The litigious nature of American society has led to poor compromises in the disclosure of data. The profession, stung by criticism and litigation, has often decided and set standards for single number disclosure on stochastic assessments. The profession has not issued attestation stating that a particular financial statement is reliable to the 95%, has not allowed for management earnings forecasts to be stated in ranges, and has not stated that most mineral reserves are of a certain value based on the commodity prices in the last 12 months. However it is pretty clear that statements of this type would be preferable for sophisticated users.
While many statistical estimates pervade annual reports (e.g. pensions, bad debt, etc.) these are stated in a deterministic format emphasizing the basic weakness of traditional reporting. When the distance between the report and its underlying stochastic reality gets too big, the credibility of business reporting disappears. If the variances around the values of estimates are very large the credibility of point estimates are very small.
The new business reporting model will have to rely on a wide set of disclosed probabilistic assessments for past results, current actions and future estimates. It is better to be about right than exactly wrong.
proposes a set of probabilistic oriented reports whereby all items in the traditional statements (BS, IS, & FF) are reported as point estimates with a variance measure. For example, the corporate cash level at 12/31/xx was 20m plus or minus 5%, and our best estimate for the value of inventories is 60m plus or minus 15%, and that our current estimate for P,P & E is 71m plus or minus 25%, and that the intangibles in our balance sheet originated by the merger with ABC corporation are 75m plus or minus 100%. Each of these numbers is composed of numbers from each division and each of these numbers has its intrinsic variance. As the public today glazes at the thought of point estimates and variances, a targeted educational effort could help significantly while the investment public can ultimately use the point estimate at a deterministic estimate if so desired.
Extending the reporting range for non-financial variables, key numbers in financial and non financial units would describe non-financial items and hyperlink to the bases of estimates using point of comparison indices developed by the specific industry relevant to the particular line-of-business. Assessments of quality control probability based scorecards would complement this picture. [m6]
Figure 2: Probabilistic reporting