Interesting paper from Paul Povel, Rajdeep Singh and Andrew Winton from the University of minnesota on the cyclical nature of fraud and the implications for public policy and legislation.

The paper, Booms, busts and frauds, reveals that fraud is most likely to occur in relatively good times, that the link between fraud and good times gets stronger as monitoring costs decrease, and that fraud peaks towards the end of a boom and is then revealed in the bust that follows.

The interesting claim is that fraud can increase if firms make more information available to the public:

"We have already seen that lower monitoring costs sometimes lead to more fraud. Increased disclosure may have the same effect. Suppose that improved disclosure makes investors trust public information more, so that they are more likely to fund firms with positive information and deny funding to firms with negative information; then bad firms are more likely to resort to fraud so that they can produce such positive public information. To be effective against fraud, disclosure standards must directly make fraud more difficult."

The authors say the level at which investors can be expected to scrutinize the firm in which they may invest depends on their expectations about the state of the economy, and their expectations about firm's fraud decisions. And that, they say, has enormous implications for Sarbanes-Oxley.

"The Sarbanes-Oxley Act was introduced to improve the public disclosure of financial information; the consequence may well be that investors monitor less (instead relying more on the publicly available information), and more firms commit fraud. In other words, the effectiveness of the new constraints that were imposed on businesses to prevent fraud may be more limited than expected, and they may even have unintended consequences."