In a previous paper,' I demonstrated how the problems faced by the auditor of a firm's published annual financial statements in deciding (1) how much audit evidence to obtain and (2) what set of balance sheet valuation numbers to choose in the light of this audit evidence can be regarded as a problem in Bayesian point estimation. In effect, the auditor was viewed as picking a vector of numbers (the balance sheet) to summarize his posterior probability distribution (i.e., posterior to his audit examination) of those balance sheet parameters for which he is held responsible by the nature of his audit engagement. He does this subject to a loss function whereby he is penalized for discrepancies between the balance sheet numbers so chosen and their subsequent realization. Given the professional nature of the auditor's engagement, it seems reasonable to suggest that this loss function really derives from the various financial statement users. If they rely on the audited statements as inputs into decisions when those statements contain auditor's errors, they will suffer an opportunity loss of expected utility. This raises the question, of course, of who these users are and how financial statement errors operate to cause utility losses. The approach in this paper is to pick a specific, well-defined normative decision problem for which audited financial statements have the potential to be useful and find the loss function that is implicit in it. The problem I have chosen is an individual's consumption-investment
[Models of the behavior of populations of self-reproducible natural resources in an economic framework have rarely anticipated the consequences of different forms of production functions. This paper investigates sufficient conditions for extinction in a very general model as well as a model having a specific production function. In the second section additional considerations relating to extinction are deduced as well as the existence of a watershed level of population. These conclusions are exemplified using data from one particular population of red deer.]
[The model extlesstex-math extgreater$Y_\it\= extbackslashSigma _\k\( extbackslashbeta _\k\+ extbackslashdelta _\ik\+y_ k\)x_\ikt\= extbackslashvarepsilon _\it$ extless/tex-math extgreater with extlesstex-math extgreater$ extbackslashdelta _\ik$ extless/tex-math extgreater and extlesstex-math extgreatery_ k\ extless/tex-math extgreater random is considered as a means of pooling the time series of a cross-section sample. The model is placed in a mixed analysis of variance framework. Relationships between various estimation criteria are derived and their asymptotic properties compared. Some implementation problems are also discussed.]