The Map and the Territory

by Alan Greenspan

  Moreover, changes in prices of owner-occupied homes affect not only expenditures by households on goods and services, but along with long-term mortgage rates, are a major factor in the decision of homebuilders to construct new homes.9 Homebuilding, however, does not appear to be significantly affected by stock prices, with the possible exception of the stock price of a homebuilder’s own company.

  GOVERNMENT RESPONDS TO CAPITAL GAINS

  Asset prices have a measurable effect not only on consumption expenditures and private capital investment but also on government outlays as well. Municipalities that depend heavily on property taxes do ratchet spending up and down as the market values of assessable properties fluctuate. As federal budget receipts rise from taxed capital gains and stock option grants, the fiscal status of the government improves, creating leeway for Congress to spend more. The dot-com boom, for example, produced the budget surpluses of 1998 to 2001 for the federal government and many states. No elected officials in modern times seem to resist being enticed by unexpected surplus cash sitting around, uncommitted. But outside of the extremes of boom and bust, the effect of more workaday fluctuations of asset prices on ups and downs in federal spending is surely minimal.

  In summary: In Exhibit 4.6, I trace the overall effect of year-over-year changes in equity values (with a one-quarter lead) on year-over-year changes in real GDP.10 I find that between 1970 and 2012, a 10 percent increase in the market value of holdings of U.S. residents of stocks, bonds, homeowners’ equity, and other assets is associated with an annual change in real GDP of 1.3 percentage points.

  EQUITY STIMULUS

  Since 1952, the combination of equity from stocks and homes has averaged a net annual gain of 7.5 percent.11 The American economy thus has been “stimulated” by equity in a manner similar to the way fiscal stimulus affects GDP. Both, of course, require debt to convert the programs into spendable funds. Government deficits are funded by government debt. Private realized capital gains cannot be monetized and spent without incurring private debt or issuing stock. But there are, of course, important differences. Fiscal stimulus is initiated by deliberate action of government. Equity stimulus is a result of private sector decision making, the economic consequences of which can be affected by government actions for good or ill. In Chapter 7, I compare the contribution of fiscal stimulus to the economic recovery of 2009 with that of equity stimulus.

  CAPITAL EXPENDITURES

  I was a director of fifteen listed corporations at various times in the quarter century before I joined the Federal Reserve. I cannot remember a single instance when a chairman12 presented a new capital project to his board and cited the corporation’s rising stock price as a determining factor for authorizing a capital expenditure. Yet the data unambiguously indicate that the market value of corporate equities (stock prices) is an important determinant of capital spending (see Box 4.1).

  BOX 4.1: STOCK PRICES AND CAPITAL SPENDING

  In a paper I published in 1959,13 I related the ratio of the market value of existing U.S. corporate net assets (stock prices) to the replacement cost of those assets (the price of newly produced private plant and equipment).14 In the simplest example of this process, a new office building will tend to be funded if the market value of contiguous office buildings significantly exceeds the expected construction cost of the new investment. (See “Asset Prices,” page 80.)

  The ratio of stock price to cost of construction of capital assets correlated quite well with machinery orders (capital investment) going back into the 1920s. I recently updated the 1959 analysis and was amazed at how well this simple relationship still works, even tracing the recent years’ sharp fluctuations in real private capital investment (Exhibit 4.7). Since 1993, for example, a 10 percent change in stock prices relative to the cost of replacing plant and equipment from scratch has been associated with a 4 percent change in real capital expenditures relative to the stock of fixed assets.

  I know of no corporate executives who explicitly determine a corporation’s total capital budget based on such calculations, but implicitly they all do.

  HOW IT’S DONE

  In theory, to optimize the long-term market value of a firm, the ultimate goal of corporate investment, corporate executives need to simultaneously consider the entire universe of potential investments: their prospective rates of return and whether they should be funded wholly by equity, wholly by debt, or some combination of the two. This implies, for example, that all capital projects be identified and evaluated for any year on January 1, and then management continually reevaluate throughout the year all the assumptions that went into the initial conclusions as new evidence inevitably emerges. But no corporate executive committee has the ability to go through that exercise.

  In practice, as a rough approximation of the theoretically optimal procedure, most corporations constrain capital spending to the level of cash flow, adjusted for the preferred degree of leverage (or de-leverage) of their balance sheet. These choices reflect management’s general level of confidence about the future over the time frame of prospective capital investments. Of course, all capital budgets are made up of myriad individual investment projects that must be evaluated.

  FROM GUT TO REASON

  The cap-ex ratio (capital expenditures as a share of cash flow) captures the essence of business investment decision making. But how do corporate managers make such judgments? In years past, it was too often the “gut feel” of the CEO (not to be probed further). In more recent decades, the process has become more formal, though gut feel (intuition) has never been fully discarded (see Chapter 1).

  The investment process obviously differs from company to company, but all, especially larger corporations, follow more or less the same approach. For example, the executives of an oil company contemplating whether to expand capacity may instruct the company’s technical and marketing people to create a best estimate of the potential future market and profitability of a newly introduced petrochemical feedstock over the expected life of the proposed production facility.

  If that rate of return exceeds the corporation’s cost of raising new equity capital and is within the bounds of any overall corporate leverage constraint,15 the analysis then proceeds to stage two. What is the expected range, or variance, of best estimate outcomes? This judgment, as I have observed over the years, largely determines whether the investment is implemented. An investment with a respectable 20 percent average annual expected rate of return could nonetheless be rejected if the long-term outlook is so cloudy that the range of that estimate is, for example, between negative 20 percent and positive 60 percent.

  The fewer the variables that need to be evaluated in making a forecast, the narrower the range of possibilities and the lower the variance. A clouded business climate owing to an uncertain outlook of future tax regimes, for example, obviously increases investment outcome variance by adding new uncertainties to the decision-making process. That can, and usually does, exact a toll on the level of a company’s capital investments. Some uncertainties can be ameliorated by slightly altering the nature and terms of a project. While the calculation of the expected rate of return is straightforward, the determination of the level of variance that is acceptable is not. In my experience, the relative skills in making variance judgments determine to a large extent which companies are most successful.

  The ratio of capital investment to cash flow is thus not only an important measure of the level of corporate confidence, but it is also a useful indicator of corporation leverage. The acceptable degree of overall corporate leverage is revealed in the choice of total capital expenditures’ share of cash flow (the cap-ex ratio).16 A corporation that finances capital expenditures wholly with cash flow, that is, with no net borrowing, has a ratio of 1.0.17 A ratio exceeding 1.0 defines the degree of increase in leverage. A ratio less than 1.0 indicates the degree of de-leveraging. For nonfinancial corporate business between 1952 and 2007, for example, the (unweighted) average of that ratio was 1.05, with all of the
annual observations falling between 0.82 and 1.29 (excluding 1974). Thus, some net borrowing, on average, is associated with corporate investment. Corporate managers are always aware of the value of the equity buffers that stand between debt and bankruptcy as they set the appropriate amount of leverage that their corporation will take on.

  CORPORATE CULTURE

  During my tenure as a director of JPMorgan (just prior to joining the Federal Reserve), I was impressed by the value the bank accorded to its AAA rating. They recognized that in the short run, they could achieve a higher return on equity through increased leverage. But they feared that that could lower the bank’s AAA rating, an important factor in their long-term ability to attract low-cost liabilities. Most important, the rating was required to sustain a reputation for prudence, an essential characteristic of their historic franchise that dated back to the time of John Pierpont Morgan himself. Similar considerations led to constrained leverage on the part of many nonfinancial corporations for which I have worked over the years.

  Most firms will borrow only so much, even though they recognize they may be forgoing profitable opportunities by failing to more fully leverage certain investments (a reluctance far greater than is evident among financial firms). This business reluctance to borrow beyond a determined limit relative to cash flow is evident in Exhibit 7.2. Because investment less cash flow is necessarily equal to net borrowing, the data indicate a relatively narrow range of leverage. This is true for business in general. If at any interest rate or level of cultural restraint business borrows less, governments are free to borrow more.

  Regression analysis indicates that higher mortgage rates suppress home prices (Exhibit 3.3), and home prices are critically tied to home investment (Exhibit 7.4). That accounts for about a fourth of crowding out of private investment. I suspect that interest rate suppression of less than investment grade companies accounts for a significant additional amount. But in general it appears that roughly half of the displacement of gross private savings by budget deficits can be attributed to interest rate crowding out, and the residual other half to corporate culture restraints.

  Both crowding out owing to elevated interest rates and crowding out owing to corporate culture restraints in the end come out in the same place. Elevated costs of funding will often render a relevant prospective investment unprofitable. But that is also what occurs when an individual company self-limits its degree of leverage. Even triple-A firms restrict their debt issuance, and that implies that certain contemplated investments are stillborn.

  A company with very little debt has considerable leeway to leverage and maximize the rate of return on equity. It can thus accept a relatively low potential operating rate of return18 and, through leverage, bring the return on equity up to a satisfactory level. A firm with an already high degree of leverage does not have that opportunity. Since it cannot borrow, it needs a higher operating rate of return. Therefore, other things equal, an already highly leveraged firm will invest and borrow less than a firm not so debt encumbered. In the end, it behaves like an unleveraged firm inhibited by a rise in interest rates.

  To be sure, that’s not the way investment judgments are made in the textbooks on optimum capital allocation. Human propensities too often warp objective decision making. But most large firms do adhere to a rigorous paradigm. In my experience, most medium- and, especially, smaller-sized companies do not. When they reach a certain debt level, they behave as though they were crowded out by higher rates. A model that fails to embody such economic forces (and most do not) misses an important constraint on the forecast results.

  FIVE

  FINANCE AND REGULATION1

  In recent years, the ongoing debate on the merits of capitalism has homed in on a crucial pillar: Adam Smith’s fundamental premise of free markets, the idea that people acting in their own self-interest spur competition that advances society as a whole.2 In the purest form of that paradigm is the idea that markets are essentially self-regulating.

  Although I always have been, and remain, a strong supporter of free market capitalism, my support is not based on the belief that all market participants behave solely in their own rational self-interest at all times. As someone brought up in the canyons of Wall Street, I saw too much of what we now describe as the influence of animal spirits to entertain such a view.3 I nonetheless found the broad success of free markets, with all their shortcomings, intellectually and empirically too compelling to ignore and the arguments for alternative economic systems flawed and unpersuasive.

  I believed in years past that the aberrations from rationality and efficiency—often reflecting the effects of animal spirits—were sufficiently infrequent and random to evoke little more than economic noise. I was thus especially shaken by the breakdown in 2008, which could scarcely be characterized as economic noise. Moreover, it had an eerie historical familiarity about it—1929 and 1907 for example. Did these breakdowns reflect “noise” or some systematic propensity of human nature? The crack in my view of the economic world manifested itself in a widespread failure of one of the most important pillars of a stable market economy, whose roots lay deep in postwar academia: rational financial risk management.

  I noted in an op-ed piece early in the crisis, “those of us who look to the self-interest of lending institutions to protect shareholder equity have to be in a state of shocked disbelief.”4 That episode led me to rethink my view of the importance of animal spirits. And in the aftermath of the failure of financial risk management, I was led to conclude that a tightening of regulatory capital standards was long overdue.

  The events of the crisis demonstrated that financial managers could not be counted on to maintain an equity buffer adequate to withstand a broad range of economic outcomes. The willingness of many financial firms to allow their tangible capital, at the height of the boom, to become razor thin was a folly largely explained by herd behavior and an underestimation of the ephemeral nature of market liquidity. The financial and economic instability that followed in its wake induced a sharp political response toward increased regulation, most notably in the form of the Dodd-Frank Act. In political terms, that response was understandable. The problem is that substituting government regulations is rarely an improvement and, indeed, as I note later, if overdone, turns out to be counterproductive. As poorly as certain private financial managers performed leading up to the crisis of 2008, government regulators fared no better (see Chapter 2).

  FINANCIAL INTERMEDIATION AND REGULATION

  The Purpose of Finance

  The role of regulation in a market economy is determined by the nature of what is being regulated. The ultimate goal of a financial system in a market economy, for example, is to direct the nation’s savings (including depreciation5), plus any savings borrowed from abroad (the current account deficit), toward investments in plant, equipment, and human capital that offer the largest risk-adjusted returns on capital and presumably the greatest increases in the nation’s output per worker hour. Nonfinancial output per hour, on average, rises when obsolescent facilities (with low output per hour) are replaced with facilities that embody cutting-edge technologies (with high output per hour). This process improves average standards of living for a nation as a whole. There is no alternative to what Joseph Schumpeter aptly called “creative destruction.” In the United States, the success of finance through the last decades of the twentieth century in directing our scarce savings into real productive capital investments may explain the generous compensation that nonfinancial market participants had been willing to pay to the domestic producers of financial services (see Box 5.1).

  BOX 5.1: COMPENSATION

  Prior to my Federal Reserve tenure I spent a good deal of my time as a director of a number of American corporations, large and small, including two banks and a savings and loan holding company. In The Age of Turbulence (page 426), I complained about the consultants hired by corporations to advise boards of directors on corporate compensation packages. I was distressed that all seemed
to be arguing that boards of directors needed to offer above-average compensation. It is a neat trick to get everyone doing better than the average of all. At the same time, it has been my experience serving as a consultant and, at times, as a director of many large financial corporations that small differences in the skill level of senior bankers tend to translate into large differences in the bank’s bottom line. Competition for even the slightly more skilled is accordingly fierce. Senior bankers operate as largely independent entities whose “clients” are often more theirs than the banks’; they leave with the “star” when he or she changes organization. It is doubtful that legislation can work in such an arena. My experiences in such issues have made me skeptical, and data do indirectly support such skepticism. If directors, in seeking bank executives, believe that senior officers of the bank contribute to the bank’s bottom line, you would expect the bank’s executive compensation to mirror the market value of the bank. The larger the bank, the greater the dollar gains and losses from management decisions. I have not tried to match the figures bank by bank, but the aggregate CEO compensation of the S&P 500 corporations, a good proxy, does exhibit a surprising stability over time with the market value of the firms (Exhibit 5.1).