky thing to measure. Microsoft, for example, spends close to $10 billion a year on R&D, which is a staggering figure. But huge chunks of that money go toward the development side of the equation, which means tuning future versions of Windows and Office rather than inventing teleportation devices. Is Microsoft then the best-funded research laboratory on the planet or a company mostly focused on making incremental improvements to productivity software? It’s hard to tell sometimes.