New Math for a New Economy

And while the benefits that come with knowledge assets can be enormous, they are much more uncertain than the benefits of tangible assets. When you invest in a tangible asset, such as an office building, you always get some kind of return -- even during a recession. And when boom times come, your property really pays off. But when you're building a knowledge asset, you could quite possibly end up with nothing.

Given the nature of knowledge assets, what is the conflict between these new assets and the old laws of accounting?

One problem is that you end up with accounting practices that are virtually antithetical to the business practices that they're trying to measure. Let me give you an example. In 1994 and 1995, America Online capitalized some of its customer-acquisition costs -- which means that it considered part of those costs assets. In other words, AOL was saying that, in acquiring new customers, it was creating a unique asset -- one that would help the company become even more profitable in the future. Financial analysts called that cheating! It was a new industry, competition was fierce, and analysts thought that AOL was trying to manipulate its earnings. Finally, in October 1996, AOL gave up and completely expensed its $385 million in customer-acquisition costs.

Today, AOL has a market value of roughly $140 billion. Compare that with the $385 million that it tried to capitalize, and it's almost humorous! And yet only five or six years ago, financial analysts were proclaiming that AOL was a cheat.

Or take, for example, what happened when IBM acquired Lotus in 1995. As an accounting requirement, IBM had to estimate the fair-market value of the assets that it had acquired. IBM estimated that the portion of Lotus's R&D that was in process -- R&D for which there was not yet a product -- was worth $1.84 billion. That's 53% of the entire $3.5 billion acquisition price. IBM expensed the entire thing -- because those are the rules of accounting: Once you estimate that something is in-process R&D, you have to expense it. As a result, no trace of an asset remains.

This kind of mindless writing-off of all investments in knowledge assets means that there is no accountability -- and no ability to measure the performance of an investment or to learn from it. And the problem is only getting worse: Over the past 20 years, as the actual value of companies' intangible assets has been going up, that value, as it is represented in financial reports, has appeared to be consistently going down.

How do you account for this failure of accounting?

Accounting is based on the matching principle. To determine your earnings, you match your revenues against your expenses. It's that simple -- that's what an accounting system does. And if the matching is good, you get a reliable income number.

Now here's the problem: With knowledge assets, you get a complete mismatch, and the system breaks down completely. Take the AOL example. During its period of tremendous growth, AOL immediately expensed all of its customer-acquisition costs. So for that period, the company was showing those costs as big losses. Then, once the customers were acquired, the company realized large benefits -- which then increased its income without any associated costs! So both periods are misstated in financial reports. Companies that are on a steep growth curve -- for example, Internet and biotech companies -- are most likely understating their results. And older companies that have plateaued are most likely overstating their results. The outcome is a disconnect with the market, which is supposed to reflect reality.

There's another disconnect between the world of accounting and the world of knowledge assets: Accounting records transactions, but much of value creation or value destruction precedes any transaction. Look at regional telephone companies. In the late 1980s, deregulation started to hit the regional phone system. The old system was based on a guarantee of a reasonable rate of return for phone companies. The companies had an assured monopoly and assured profits. The new system was more open and competitive.

Investors immediately understood the implications of moving from a secure monopoly to a competitive system: higher risks, lower returns. But the accounting system didn't reflect any change at all -- because deregulation is not a transaction! Five or six years after deregulation began, the Baby Bells finally said that their assets must be much lower than before, and wrote off $27.6 billion of assets. But in the preceding five years, there was an almost total disconnect between what the company was actually worth, what the accounting system showed, and what the markets understood.