SAS No. 59, The Auditor's Consideration of an Entity's Ability to Continue as a Going Concern, requires the auditor to evaluate whether there is a going concern problem. This article describes a relatively easy wasy to calculate a measure of the likelihood of bankruptcy that would be easy to incorporate as an analytical procedure. It could provide the auditor with another method of review to avoid oversights or check out hunches. In this model, which we call the Zavgren model, a ratio from each of seven important functional dimensions is used. These ratios, their interpretation, and their method of calculation are provided in Table 1. Note that all of these ratios are readily computed from trial balance data and are easy to interpret.

For evaluation of the financial condition of a firm, trial balance data are required to compute the ratios, and the model coefficients are needed so that the firm's probability of failure may be calculated. The Zavgren model coefficients are presented in Table 2. Multiplication of each ratio by 100 and by its respective coefficient (from Table 2) and adding these products to -.23883 results in a number of that can be converted to a probability of failure for that firm. Specifically, if the resulting number is y, the probability of failure equals 1 / 1+e.sup.-y., where "e" is the base of natural logarithms. Many calculators have the "e" function and, accordingly, could be used to compute the probability of failure. In addition, a template could be developed using available spreadsheet software to generate those probability numbers. Thus, a firm's probability of failure, as calculated from the model, can be interpreted by the auditor using his/her knowledge about the firm and the context in which it operates.

Another interesting aspect of the Zavgren model is the isolation of "trouble" ratios that play the biggest role in explaining a firm's probability of failure. Since managerial and environmental problems often are reflected in the account balances, the ability of the model to highlight ratios needing attention is a valuable feature for the auditor. The product of each ratio and its coefficient indicates the extent of that ratio's impact on the probability of failure, with increasingly larger positive products resulting in a greater probability number. The heavy use of debt, for example, is a frequent culprit, and such a trend would be readily observable from the calculations.

An example is now provided in which an actual firm in the motor vehicle equipment industry, referred to as Company A, that had known problems and has since successfully reorganized, is analyzed. In analyzing this company over a five-year period by using the Zavgren model, several disturbing trends become evident.

Company A was one of the giants in the heavy-duty trucks and agricultural equipment industry. The company's core businesses included medium- and heavy-duty trucks and agricultural equipment produced in the U.S. and Canada. It maintained a worldwide marketing presence through approximately 3,910 independent dealers and distributor outlets. Although it seemed to be a healthy company in 1978, its situation quickly deteriorated, and it was in grave difficulty five years later in early 1982.

In the case of Company A, its bankruptcy was predictable as far back as 1979, and the Zavgren model clearly highlights steadily increasing deterioration of the company's situation: in 1978 the probability of failure is 29%; in 1979, 56.6%; in 1980, 78.7%; and in 1981, 99.9%.

In 1978, the probability of failure is 29%. This ambiguous figure does not indicate whether a firm is healthy or failing. However, when individual ratios are considered, inventory turnover and capital turnover significantly affect the probability. Even though net income at the end of the fiscal year was $186.7 million (restated upward to reflect a change in 1979 from the FIFO inventory flow assumption to LIFO), Company A's president laid off 3,000 employees and offered early retirement benefits. This move was probably a consequence of the 2.8% profit margin: it would help improve manufacturing efficiency (increase capital turnover) and keep better control on quality (improve inventory turnover by greater sales).

The probability of failure suddenly increases to 56.5% in 1979. Inventory turnover problems increase (the ratio of .37 may be understated because of the change from FIFO to LIFO). The company's cash position factor doubled, reflecting a relatively low cash balance of $17.2 million as of April 30. The situation may still appear healthy to outsiders (net income is up 42% to $95.3 million in the second quarter), but the company's balance sheet does not appear to be as strong. The numbers show a high debt-to-equity ratio of 72% (the industry leader is at 54%).

In 1980, the fiscal year in which the company faced a strike, the probability of failure jumps to 78.7%. In November of 1979, 35,000 unionized employees (representing 36% of the work-force) went on a six- month strike, which closed 17 of a total of 21 company plants.

The impact of the strike is reflected in the company's financial ratios, not so much by their absolute values, but rather by the yearly change from 1979: its return on investment, which was already negative in 1979, dropped to a dramatic -15.6%. This reflects two important figures: a loss of $397.3 million (down from a $369 million net income in 1979), and a 100% increase in short-term debt to $808.9 million. This increase is also reflected in the financial leverage ratio, even though long-term debt increased to $1.33 billion from $948.2 million in 1979.

The probability of bankruptcy increases in 1981 to 99.9%. This fiscal year showed record sales, indicated by a 52% decrease in the model's inventory turnover factor, as well as a 49% decrease in receivables turnover factor. Yet, in the first quarter, the company reported a $96.4 million loss (yearly ROI drops to -127%, from -16%), increasing the ROI factor dramatically. The interest payments on its debt hurt profits badly and forced the company to negotiate a financial restructuring plan to keep it afloat an additional year. This is reflected in the financial leverage factor (9.08, up from .88 in 1980) which shows the change from short-term debt to less expensive long-term debt.

Thus, the Zavgren bankruptcy prediction model highlights Company A's problems. Given this analysis, the bankruptcy model would have clearly indicated to he auditor that the company was in trouble in 1980. It also would have indicated that the poor financial situation of the company worsened considerably in 1981, such that the 99.9% probability of failure provided by the bankruptcy model for that year accurately coincides with the near bankruptcy of Company A in May of 1982.

Though a bankrupcty prediction model can provide the auditor with valuable insight, it obviously should not be used exclusive of the other evidence-gathering procedures undertaken by the auditor to form the basis for the audit opinion. If the model indicates that a client firm's potential for failure is high, the auditor needs to weigh the risk of losing the client if a going concern explanatory paragraph were included in the audit report against the risk of litigation if the company failed within a few months after the issuance of an unqualified opinion. The evidence obtained from client representations, minutes of directors' meetings, discussions with creditors, and other such sources would need to play a major role in the assessment of tradeoffs associated with such risks. On the other hand, as was illustrated in the example discussed earlier, the use of a bankruptcy prediction model as an analytical procedure also could assist the auditor in his/her role as the clienths business advisor by isolating problem areas needing management attention even before going concern issues loom large on the horizon.

Conclusion

We believe that the use of a model such as the Zavgren model would serve to reduce the "expectation gap" problem currently facing the profession by providing the auditor with a more efficient and effective method of review. Specifically, such objective data would provide the auditor with additional insights about the client company's financial condition that a more traditional analysis of trial balance data might not reveal. Finally, those insights would assist the auditor in assessing at what point he/she believes substantial doubt exists about the client's ability to continue as a going concern and whether to include an explanatory paragraph in the audit report.