(FORTUNE Magazine) – In the pages of FORTUNE that follow are thousands upon thousands of statistics that reveal very little that's meaningful about the corporations they purportedly describe. At least that's the verdict of a growing number of forward-thinking market watchdogs, academics, accountants, and others. Convinced that accounting gives rotten information about the value and performance of modern, knowledge-intensive companies, they're proposing changes that would be earthshaking to the profession--and a big deal to everyone else too.

It's about time. For years scholars, investors, and managers have grumbled that Generally Accepted Accounting Principles (GAAP) generally do an unacceptable job of accounting for the principal activities of Information Age companies. In Relevance Lost: The Rise and Fall of Management Accounting, H. Thomas Johnson and Robert S. Kaplan write: "Corporate management accounting systems are inadequate for today's environment," citing technological change, globalization, and expanding information processing as three reasons old charts of account aren't useful. That lament dates from 1986, before the Internet rose and the Berlin Wall fell, when men were men and P/Es were ten.

Since then it's only become clearer that accounting is "lousy" (McKinsey Chairman Rajat Gupta's word) at describing today's most important assets and activities--intellectual capital and knowledge work. The value of R&D, most software, brand equity, or the user base of an e-tailer slips through its fingers. In January 2000, Federal Reserve Board Chairman Alan Greenspan complained that accounting wasn't tracking investments in knowledge assets and warned, "There are going to be a lot of problems in the future."

Actually, problems are here now. Former SEC Chairman Arthur Levitt told the Economic Club of New York, "As intangible assets grow in size and scope, more and more people are questioning whether the true value--and the drivers of that value--are being reflected in a timely manner in publicly available disclosure."

Make no mistake: This is not just theoretical. It costs investors money--perhaps you, dear reader, among them. We're not talking fraud here--we're talking irrelevance, with the result that investors are in the dark and managers operate by guess and by gosh. At the very least it reduces prosperity by distorting flows of investment capital, which should go where it can be most productively employed.

The Financial Accounting Standards Board, the profession's vestal virgins, says that accounting's fundamental purpose is to "provide information that is useful...in making rational investment, credit, and similar decisions." By that standard, it flunks. Take a big, obvious question: What's a company worth? If the books tell a story investors find useful, then a company's market value should roughly (not precisely, because the market looks forward and the books back) correlate with the value accountants ascribe to it. It doesn't: Arthur Andersen consultants Richard Boulton, Barry Libert, and Steve Samek compared market value with book value for 3,500 U.S. companies over a period of two decades. At the beginning, in 1978, the two were pretty well matched: Book value was 95% of market value. Now they've gone astray, every one to his own way. Twenty years later, book value was just 28% of market value. Investors simply don't value what accountants count. Not even earnings and cash flow numbers seem to help investors. Baruch Lev, a professor of accounting at the Stern Business School at New York University, has found that changes in earnings and cash flow have trivially weak relationships to changes in market capitalization.

Lev is one of several leading accountants mapping paths from GAAP to genuinely informative accounting procedures. He's in the more radical camp. He breaks with accounting tradition, which infers value by adding up spending and other transactions. His different approach--a breakthrough or a breakaway, depending on whom you ask--allows him to infer a dollar value for almost any company's intellectual capital from its performance.

A less radical group recognizes that some items currently expensed should really be capitalized but stays within accounting's transaction-based tradition. Prominent among them is Robert A. Howell, visiting professor at the Tuck School at Dartmouth and an authority on the changing role of finance and accounting. To Howell, the "big three" financial statements--income statement, balance sheet, and statement of cash flow--are about as useful as an 80-year-old Los Angeles road map.

Their very logic is outmoded, Howell argues. The income statement, boiled to its bones, looks like this:

Traditional Income Statement

Revenues minus Cost of goods sold Gross margin minus Operating expenses

Earnings before interest and taxes minus Interest Taxes

PROFIT

It was set up that way to highlight the most important cost--the cost of goods sold. When raw materials and direct labor made up most of a product's cost, that was a number managers and investors had to know. Today it doesn't deserve such prominence. For Microsoft, the cost of goods ("cost of revenue," they say) is just 14% of sales, for an 86% gross margin. Coca-Cola's gross margin is 70%. For Revlon, it's 66%. When gross margin is about as meaningful as your ex's promise of eternal love, why bring it up?

Further, the income statement long ago ceased to be a decent measure of profit or loss--"profit" having become a subjective number that depends on when revenues and expenses are recognized. Many companies manage their earnings better than they manage anything else.

Therefore, Howell argues, forget about profit. Focus on cash. Call it an "operating statement," not an income statement. Remove interest expense, a financial cost. Add back all the noncash items. Format it like this (again, a bare-bones view):

Howell's Operating Statement

Revenues

minus Cost to serve customers Cost to produce products/services Cost to develop products/services Administrative costs Earnings before interest and taxes

minus Taxes plus/minus Non-cash adjustments

CASH EARNINGS

That change would give readers a decent idea of where a company spends money without spilling proprietary beans. It replaces manipulable "profit" with gen-yew-wine cash money. It focuses on the real work of 21st-century corporations: taking care of customers (sales and marketing, shipping, service), producing things to sell (manufacturing or providing services, materials, equipment), and producing future offerings (research and development, knowledge creation). It draws a tight box around administrative expense, a goad to efficiency. Today's oft-used "sales, general, and administrative cost" is anachronistic. It assumes selling is a trivial cost to be minimized. For many service companies, the cost of selling and the cost of producing are pretty much the same thing, so Linda Loman, Willy's wife, was right: Attention must be paid. Also, "SG&A" can and does hide a multitude of sins. Does Bristol-Myers Squibb run a tight ship? An investor can't tell, because the company lumps administration with marketing and sales, a category that totals about 24% of costs.

Howell would also rewrite balance sheets to correct their failure to account for intellectual assets and to bring them in line with high-speed, real-time business. Balance sheets are snapshots-- frozen-in-time pictures of what resources (assets) a company controls and where it got them (from investors of equity or by borrowing). The old balance sheet--which took its present form in 1868--did a reasonable job of portraying old realities. As Howell says: "To support a dollar's worth of sales, old-line companies might have had to put 15 to 20 cents' worth of earnings back on the balance sheet in the form of working capital"--inventories and receivables. Another big tranche of earnings had to be turned into fixed assets to make up for depreciation and support growth, which depended on adding factories, lines of track, or stores.

That's a lot less meaningful for companies like Dell, which has negative working capital. Thanks to e-commerce, even traditional manufacturers need less working capital: Durable-goods makers would be carrying an extra $115 billion in inventory--about a third more than they do--if they operated by the ratios that prevailed in 1988. Today's companies need substantially less in the way of physical assets too.

The fundamental balance sheet equation is, of course:

Assets = LIABILITIES + EQUITY

Howell would change it, to

Investments = FINANCINGS

The money invested in a business has to equal the money raised for it. (If it doesn't, look for your missing CFO on Grand Cayman.) On the investment side go the three usual suspects--working capital (cash, receivables, and inventory), fixed assets (but at market value, not cost), and investments (at market value). To them Howell would add a set of intangibles. The know-how of a work force, intellectual property, brand equity, and relationships--these are the real assets of knowledge-intensive companies. Money spent on them should therefore, as much as possible, be booked as investments.

How should these intangibles be valued? Various tested ways of valuing R&D exist. A transaction-based evaluation, the most compatible with traditional accounting, would book the cost and amortize it over average product lifetimes. Likewise there are established ways to measure brand equity and evaluate loyal customers: They're used all the time in mergers and acquisitions work. Valuing people is harder. Says Howell: "At a minimum, I would book as assets such costs as recruitment and training and development, and amortize them over some sort of employment life. Why aren't they as much an asset as some piece of machinery?" The counterargument is that by definition an asset is something that can be owned, and you can't own people. Yesss, that's the definition--but what of it? The whole point is that the current definition is inadequate to the realities of a knowledge economy.

The financings side might need adjusting too. For instance, if recruitment and training are booked as assets, then some employee contracts might be booked as liabilities.

A balance sheet like this answers questions managers and investors need to ask: What are you really doing with the money you raise? What did your spending on people, brands, and research get you?

Last but not least, Howell would revise the statement of cash flows--now "an abomination" that looks like this:

Traditional Cash Flow Statement

plus/minus Operating cash flows Investing cash flows Financing cash flows

CHANGE IN CASH

The purpose of the cash flow statement is to illuminate a company's liquidity and show how management uses financial resources. A new operating statement (vs. an income statement) would do the latter job better. As for revealing liquidity, the candlepower of cash flow statements is kinda low. The story they tell--is the account up or down at the end of the year?--is trivial, says Howell: "It's useless to a manager." Managers (and investors) need to know how much cash a business produces over and above what's needed to operate it--free cash flow, that is. To get that, the cash flow statement ought to look like this:

Howell's Cash Flow Statement

Cash earnings* minus Investing activities**

FREE CASH FLOWS

*From operating statement. **Plus or minus working capital and minus investments in fixed and other assets.

This architecture, which Howell calls a Digital Age accounting framework, makes sense for lots of reasons: It tells a clear story, makes flimflam hard to hide, and--most important--focuses on the real concerns of business: producing cash and creating value.

Howell's approach doesn't go far enough for Baruch Lev, who doesn't believe you can build a "Digital Age framework" on the ruins of Analog Age accounting. Instead, he wants to build a knowledge-measurement system suited to the new age from the get-go.

Lev got interested in intangible assets a decade ago, when he was teaching at Berkeley and consulting about valuation issues that arose in the course of litigation. NYU wooed him east with money to set up an accounting research institute, from which he sends out a fast-moving stream of innovative work. Lev is a radical insider: an indefatigable, controversial critic of his profession's failure to measure intangibles. His own methodology, first described in print in CFO magazine, is a ballsy attempt to put a dollar value on intellectual capital in a way that can be tested in the market.

Forget about adding up transactions to find the value of intellectual capital, Lev says. They show only the tip of the iceberg. Most knowledge assets are homegrown, so no transaction takes place. A truck is worth what you pay for it, but the value of R&D is not necessarily the amount you spend--there's no proven correlation between cost and worth. Nor can you try to mark intangibles to market; for human capital or customer relationships, there may be no market except M&A.

Lev rejects a common measure of knowledge assets, which subtracts a company's book value from its market value and labels the difference "intellectual capital." The most egregious of this crude measure's many problems is this: The number rises and falls with market exuberance. It assumes that tangible assets have a fixed underlying value, regardless of the market, but intangibles do not--yet trying to find the real value of intangibles is the very purpose of the exercise.

Lev therefore approaches the problem from an entirely different direction. He starts with performance--company earnings--and then digs inside to identify what assets produced the earnings. It's a "watch what we do, not what we say" method.

Here's how it works. (This is oversimplified; there's more in the footnote at the end of the story.) Start with a company's earnings--$1 billion, say. Then look at its balance sheet to see what it has in the way of financial assets. Say that figure is $5 billion. The expected after-tax return on financial assets is in the neighborhood of 4.5%, so the $5 billion in such assets explains $225 million of the earnings. Say the company also has $5 billion in physical assets--property, plant, and equipment. The average after-tax return for tangible assets is about 7%. So $350 million of earnings can be credited to them.

That leaves $425 million that must have been produced by assets not on the balance sheet. Lev calls that residual "knowledge-capital earnings" (KCE). He then calculates the knowledge capital itself by dividing the earnings by an expected rate of return on knowledge assets:

Knowledge Capital=

KCE [DIVIDED BY] KNOWLEDGE CAPITAL DISCOUNT RATE

The question: What's the right discount rate? There's no historically tested rate of return for knowledge assets, so Lev uses a proxy. He takes the average after-tax profits of two industries that depend on knowledge assets almost to the exclusion of any others: software and biotechnology--10.5%. By inserting it in the formula, one can put a dollar figure on a company's knowledge capital. To produce $425 million in earnings, our imaginary company would need $4.06 billion in intangible assets, assuming they produced a 10.5% return.

Now, at first glance it may seem that Lev's intellectual-asset discount rate was plucked from thin air. But the market validates Lev's formulations. Working with Seng Gu of Boston College, Lev looked at whether cash flow, traditional earnings, or knowledge earnings most correlates with total return from the stock. The results: They found just a 0.11 correlation between stock returns and cash flows, a 0.29 correlation with earnings, and a strong 0.53 correlation with knowledge earnings.

Lev's work has lots of implications. By themselves the three after-tax rates of return--4.5% for financial, 7% for physical, and 10.5% for intellectual assets, however imprecise--ought to tell managers something about where to invest. Lev did a study for the chemical industry that dramatically confirms the point: Looking at 83 companies over a span of 25 years, he found that their R&D investments returned 25.9% pretax, whereas capital spending earned just 15%.

With Marc Bothwell, a vice president of Credit Suisse Asset Management, Lev has been investigating the investing implications of his work. For example, calculating knowledge capital can change your opinion about whether a stock belongs in the growth category or the value camp. Today a company is put in one category or the other mostly on the basis of its market-to-book ratio. Lev, who says, "That is ridiculous," would instead add the knowledge value to the book value, call the sum "comprehensive value," and compare it with market value. In case after case, Lev and Bothwell find that stocks move in the direction suggested by comprehensive value, not book value. One example: On Aug. 31, 2000, Costco and Target had almost identical market/book ratios (3.3:1 and 3.5:1), but Costco had a much higher market-to-comprehensive/book ratio (1.52:1 vs. 0.98:1), indicating that Target was a better value. In the six months that followed, through Feb. 28, 2001, Target returned 68% to shareholders, Costco 21%. This is just one case, over a short period, and though there are many such examples, it's far too soon to know if comprehensive value will prove to be a great market indicator--but at this stage the anecdotal evidence is compelling.

Lev's method is not unprecedented: Under some circumstances the Internal Revenue Service lets companies set a market value for intangibles by calculating their probable contribution to earnings. But it is highly unconventional and needs testing over time--especially given the dot-com tulip craze of 1999 and 2000. The proxy for knowledge-capital returns also needs testing and should include returns for pure-knowledge service companies, such as advertising agencies or publicly traded professional firms. In addition, Lev hasn't yet figured out how to apply his method to banks and insurance companies; the financial assets they manage act like lead aprons, preventing him from X-raying their operations. It would also be fascinating to test the validity of historical-cost transaction data for tangibles and intangibles (as Howell, for example, would book them) against Lev's data.

But Lev is clearly the advance scout in a territory others are entering. Last fall a Brookings Institution group led by former SEC commissioner Steven Wallman and Georgetown law professor Margaret Blair called on regulatory bodies to help correct "a serious potential problem for business managers, for investors, and for government." Adding their voices are accounting professionals like Robert Herz, a partner at PricewaterhouseCoopers, who vows, "One of these days there will be standards for the reporting of intellectual capital." FASB itself takes a baby step in that direction in a just-issued hundred-plus- page special report on accounting's treatment of "new economy" assets. In typical rock-the-boat-gently style, the report suggests that the board might consider undertaking a quartet of projects, two to examine whether internally generated intangibles (like intellectual property) should be capitalized, another to look at using footnotes to disclose more about intangibles, and a fourth to see if nonfinancial performance measures (such as capacity utilization and customer lists) might be disclosed to the public.

What holds them up? Fear, basically--the dread of something after GAAP, the undiscovered counting. One argument is that measures of intangibles are imprecise and subjective. "People in glass houses..." is the polite response--or maybe "Balderdash." There's plenty of hooey in supposedly objective valuations of fixed assets, and the valuation of intangibles gets better all the time.

True, one shouldn't lightly discard a system that has, for all its faults, served business and the public well, like an octogenarian butler. It is no small thing that GAAP makes fraud difficult and, in its absence, produces pretty good numbers--even if they are not always useful. Today, says Steven Wallman, "some of the most useful information is not necessarily the most reliable, and some of the most reliable is not necessarily the most useful." He's right. Though "consistent" might be a better word than "reliable," consistency is too important to cast aside. At the same time (as Wallman agrees) there's increasing evidence that the faithful servant isn't just misplacing a spoon here or there but has lost track of some valuable jewels, paid no attention to the furnace and the water heater, and put the place at risk.

FEEDBACK: tstewart@fortunemail.com