How Much Cash Does Your Company Need?

The value of intangible assets is highly dependent on a company’s own ability to fund those assets, while the value of tangible assets is independent of the company.

More than you think—a lot more—if yours is a knowledge-based corporation. That’s because you need to provide for intangible liabilities—the investments a company has to make to realize the benefits of its knowledge.

After its merger with rival Warner-Lambert in 2000, New York-based pharmaceutical giant Pfizer found itself sitting on a net cash position approaching $6 billion. That seemed extraordinarily conser­vative for a company whose products generated close to $30 billion in reve­nues. Those products included some of the world’s best-selling drugs. The anti­-cholesterol blockbuster Lipitor alone generated worldwide revenues in excess of $7 billion in 2001.

Most large companies with revenues that healthy would increase their lever­age, or the amount of debt they carry, thereby unlocking tremendous value for their shareholders, both from tax benefits and from the market’s well­-documented perception that managers with less money to spend will spend it more wisely. Consider Bank of America. Its capital structure, like that of most banks, relies heavily on debt. The value of the tax shields alone accounts for approximately one-third of the com­pany’s $120 billion market capitaliza­tion. But is this kind of strategy appro­priate for a knowledge-based company like Pfizer? To answer that question, Tim Opler of Credit Suisse First Boston and Richard Passov undertook an in-depth study of the knowledge-based businesses that were most closely comparable to Pfizer. There, we saw a rather different picture: The world’s largest and most successful technology and life sciences companies were consistently holding significant net cash positions.

Like Pfizer, these companies had mar­ket valuations that were much greater than the value attributable to their on­going businesses, a premium that re­flects these companies’ ability to create new products through R&D. And like Pfizer’s, these companies’ assets were very risky—a fact often obscured by the companies’ balance sheet structure. Pfizer shares, for example, had approxi­mately the same price volatility (30%) as those of Bank of America in 2001. But Pfizer had a negative leverage ratio of 0.3:1 where Bank of America had a ratio approaching 10:1. If the equity volatili­ties are adjusted to eliminate the effect of the two companies’ balance sheet structures—treating both companies as if they were wholly equity financed and had no cash—we see that Pfizer has an underlying asset volatility of close to 30% while Bank of America has a vola­tility closer to 5%. Because of this higher underlying volatility, Pfizer and the other knowledge companies we looked at were in a group apart from other large corporations.

We believe that these companies’ decisions to run large cash balances is one of the key factors in sustaining the value of their intangible assets—which typically comprise a substantial portion of overall valuations for knowledge companies. Only by consistently invest­ing in their intangible assets can knowl­edge companies hope to preserve the value of those assets. A company that finds itself unable to meet such com­mitments because unfavorable market conditions reduce its operating cash flows will find its share price suffering almost as much as if it were to default on its debts. By the same token, with the right balance sheet, knowledge com­panies can profitably insure against the risk of failing to sustain value-added in­vestments in difficult times. An optimal capital structure that calls for significant cash balances is at odds with the results of a traditional capital structure analysis but explains the financial policies of many well-run knowledge companies.

Pfizer’s intangible assets are the prod­uct of heavy ongoing investment. The company’s R&D alone consumes about $7 billion a year. What’s more, the pro­ductivity of the company’s research sci­entists depends on maintaining a vast, interconnected IT infrastructure. Such investments are not treated as liabili­ties in a traditional capital structure analysis. To see why that’s a problem, consider what would happen if Pfizer found itself in a situation where inter­nally generated funds could no longer sustain R&D. Finance theory maintains that the market will always be willing to provide funds for a good investment opportunity. Based on that reasoning, companies with promising pipelines should always be able to find funding for R&D. History has shown, however, that in times of need, external financing can be exorbitantly expensive or simply unavailable for knowledge companies.

Intel experienced just such a funding crisis in the early 1980s. At that time, the company’s 80286 microprocessor chip had just emerged as the key hard­ware component for IBM’s burgeoning personal computer business. Unfortu­nately, Intel was at the same time in­volved in producing DRAM memory chips, which were becoming commodi­tized due to competition from Japanese manufacturers. Intel’s cash position dwindled and its debt rose to the point where the company was unable to make the capital expenditures necessary to complete the development of its micro­processors. Intel was forced to raise new equity capital from its primary business partner, IBM, which purchased 12% of the company for $250 million. The cap­ital infusion allowed Intel to continue its R&D program and build manufac­turing plants for new microprocessors while simultaneously sustaining a costly transition away from the DRAM market.

Arguably, Intel’s inability to meet its R&D commitment cost its shareholders as much as a debt default would have. Like a debt crisis, the funding crisis had forced the company into financial dis­tress. Had Intel not shrewdly repur­chased the stake in the late 1980s, by 2001, IBM would have earned a 100-fold increase on its initial investment.

Of course, Intel and Pfizer are not the only companies that have large capital commitments. Oil companies, for exam­ple, spend huge amounts of money on exploration and development. Yet they often use more leverage and seem less vulnerable to the whims of the capital markets. The difference in financial strategies seems to lie in two important distinctions between tangible and in­tangible assets. These factors also ex­plain the relatively high asset volatility of Intel, Pfizer, and other successful knowledge-based companies.

Intangible assets are company de­pendent. The value of tangible assets—even those that require considerable investments to exploit—is usually widely recognized by outside investors. An oil reserve, for example, has a generally agreed-upon value, regardless of the company that owns it. Energy giant ChevronTexaco has to spend billions to exploit its reserves. But because the value of those reserves can be estimated and easily communicated, the company usually can find the money to fund de­velopment regardless of the state of its finances. By contrast, the value of a com­pany’s intangible assets is typically un­derstood only by the company itself or its close partners. If the company fails to invest in maintaining the value of its intangible assets, no one else is likely to volunteer. In other words, the value of intangible assets is highly dependent on a company’s own ability to fund those assets, while the value of tangible assets is independent of the company. In Intel’s case, nobody had any idea in 1983 what a huge market microprocessors would become, which explains why the company could obtain funding only on expensive terms from IBM. This experi­ence cast into sharp relief the value of holding cash reserves, and Intel went on to build a strong balance sheet to pro­vide insurance against potential future funding needs.

Intangible liabilities cannot be hedged. The second reason intangible assets are different from tangible ones is that the risk that a company will be unable to meet commitments on intan­gible assets cannot be easily hedged. The value of an oil company’s explora­tion and development budget is subject to the variable market pricing of oil; a sharp fall in oil prices reduces the value of those projects. However, the risk of a fall in oil prices can be hedged in the fi­nancial markets, which allows the com­pany to preserve the value of its explo­ration and development projects even when business conditions deteriorate. Even if it decides not to protect its ex­ploration and development projects in this way, the company enjoys a natural hedge because the price of oil is highly correlated with cash flows. When the firm’s cash flows are low, so is the ex­pected value of exploration and devel­opment. That means the company is most likely to be short on funds when it least needs to spend the money on the asset. By contrast, a knowledge com­pany’s primary risk—whether or not the various molecular compounds in its pipeline will react as hoped, for ex­ample—is impossible to hedge in the financial markets. That risk is also un­likely to be correlated with the com­pany’s cash flows. A particular molecule may react as planned, but the company may run out of funds before discovering that. Unlike an oil company, therefore, a pharmaceutical company may face a funding crisis just when the value of continuing its research is highest. The only way to manage that risk is to en­sure that the company always has on hand enough liquid assets—essentially, cash—to meet its R&D liabilities.

If the defining characteristic of a lia­bility is that the company’s inability to meet it triggers financial distress, then it is only logical that R&D expenditures of Intel, Pfizer, and their peer companies should be considered liabilities—just as inescapable a commitment as if they were a debt obligation. Companies do not currently treat R&D and compara­ble investments in intangible assets as balance sheet items. But from an economic perspective, they probably should. After all, if the market is placing a value on the promise of future success in drug discovery, for example, it is also expecting that the resources necessary to discover these drugs will be available to spend as needed.

Let’s look at the numbers in more detail. The tax benefit, or tax shield as it is usually called, is determined by the corporate tax rate: Simply multiply the amount of debt by the marginal corpo­rate tax rate. Arriving at the cost of de­faulting on debt is a bit more compli­cated: The probability that the company will not be able to meet its debt obliga­tions is multiplied by the likely impact of that default, should it happen, on the company’s value. One practical way to estimate the probability of default is to look at the historical volatility of a com­pany’s cash flows. From that, you can determine through statistical analysis how frequently a company’s cash flows are likely to be less than the level of in­terest payable for a given level of debt. Obviously, the greater the size of the in­terest bill, the higher that probability. If a company’s historical cash flows are not available, analysts can use industry data or empirical studies that provide default rates. Debts rated Aaa by Moody’s, for instance, have historically had a 0.1% chance of defaulting within five years, a Baa rating a 1.8% chance, and a B rat­ing a 32% chance. A more ambitious an­alyst could also use default rates implied by an analysis of spreads ort corporate bonds or credit derivatives.

The impact of default can be esti­mated by looking at the empirical data. Research shows that a typical company will lose roughly 20% of its enterprise value (market value of the company’s shares plus the value of its debt less cash) in times of financial distress. Multiplying this number by the probability of default determined above produces the expected cost of default, or the amount a company would want to in­sure itself for, if it could. So for an aver­age company with a default probability of 5% over a five-year horizon, the ex­pected cost of default, over that same time period, would be 1% of the firm’s enterprise value.

The underlying assumption in using this method of calculating optimal cap­ital structure is that a company’s debt level is the principal determinant of whether or not a company will suffer financial distress. But as we’ve argued, a company can lose just as much value, if not more, if it cannot fund the intangible liabilities associated with its in­tangible assets. In other words, financial distress costs can kick in even while a company has a net-cash position. On that basis, the calculation needs to be adjusted. First, the probability of de­fault must be adjusted to encompass the probability of distress: It should be determined not by the probability of interest costs exceeding cash flow but by the probability of interest costs plus R&D expenses (or other capital expen­ditures) exceeding cash flow. Similarly, the impact of default must be adjusted to reflect the fact that the value of in­tangible assets tends to be much more volatile than that of tangible ones and their higher volatility exposes the com­pany to greater financial risk.

Companies may wish to make even more precise estimates. At Pfizer, for instance, an analysis of our historical cash flows allows us to quantify the im­pact of losing patent protection on cur­rently marketed products. Another key risk we consider is the possibility that a drug might be withdrawn from the mar­ket due to safety concerns. To account for that, we use industry data to model the likelihood of an approved drug being pulled from the market.

Running the numbers for Pfizer, we found that the company’s optimal capi­tal structure—the structure that maxi­mized the company’s value—called for holding a positive net financial balance, as shown in the exhibit "Pfizer’s Optimal Capital Structure." That stands in sharp contrast with the optimal capital struc­ture predicted by a conventional ap­proach—a net debt position entirely inadequate to maintain the value of Pfizer’s intangible assets.

The bulk of ChevronTexaco’s capital commitments consist of exploration and development expenditures. As we have pointed out, development costs can be easily hedged. Exploration, how­ever, counts as a true liability. To calcu­late the probability of distress at various levels of leverage, therefore, we looked at the volatility of ChevronTexaco’s cash flows and applied that to the combined average historical interest and explo­ration costs. To estimate the impact of financial distress, we looked at our em­pirical data, which showed that oil com­panies typically lose about 20% of their enterprise value in times of distress. Our analysis indicated an optimal net debt level of approximately $10 billion for the company. This compares with ChevronTexaco’s actual net debt posi­tion of $12 billion as of year-end 2002.

A conventional model based on the weighted average cost of capital ap­proach would suggest that Chevron­Texaco should have debt in excess of $20 billion. Clearly, our model does a better job of explaining this well-run company’s financial policy.

Does our model always find that tra­ditional firms can operate with leverage while knowledge-based firms shouldn’t? Not necessarily. Consider the case of Oracle, the enterprise software giant that had nearly $9.7 billion in revenues in 2002. Approximately 60% of Oracle’s revenues come from licenses associated with its software, while nearly 40% come from product support and consulting services. To counter fierce competition from giants like Microsoft and IBM, as well as newer competitors like SAP, Siebel, and PeopleSoft, Oracle invests considerable sums in R&D, marketing, and training. More than 10% of annual revenues are committed to the research and development of new products. In 200l and 2002, the company spent over $1.4 billion each year on capital expenditures and R&D. To determine the probability of default for various net cash-to-debt scenarios, we used histori­cal cash flow volatility, analysts’ projec­tions for the company’s future debt, and R&D expenditures. To estimate the im­pact of default, we looked at the cost for companies with asset volatility compa­rable to Oracle’s.

Our analysis suggested an optimal net cash position of approximately $1 bil­lion, which contrasts with Oracle’s ac­tual net cash position of around $6 bil­lion. This finding raises some strategic questions. If current predictions about the slowing growth potential in the software industry come to pass, then there may come a time when even Ora­cle shareholders would fare better with a nice, steady dividend and leverage in the company’s capital structure. (It is interesting to note that Oracle’s move to acquire PeopleSoft, were it funded by debt, would bring its net cash position much closer to the optimal level pre­dicted by our model.)

Business strategy and financial strat­egy are inextricably linked. Therefore, companies must develop capital poli­cies in light of their business risks. In­deed, balance sheet management is best viewed as a form of risk manage­ment to be coordinated with the other ways in which companies manage busi­ness and financial risks. Johnson & Johnson’s consumer products business, for example, has had strong and stable operating cash flows, which tend to buffer the potential liquidity requirements of the company’s riskier phar­maceutical business. As a result, John­son & Johnson can afford to have a smaller financial asset position than would a pure-play pharmaceutical com­pany.

Intangible assets are the "dark matter" of the business universe. Since we can only observe them by their effects, it’s difficult to understand and catalog them, which is precisely why the ac­counting balance sheet differs from the economic one. Accountants like to deal in concrete fact, while economists are happy enough with theory. But even ac­countants cannot deny that the effects of intangible assets are far-reaching. The ability of intangible assets to influence the likelihood and degree of financial distress through the liabilities they cre­ate is not the least important of those effects. By suggesting ways to measure the effects, we hope to have made a small contribution toward incorporat­ing intangible assets into capital structure analysis.