New Math for a New Economy

My approach looks at the past. But I also look at the consensus forecasts of analysts. Based on those forecasts, I create an average, and I call that average normalized earnings. From those normalized earnings, I then subtract an average return on physical and financial assets, based on the theory that these are substitutable assets. Merck & Co., for instance, has lots of laboratories and manufacturing facilities. The equipment there is not unique. What is unique are the people, the patents, the knowledge that is being developed there. So when I subtract from the total normalized earnings a reasonable return on the physical and financial assets, I define what remains as the knowledge earnings. Those are the earnings that are created by the knowledge assets.

For example, my recent computations show that Microsoft has knowledge assets worth $211 billion -- by far the most of any company. Intel has knowledge assets worth $170 billion, and Merck has knowledge assets worth $110 billion. Now, compare those figures with DuPont's assets. DuPont has more employees than all of those companies combined. And yet, DuPont's knowledge assets total only $41 billion -- there isn't much extra profitability there.

Take a look at other companies where different kinds of knowledge assets make a big difference: I calculated that Phillip-Morris has knowledge assets worth $160 billion, largely because of its huge brand value. Coca-Cola is also a huge brand, and its knowledge assets are worth $60 billion. I identified another type of knowledge asset -- structural capital. Structural capital is a unique way of doing business. In the case of Dell Computer, the company doesn't produce computers that are better than other companies' computers, but the way in which it markets its computers is entirely different. Which is why Dell's knowledge capital totals $86 billion -- higher than that of Wal-Mart.

This is my first measure. I call it a top-down approach because it's an overall measure. According to the calculations that I've made, it performs far better than earnings or book value.

Is the top-down approach sufficient?

To complement the knowledge measure, we need to identify the drivers of knowledge. Here's how I think about it:

Economic theory and research tell us that almost all industries share a similar development pattern. They start out with a large number of companies, and then, usually after some kind of big innovation, there's a shakeout period that eliminates many of those companies. (At the turn of the century, there were more than 100 American car manufacturers; now there are only 2.) During the shakeout, most of the original companies fall by the wayside -- even those that previously were large and successful. General Electric was once a major semiconductor manufacturer, but it had to get out of the industry when the shakeout hit. So the question is, Who can survive the shakeout that invariably hits every industry?

The survivors are those companies that have good technology, because they have the ability to innovate. For me, technology also includes structural capital like that of companies like Dell and Home Depot. I'm in the process of developing a technological-capabilities index: an index based on measures that are quantifiable, publicly available, and linked to value. These aren't stories -- about good customer relations, good public relations, or good service -- but measures that can be supported by real research.

Let me give you an example. Some people use patents as an important knowledge attribute, but to me the number of patents that a company has is not very meaningful at all. You can get patents on almost anything, so simply having a large number of patents is absolutely meaningless. But there are ways to measure the attributes of the patents that have real significance. For instance, one powerful measure of the real value of a patent is how many times subsequent patents refer to it. If you have good science, then people will refer to you a lot, and you'll contribute a lot.

My technological-capabilities index is based on measures of inputs, such as investment in R&D, investment in product development, investment in information systems; on measures of intermediate outputs, such as patents and trademarks; on measures of competitive position, such as the number of people who access a particular Web site; and, of course, on measures based on the ultimate output -- commercialization. Commercialization of R&D is a powerful predictor of a company's success. Recently, a couple of French economists conducted a study using data that French companies are required to publish showing the percentage of their revenue that comes from new products. The results of their study demonstrate how important commercialization of new products is to a company's success in the marketplace.