[ 2]
The accounting profession has struggled for years with the concept of materiality. The audit opinion states that financial statements “fairly represent” the financial health of an organization. The materiality threshold is in engineering jargon an indication of “allowable error in measurement.” Current audit practice relative to materiality has been in place for three decades. It represents a compromise between the cost of audit investigation in manual records and the benefit for stockholders of this investigation. Information technology has dramatically changed allowing for cheaper and more effective controls and investigation, unbalancing this archaic compromise. While the tradeoff between the accuracy of measurement and the cost of assertion continues to be real, the break even point has changed in reality but not in practice.
A likely and desirable change, the leveraging of technological change, is the justification of the audit turning towards the improvement of data quality at the client (Vasarhelyi and Cohen, 2005[2]) and providing a variable level of assertion depending on asserted process. Clients and auditors would agree on the assertion needed on different processes subject to minimal requirements set by statute. Business entities that have real needs of data quality and validation would decide where the optimum tradeoff would be and pay accordingly. This would create a much larger economic threshold for assurance services as companies already pay much attention to data quality. In the future world of a universal data bus and balkanized information being transferred among interoperable Web Services will create even larger concerns for data quality.
While the concept of financial statement audit will continue for a while, a new set of assurance types will emerge where auditors, or other assurors, will place an imprimatur on data at the tag level. This imprimatur can be at the data accuracy level (this data is 98% correct) or at the process level where effective controls that act on the data would be either listed or rated. Obviously these two approaches dovetail and can be used simultaneously. Furthermore, it must not be lost that a wider set of assurance service may emerge with classes such as wider audits, intervening audits, ubiquitous audits, control rating audits, causal audits, etc.
As continuous audit techniques, become more prevalent, the entire economics of auditing and financial report preparation will change. With the cost of automatic procedures becoming negligible in an ERP environment, so will the ability to conduct analytic procedures on a real time basis. The tradeoff between sampling and full population testing will shift akin to the change in the materiality threshold. More generally, the evolution and ubiquity of ERPs will fundamentally lower the costs of compliance and reporting. The basic cost of preparing a report that obeys a particular auditing/accounting standard will become slight as it is prepared by the ERP provider and pulled out as a standard product. Setup costs however, may vary among installations as the basic data for the new requirement may not be available
