SEEING IS BELIEVING

The 18th-century philosopher Immanuel Kant spent much of his life pondering experience beyond the physical realm. Although a stranger to modern finance, he might feel right at home in the current debate about measuring and reporting knowledge assets. Just as Kant concluded that metaphysical reality exercises more influence on human behavior than objects we can see and touch, some observers today recognize that intangible knowledge assets increasingly drive financial performance.

Where knowledge assets are concerned, accounting practice has not changed much since Kant's day. After several hundred years of double-entry bookkeeping, accountants still do not treat knowledge assets as assets. CFOs can keep close tabs on spending tied to patents, brands, trademarks, and research and development, but unlike routine accounting for tangible assets, there are no rules for estimating or reporting the cumulative consequences of these investments.

Meanwhile, the industrial revolution has given way to an age of computers, information, and global competition. Decisions are made more quickly, and superior knowledge of products, markets, methods, and cultures often determines who prevails. The desire for mobility favors assets that can move over those assets rooted to the ground. This spurs additional investment in assets that may affect the balance sheet, but are not recorded there.

Despite increasing awareness that the value of knowledge assets now approaches or even exceeds the value of reported book assets, rulemakers in the United States have largely dodged the issue. Since hosting an exploratory seminar on intangible assets in 1996, the Securities and Exchange Commission has been silent. At the Financial Accounting Standards Board, which often leads the way on such matters, knowledge assets are "not on the horizon," says a FASB spokesperson.

Whether or not regulators choose to deal with it, a gap exists between a knowledge-based economy and accounting measures geared to an era when most corporate assets resided on factory floors and shelves. Debate centers on the merits of measuring knowledge assets, not on whether they exist.

Skeptics, including many CFOs, insist that knowledge assets cannot be measured in a meaningful way, chiefly because nothing actually changes hands. Barring calamity, tangible assets won't vanish overnight, but the value of a knowledge asset can. A high-tech patent can be improved upon or superseded by a new technology; an unfounded rumor can diminish the reputation of a brand name; promising R&D can come a cropper; a top executive can go elsewhere.

"The closest you'll come," says David Shedlarz, CFO of pharmaceuticals giant Pfizer Inc., "is by determining the quality of existing and future product pipelines, as well as the capacity to develop, manufacture, and market new pharmaceutical products, taking risk factors into consideration. None of that is on the balance sheet." Until a mechanism for evaluating knowledge assets wins wide acceptance, says Shedlarz, they won't belong on the balance sheet.

Skeptics predict unintended and unwelcome results if knowledge assets find their way onto financial reports. Far from clarifying performance, in their view, knowledge assets will instead mislead investors, distract managers, and foster uncertainty.

These fears are overblown. There are more-serious deficiencies in an accounting system anchored to physical assets. The value of a single knowledge asset may indeed rise or fall unexpectedly, but successful long-term corporate performance demands optimal control over the levers of value creation. This cannot be accomplished fully without the means to assess how well companies manage knowledge capital.

Companies need to answer such questions as: Are returns on R&D satisfactory? Are patents worth renewing? Are brands worth defending? Failure to address these deficiencies already undercuts prospects for optimal decision making. Companies that fail to address these deficiencies will ultimately lose out to competitors that learn to measure, manage, and leverage their knowledge assets.

The momentum for change is growing. "We are just at the stage of forming a consensus," says Steven M.H. Wallman, co-chair of a Brookings Institution task force charged with advancing the visibility of knowledge capital. During his tenure as an SEC commissioner, Wallman helped initiate the move toward consensus. "Once the intellectual foundation is laid for the need for change," he declares," and once world-class academic research shows the importance of this evolution, we will see a serious discussion about how to make the system more reflective of the economy as it is today and as it is likely to be in the future."

Absent a secure handle on knowledge assets, companies address the knowledge-capital gap either in a piecemeal fashion or by resorting to shortcuts with obvious flaws. Attempts to assess knowledge capital one product at a time stumble on such questions as how to allocate overlapping benefits from companion products and services.

Developing A Knowledge Index
A broad proxy for knowledge capital is the subtraction of book value from market value. This proxy assumes, however, that book assets have no value above reported cost--an assumption that is plainly suspicious, if not dead wrong. The resulting figure overstates knowledge assets and understates the value of book assets. Moreover, because it is a stock-market-based measure, book value fluctuates from moment to moment, in a manner not related to the underlying value of the assets.

Calls for companies and regulators to start measuring knowledge capital caught CFO magazine's attention in 1996, when we published "Getting a Grip on Intangibles" (September), by contributing writer Randy Myers.

Response to the story prompted us to ask if intangibles, or knowledge assets, can indeed be measured. We turned for guidance to Baruch Lev, the Philip Bardes Professor of Accounting and Finance at New York University's Leonard Stern School of Business. Lev, an outspoken advocate of greater recognition for knowledge assets, embraced the challenge by formulating the first Knowledge Capital Scoreboard.

With critical assistance from portfolio manager Marc Bothwell, of BEACredit Suisse Asset Management, Lev developed this performance-based calculation of knowledge capital as a tool for measuring the economic consequences of investment in knowledge assets.

As a first step, we decided to test the merits of the Scoreboard in two industries. Lev and Bothwell computed levels of knowledge capital at 27 chemical companies with sales in excess of $1 billion, and 20 pharmaceutical companies with sales in excess of $250 million.

Lev's novel interpretation of intangible-asset values chiefly reflects the assumptions underlying conventional appraisals of tangible assets. Standard aftertax return expectations for tangible and financial assets drive two-thirds of the analysis. The other third of the analysis rests on an estimate of long-term expected returns on knowledge assets. Since no such historical calculation yet exists, the Scoreboard substituted a proxy: average aftertax expected return (Ibbotsen & Associates's cost of equity) for three industries that consist almost entirely of knowledge assets--computer software, biotechnology, and pharmaceuticals.

The Formula For Knowledge Capital
Calculating knowledge capital starts with an estimate of each company's annual normalized earnings. For the Scoreboard's purposes, the estimates encompass three years of historical data through year-end 1997 plus earnings forecasts for one, two, and three years taken from IBES International consensus estimates. To accommodate business changes that are likely to affect future results, averages give slightly greater weight to earnings forecasts. Performance improvements expected beyond three years out do not affect the Scoreboard, Lev warns.

Once normalized earnings are established, a figure for knowledge-based earnings, the Scoreboard's primary building block, is determined based on a residual calculation. Expected rates of return for each broad asset class are applied to tangible book assets and financial assets.

The Scoreboard multiplies the recorded assets by their respective aftertax expected returns--7 percent for tangible assets and 4.5 percent for financial assets. The rates of return apply to all companies in both industries equally, irrespective of each company's risk profile or cost of capital. With more time and resources, companies can adjust these rates to reflect their specific track records. This limitation notwithstanding, the process computes credible measures of earnings linked to tangible and financial assets.

Subtracting tangible and financial earnings from normalized earnings leaves a portion of normal earnings unaccounted for. This residual, says Lev, represents earnings generated by knowledge assets, or knowledge capital earnings (KCE). This calculation by itself suggests a range of new financial metrics, including a knowledge capital margin (KCE/Sales) and a knowledge capital operating margin (KCE/ Operating Income), to name just two.

With these measures, companies can, for the first time, estimate the contribution of knowledge assets to their profitability and performance.

The formula solves for knowledge capital by reversing the steps for calculating earnings that tangible and financial assets generate. Instead of multiplying assets by their expected returns, knowledge-based earnings are divided by an expected rate of return for knowledge assets. It's a familiar process, with one hitch: no one has ever computed an historical expected rate of return for knowledge assets.

This is where the proxy comes in handy. Absent an acceptable alternative, the average aftertax expected rate of return for three knowledge-rich industries (software, biotechnology, and pharmaceuticals) supplies the aftertax knowledge capital discount rate of 10.5 percent.

To take an example, normalized earnings at pharmaceuticals company Merck & Co. are $5.5 billion, according to our estimate. Average tangible assets of $4.9 billion (net of long-term liabilities and reserves) can be expected to generate earnings of $343 million, while financial assets of $624 million can be expected to generate $28 million. The residual, $5.1 billion, represents Merck's knowledge earnings.

Applying the 10.5 percent as a proxy for the knowledge capital discount rate to knowledge earnings yields the Scoreboard's figure for knowledge capital as of year-end 1997: $48 billion.

As expected, company size affects results. The amount of knowledge capital at large companies will often exceed that at smaller companies in the same business. There is much more to the story, however, as comparing Merck with chemical company DuPont suggests. DuPont rang up sales of $40 billion in 1997, versus $24 billion at Merck. But when it comes to generating knowledge capital, Merck's knowledge-intensive portfolio of existing medications and drug pipelines enjoys a nearly two-to-one advantage over DuPont's capital-intensive portfolio of commodity-chemical and plastics-processing capabilities. The bottom line: Merck has accumulated $48 billion in knowledge capital compared with $26 billion at DuPont.

With figures in hand for knowledge capital and related earnings, new metrics can cast conventional financial statements in a fresh light. Calculating the ratio of knowledge capital to book capital, for instance, suggests the degree to which a company is knowledge-based. For every dollar of book capital at DuPont, there is $2.34 worth of knowledge capital, among the highest for chemical companies. A ratio of knowledge capital to sales adds another dimension, the knowledge- capital margin. A downward change in this ratio hints at lower contributions to overall performance by knowledge-capital assets; a shift upward suggests greater contributions can be expected. Warner Lambert and Bristol-Myers Squibb, with the highest ratios of knowledge capital to book value, 4.27 and 4.22, respectively, are also amply rewarded by investors, as indicated by their highest and third-highest rank in the traditional market-to-book ratios.

A More Robust Appraisal
Merging a new metric with a familiar one, the Scoreboard introduces the concept of "comprehensive value," arrived at by adding knowledge capital to book value. Comprehensive value yields an appraisal of the balance sheet that takes all corporate assets into account, from machine tools to patents. Analysts frustrated by widely divergent ratios of market value to book value might welcome a more complete picture of total assets.

Gone are whopping multiples of book value, replaced by more- modest multiples of comprehensive assets. While Warner Lambert has the highest market-to-book ratio, pipeline-rich Pfizer enjoys the highest market-to-comprehensive-value ratio.

"A 1-to-1 ratio," says Lev, "would indicate that market value is mostly derived from past performance and short-term earnings forecasts." Or, to put it another way, a 1:1 ratio indicates that near-term earnings expectations are embedded in the stock price, and not much growth is expected beyond that in the absence of new developments.

As a systematic model for evaluating knowledge assets, the Knowledge Capital Scoreboard attempts to capture a critical dimension of value that conventional accounting has failed to grasp. Reliance on public information and broad assumptions make it available to everyone, but these steps also impose obvious limitations. Access to proprietary information about expected returns and itemized expenses can refine the methodology and, in all likelihood, enhance the results. If any single conclusion stands out, it is the shortsightedness of ignoring knowledge assets. They exist, they can be estimated, and they deserve recognition.

The ROI of R&D

Knowledge capital generated per dollar of R&D investment

Conventional assessments of R&D productivity leave much to be desired. Measuring R&D costs against sales, operating earnings, or net earnings, for example, mixes an investment in knowledge assets

with output governed by market climate, quality of working-capital management, or financial engineering. A book-value ratio would make sense if R&D expenditures produced property, plant, equipment, or inventory. An apples-to-apples comparison, on the other hand, measures investment in knowledge assets (R&D) against the resulting knowledge assets. In other words, how many dollars of knowledge assets does one dollar of R&D generate?

With access to a credible, performance-based estimate of knowledge capital and a three-year average investment in R&D, the ROI of R&D comes into view. According to our estimates, every dollar DuPont spends on R&D creates $16.89 of knowledge assets, while each dollar Merck spends creates $32, a gap consistent with the observation that pharmaceuticals constitute a more valuable business franchise than commodity chemicals do.

Low levels of R&D in the denominator skew results, just as negligible earnings skew price-earnings ratios. As a result, comparing R&D productivity clearly works better for companies with similar financial profiles. For this reason, these charts focus on the top three companies in each industry. The year-to-year trend is also important. Companies that trim R&D spending might at first enjoy a jump in perceived performance, but probably at the expense of building knowledge capital long-term.

R&D, of course, is not the only source of knowledge capital. The Scoreboard is limited to public data, but companies can add to R&D costs their annual expenditures on brands and worker training, as well as other appropriate investments, to get a more meaningful ratio of knowledge capital to its drivers.

The First Cut

How knowledge capital compares in the chemical and pharmaceutical industries.

Long-term success generally lies in a company's capacity to create value above and beyond the value associated with tangible and financial assets, but balance sheets do not record this information. Conventional accounting rules make no explicit provision, moreover, for computing the values of assets that exist outside the physical realm.

In an effort to supply the missing information, CFO magazine invited Prof. Baruch Lev to develop a performance-based measure of knowledge capital.

The upshot, the Knowledge Capital Scoreboard, highlights knowledge capital at 27 chemical companies with sales in excess of $1 billion, and at 20 pharmaceutical firms with sales in excess of $250 million.

Selection of these industries for the Scoreboard's maiden run reflect two primary objectives: (1) to single out companies whose businesses rely on knowledge capital, and (2) to capture a wide spectrum of companies. Hefty commitments to research and development provide intuitive evidence that chemical and pharmaceutical companies assign great importance to knowledge assets. Chemical companies occupy the low end of the spectrum; pharmaceuticals occupy the high-end.

The means to calculate knowledge capital (column one) yields whole new twists on familiar financial metrics. These new twists suggest a much more robust portrayal of corporate value than conventional metrics that reflect only tangible assets.

Size confers an edge. The largest chemical companies accumulate more knowledge assets than do the smallest companies; likewise for pharmaceuticals. On the other hand, DuPont rings up twice Merck's sales, but Merck's premium franchise commands twice as much knowledge capital.

The Scoreboard rests on several key assumptions. Normalized earnings, the foundation for computing knowledge capital, are based on three years of historical year-end results through 1997, the last available full year, and three years of projected earnings using IBES International consensus estimates published last October. Market value was computed on May 31, 1998, enough time for the market to digest each company's 1997 performance and anticipate final results in 1998.

PHARMA- CEUTICAL COS.	Knowl- edge Capital*	Book Value	Knowl- edge Capital/ Book Value	Market Value**	Market Value/ Compre- hensive Value	Market Value/ Book Value	Normal Earn- ings*	Sales*	Knowl- edge Capital Earnings/ Sales
Merck	$48,038	$12,614	3.81	$139,910	2.31	11.09	$5,464	$23,637	22%
Bristol- Myers Squibb	30,470	7,219	4.22	106,994	2.84	14.82	3,580	16,701	19%
Johnson & Johnson	29,695	12,359	2.40	92,884	2.21	7.52	3,681	22,629	14%
Pfizer	23,890	7,933	3.01	136,846	4.30	17.25	3,066	12,504	20%
Amer. Home Products	22,822	8,175	2.79	63,392	2.05	7.75	2,342	14,196	17%
Abbott Labs	19,558	4,999	3.91	56,631	2.31	11.33	2,373	11,883	17%
Eli Lily	16,505	4,646	3.55	67,968	3.21	14.63	1,943	8,518	21%
Warner Lambert	12,099	2,836	4.27	52,237	3.50	18.42	1,362	8,180	16%
Pharmacia & Upjohn	4,725	5,538	0.85	22,447	2.19	4.05	792	6,710	7%
ICN Pharma- ceuticals	1,158	796	1.45	3,092	1.58	3.88	146	752	16%
Watson Pharma- ceuticals	1,110	565	1.96	3,899	2.33	6.90	137	338	35%
Allergan	1,053	841	1.25	2,753	1.45	3.27	149	1,149	10%
Mylan Labs	972	744	1.31	3,666	2.14	4.92	147	555	19%
Rexall Sundown	766	192	4.00	2,392	2.50	12.48	89	263	31%
Alza	622	301	N/M^	4,181	N/M	13.88	81	464	14%
Forest Labs	553	614	0.90	2,653	2.27	4.32	82	427	14%
NBTY	386	117	3.30	976	1.94	8.34	43	281	15%
Barr Labs	376	156	2.41	909	1.71	5.83	49	377	11%
Perrigo	254	426	0.60	821	1.21	1.93	52	845	3%
Agouron Pharma- ceuticals	152	236	0.64	1,049	2.70	4.44	30	467	3%
Mean	$10,760	$3,565	2.45	$38,285	2.36	8.85	$1,280	$6,544	16.2%
CHEMICAL COS.	Knowl- edge Capital*	Book Value	Knowl- edge Capital/ Book Value	Market Value**	Market Value/ Compre- hensive Value	Market Value/ Book Value	Normal Earn- ings*	Sales*	Knowl- edge Capital Earnings/ Sales
DuPont	$26,422	$11,270	2.34	$86,962	2.31	7.72	3,641	39,911	7%
Dow Chemical	10,168	7,675	1.32	21,809	1.22	2.84	1,540	20,065	5%
Monsanto	6,021	4,104	1.47	33,166	3.28	8.08	751	7,514	8%
Union Carbide	3,403	2,348	1.45	6,813	1.18	2.90	514	6,502	6%
Air Products & Chemicals	3,234	2,648	1.22	10,189	1.73	3.85	519	4,638	7%
Praxair	3,116	2,197	1.42	7,782	1.46	3.54	411	4,735	7%
Rohm & Haas	2,612	1,797	1.45	6,669	1.51	3.71	398	3,999	7%
Hercules	2,310	690	3.35	4,196	1.40	6.08	307	1,866	13%
Eastman Chemical	2,296	1,753	1.31	5,289	1.31	3.02	361	4,678	5%
Millennium Chemicals	2,062	1,464	1.41	2,444	0.69	1.67	233	3,048	7%
IMC Global	1,489	1,936	0.77	3,714	1.08	1.92	235	2,989	5%
Lyondell Chemical	1,473	619	2.38	2,448	1.17	3.95	198	2,878	5%
Morton Int'l	1,354	1,514	0.89	3,887	1.36	2.57	248	2,574	6%
RPM Inc.	1,318	567	2.32	1,704	0.90	3.01	128	1,615	9%
Cytec Industries	1,167	387	3.01	2,222	1.43	5.74	138	1,291	10%
Nalco Chemical	1,134	653	1.74	2,482	1.39	3.80	149	1,434	8%
Great Lakes Chemical	965	1,307	0.74	2,362	1.04	1.81	176	1,311	8%
Betzdear- born	886	452	1.96	1,460	1.09	3.23	83	1,295	7%
Cabot	776	728	1.07	2,261	1.50	3.11	137	1,630	5%
Lubrizol	751	815	0.92	1,965	1.25	2.41	130	1,674	5%
Witco	607	644	0.94	2,184	1.75	3.39	65	2,187	3%
Ethyl	542	145	N/M	589	N/M	4.08	72	1,064	5%
W.R. Grace	473	468	1.01	1,409	1.50	3.01	78	1,480	3%
Ferro	333	273	1.22	1,071	1.77	3.92	53	1,381	3%
H.B. Fuller	179	339	0.53	876	1.69	2.58	42	1,307	1%
Geon	118	224	0.53	504	1.47	2.25	31	1,250	1%
Wellman	25	634	0.04	749	1.14	1.18	46	1,083	0%
Mean	$2,786	$1,765	1.42	$8,045	1.45	3.53	$396	$4,644	5.8%

The various measures of knowledge assets included in this survey were designed by Prof. Baruch Lev, who retains exclusive trademark rights to the measures. These knowledge-based measures should not be used in any form without the written permission of Professor Lev. Marc Bothwell, vice president of BEA, a member of Credit Suisse Asset Management, contributed to the development of the measures and performed the required programming and computations.