Leer en español
Ler em português
As a guide to the health of a company, operating cash flow data have a great vogue these days among those who watch the fortunes of corporate America from the outside—especially securities analysts. Moreover, financial executives of businesses increasingly prefer a cash-basis assessment of available funds over the traditional working capital status. Apparently speeding the trend is action by the Financial Accounting Standards Board.
How good a yardstick is operating cash flow? Not very, say Messrs. Casey and Bartczak. They studied 290 companies, 60 of which had been declared bankrupt, and found that operating cash flow data for a five-year span could not distinguish between the healthy enterprise and the one that would fail. The OCF measure was less accurate a predictor of failure than a combination of six conventional accrual-based measures, including debt-to-equity and profitability ratios.
A growing number of securities analysts, financial writers, and accounting policymakers contend that financial statements providing information of a company’s cash flows yield a better measure of operating performance than do the company’s income statement and balance sheet. According to recent surveys, corporate and government officials have accepted this view; they rated cash flow data the most important piece of information contained in published financial statements.
The trend toward wider acceptance of this yard-stick has been building since the early 1970s. Accelerating the trend have been several developments—including new financial reporting rules on such issues as foreign currency translation, equity earnings, interperiod income tax allocation, and lease and interest cost capitalization—that put greater distance between a company’s net income and its cash flow; the adoption of “liberal” accounting practices by some companies; and record inflation levels.
To many, the collapse of Penn Central and the W.T. Grant Company proved that traditional accrual accounting-based data had limited value in alerting investors to important changes in a company’s financial condition.
Accordingly, securities analysts have come to view cash flow information as a more accurate yardstick for gauging debt and dividend-paying ability. Corporate executives have penetrated the veil of accounting profits, have found them sometimes misleading, and have turned to the “real thing,” cash flow data, to evaluate their company’s performance and that of competitors.
Since 1981 the Financial Executives Institute has been encouraging companies to voluntarily report cash flows in their statements of changes in financial position, that is, “funds flows” statements. Preliminary results of an FEI-sponsored study on the structure and use of the funds flows statement suggest a growing preference for a cash-basis definition of funds over the traditional focus on working capital. As of the end of 1983, more than 750 financial executives, financial analysts, individual investors, and commercial bankers had responded to the FEI study. Of the total, 57% said they intend to use the cash flow basis in the statement, while 23% plan to use the working capital basis. Only 27% used the cash flow basis in 1980, while 52% used the working capital basis.1
Furthermore, in November 1981 and December 1983, the Financial Accounting Standards Board issued exposure drafts proposing that every corporate financial statement include information on cash flows during the particular period. In explanation the FASB contended that “the greater the amount of future net cash inflows from operations, the greater the ability of the enterprise to withstand adverse changes in operating conditions.”
While we applaud attempts to glean better information on corporate past and future performance, we fear that operating cash flow may come to be regarded as the barometer for gauging company performance. The growing legion of supporters of operating cash flow—which has great intuitive appeal—would be hard pressed to produce objective evidence of its superiority. The most compelling proof of the usefulness of OCF data that we have discovered is a study of W.T. Grant that appeared in the Financial Analysts Journal. The authors found that Grant’s operating cash flow was much more accurate and timely as an indicator of its impending bankruptcy than were traditional financial ratios or movements in Grant’s stock price.2
This one finding, while provocative, does not substitute for a broad-based study of a possible relation between the level of operating cash flow and future financial condition. A study we have made of nearly 300 companies raises serious doubt about the reliability of operating cash flow as a financial indicator.
Bankrupt or Healthy?
We selected 60 companies that had filed petitions for bankruptcy during the period 1971–1982 and matched them with 230 viable (or at least “nonbankrupt”) companies chosen at random from similar industry groupings on the Compustat Industrial Tape. We calculated three variables, operating cash flow (OCF), operating cash flow divided by current liabilities (CL), and operating cash flow divided by total liabilities (TL). OCF has a serious drawback as a measure of potential financial distress because it disregards size-of-business considerations as well as any unused borrowing capacity. CL and TL offset this drawback by relating OCF to a company’s level of short-term and long-term indebtedness, respectively.
What did we find? We found that none of the variables could discriminate between the bankrupt and nonbankrupt companies with reasonably good accuracy. In fact, overall accuracy for OCF was only slightly better than chance (50%) for the first and second years before failure and was worse than chance for the remaining years. Exhibit I shows the percentages of accurate classifications.
Exhibit I Percentage accuracy using OCF, CL, and TL
So, operating cash flow data are not the Holy Grail that some have made them out to be. Furthermore, no one number can accurately and consistently predict performance; many factors affect a company’s well-being.
One would think that the operating cash flow measure would have utility in predicting bankruptcy, since an impending collapse usually sends clear signals. But the OCF figures over five years before filing for bankruptcy were a poor predictor of failure and less good than a combination of six conventional accrual-based financial ratios (net income/total assets, cash/total assets, current assets/current liabilities, net sales/current assets, current assets/total assets, and total liabilities/owners’ equity).
Perhaps more important, the cash flow numbers failed to improve predictive accuracy when we analyzed them together with the accrual-based ratios. That is, the OCF data did not have even marginal value.
For our research we defined OCF as working capital provided by operations plus or minus changes in the noncash working capital accounts, except for short-term debt (seasonal bank loans, nontrade notes payable, and the current portion of long-term debt). This is also the way the FASB defines it. The insert supplies further information on the nature of cash flows.
Because it includes changes in working capital accounts, however, OCF differs sharply from the more traditional definition of cash flow used by many analysts: net income plus depreciation and other operating items that do not affect working capital. This definition excludes changes in current operating accounts that may greatly affect a company’s operating cash flow.
We limited the analysis to OCF figures that could be calculated from the companies’ published financial statements and from information filed in the event of bankruptcy. The failed companies represented a range of sizes and industry classifications and included, besides W.T. Grant, such well-known names as AM International, Braniff, McLouth Steel, Wickes, and Saxon Industries. Financial data for the healthy companies spanned the same period as for the failed companies.
The poor predictive ability shown in Exhibit I is due to the large number of inaccurate classifications of nonbankrupt companies as failures because they do not generate much operating cash flow. This finding reinforces our concern that too much reliance on OCF may cause investors and creditors to view otherwise healthy companies as financially distressed. Although many companies generate little OCF in some periods, most of them do not go belly up.
Compare the well-known situations of Pan American and Braniff. During these airlines’ severe financial distress in the past few years, their annual operating cash flows were negative. Yet Pan Am has survived, while Braniff filed for bankruptcy in May 1982. Pan Am stayed alive by raising $500 million each from the sale of two large capital assets, the Intercontinental Hotel subsidiary and the Pan Am building in central Manhattan. By concentrating only on its operating cash flows, Pan Am’s creditors might have forced it into bankruptcy. Braniff lacked comparable salable assets on its balance sheet, or it too might have lived.
Massey-Ferguson and International Harvester are additional cases of survival despite very poor operating cash flows for a long time. The most important factor here has been the willingness of Massey’s and Harvester’s creditors to renegotiate and restructure their indebtedness despite the insufficiency of their operating cash flows to service it. The creditors probably view the potential costs of bankrupting these vast multinational companies as greater than the costs of continuing to extend their debt.
It might be argued that focusing on the negative operating cash flows of Pan Am, Massey, and Harvester at least would have signaled their coming financial straits. By this standard, however, most growth companies would have to be called financially distressed. Growing companies, in their efforts to take advantage of market opportunities and gain market share, often cannot generate positive operating cash flows as they build receivables and inventories. Yet these companies usually have little difficulty meeting their operating cash shortfalls as long as creditors and equity investors share the perception of their growth potential.
Even mature companies may suffer operating cash flow difficulties without becoming endangered. Cyclical companies are a prominent case in point. These enterprises often invest cash to build inventories well ahead of the anticipated peaks in their operating cycles. As a result, their cash flows may appear depressed even though they are running their affairs properly.
By contrast, some companies may be regarded as financially strong because they report large, positive operating cash flows. This condition, however, may be the result of a decision not to reinvest in their businesses and to “harvest” the cash that mature businesses often generate late in their life cycles.
Finally, some companies have off-the-balance-sheet cash resources to exploit. A growing number of corporations, for example, have petitioned the Pension Benefit Guaranty Corporation to terminate and, in most cases, replace their overfunded defined-benefit plans. Two recent petitioners have been A&P, a financially troubled company, and Occidental Petroleum, a healthy concern.
Accrual-based measures
To get a more reliable yardstick than operating cash flow, we constructed a model using a standard statistical tool, multiple discriminant analysis, that allowed us to compare its predictive accuracy with that achieved via CL and TL as well as OCF. The model contained the six conventional accrual-based financial ratios mentioned earlier.
What Is Cash Flow?
The simplest way to view cash flow is to define it as the difference in the cash balances of a company on two dates. For instance, a company that has a cash balance of $1 million at December 31, 1982 and $2 million at December 31, 1983 has had a net cash inflow of $1 million during 1983. This view, however, provides no information as to how the net inflow occurred. Did it arise via operations, the sale of assets, an increase in debt, the sale of stock, or some combination of these factors?
To answer this question, publicly owned companies accompany their financial statements with statements of changes in financial position. A statement of changes gives information on the company’s important investing and financing decisions with a focus on how such decisions affected its liquidity. When the accounting rule-makers mandated the statement in 1971, working capital was considered a good measure of a company’s liquid position. Indeed, at that time most financial analysts, bank lending officers, and others were simply using either net income plus depreciation or working capital provided by operations as a surrogate measure of a company’s “cash flow.” Unfortunately, this misnomer has stuck, and many in the business community still refer to these measures as cash flow.
Lately, however, a growing number of analysts are adjusting working capital provided by operations for changes in the working capital accounts to arrive at an operating cash flow (OCF) number that more accurately reflects the rise or decline in the cash account from operations. The following example highlights the approach we used to calculate OCF for our study.
In this situation, traditional cash flow is calculated as a $200 net inflow, working capital provided by operations is a $250 net inflow, and OCF is calculated as a $250 net outflow. Which number most accurately reflects this company’s decision to invest in its operating accounts? Obviously, OCF.
We calculated a score for each company on the basis of which we classified it as either bankrupt or viable. The score is computed by multiplying the value of each of a set of the company’s financial ratios by coefficients derived from the statistical process that underlies discriminant analysis. The process ensures maximum difference between the scores of the failed and the going concerns. We built a separate model for each of the five years.
The results of the discriminant analysis, appearing in Exhibit II, show a significant improvement over the best-performing operating cash flow ratio, the one incorporating current liabilities. The reason for the improvement is the increase in the percentage of accurately classified healthy companies.
Exhibit II Percentage accuracy using accrual-based measures
The superiority of these statistical models, however, did not preclude the possibility that OCF, CL, or TL possesses marginal value if used together with the six financial ratios. Accordingly, for each year we ran separate discriminant analyses including the six financial ratios and each of the operating cash flow variables. None of the results improved significantly on the percentage accuracy obtained using the combination of financial ratios alone.
Implications for Managers
The results suggest that other factors, such as a company’s debt level, its access to the debt and equity markets, the salability of its capital assets, and its reservoir of liquid assets, may be better indicators of its survival prospects than cash flow data.
While OCF data proved inaccurate in this study, operating cash flow possibly could perform better in other applications. For example, it might be a more reliable predictor of loan default than of bankruptcy, since a decision to default is usually less subject to political and other extramarket forces than is a decision to file a bankruptcy petition.
Business acquisitions, particularly leveraged buyouts, are another area in which operating cash flow data may have predictive value. Since the ability of an acquired company to contribute heavily to service debt is a critical factor in many acquisition decisions, operating cash flow and related measures may be useful in identifying potential targets.
Nevertheless, current speculation on the best uses for operating cash flow data may be missing a bet. Thus far, attention has centered on historical operating cash flows; a potentially more worthwhile kind of data, cash flow forecast, is already available to financial executives. We doubt, however, that companies’ managements would voluntarily accept a refocusing toward prospective operating cash flows. Despite recent encouragement by the Securities and Exchange Commission that businesses issue more detailed forecasts of earnings and financial condition under the protective guidelines of the SEC’s “safe harbor” rules, few companies have done so. It is unlikely that they would view forecasting of cash flows any more favorably.
Our finding that OCF data do not accurately distinguish between healthy companies and dying ones raises a question about the presumed value of cash flow data for analyzing and forecasting a company’s performance. Elevating cash flow, without testing its applicability, as the panacea for the problem of assessing performance is akin to the euphoria in the 1960s surrounding growth in earnings per share as supposedly the best indicator of company value. We hope that unbridled enthusiasm for cash flow data will not produce a repeat of the debacles that resulted from blindly following earnings-per-share growth.
[For summary statistics, see the Appendix.]
Appendix
After calculating OCF, CL, and TL for each company for each of the five years before bankruptcy and for the same period with nonbankrupt companies matched to the bankrupt ones, we applied statistical tests to the individual cash flow variables in order to assess their ability to discriminate between bankrupt and healthy companies. We compared the predictive accuracy from these analyses with that produced by application of a standard statistical tool, multiple discriminant analysis, to a set of six accrual-based financial ratios. To test for their marginal predictive value, we added each of the OCF variables to the discriminant analysis models.
Summary statistics for the operating cash flow variables appear in Table A. In general, the differences between the averages (means) of the two groups were statistically significant; they were not the product of mere chance. Despite the differences between the group means, however, none of the OCF variables could discriminate between the bankrupt and healthy companies with reasonably good accuracy.
Table A Summary statistics for OCF, CL, and TL in $ millions
The poor predictive accuracy was due to the many inaccurate classifications of nonbankrupt companies as failures. Examine Figure 1, which plots the distributions of values for the OCF variables for the bankrupt companies for the last year before failure and similar data for the matched nonbankrupt companies. The distributions overlap considerably, making it difficult to distinguish between the two groups. (The overlaps in charting the CL and TL variables are only slightly less.) Causing the overlap is the large number of nonfailed enterprises whose OCF variables closely resemble those of the bankrupt companies. The graph indicates that although a large number of companies generate little operating cash flow, most of them do not file for bankruptcy.
Figure 1 Distributions of OCFs one year prior to bankruptcy
1. The FEI plans to publish the survey this summer.
2. James A. Largay, III and Clyde P. Stickney, “Cash Flows, Ration Analysis and the W.T. Grant Company Bankruptcy,” July–August 1980, p. 51.
A version of this article appeared in the July 1984 issue of Harvard Business Review.
