I recently highlighted a paper by Alistair Murdoch of the University of Manitoba in which he showed that, in terms of the effect on the CAPM Beta of the common stock, preferred shares that were not convertible at a fixed rate into common were perceived by the market as being debt-like.
This paper was used as evidence to support the accounting reclassification of retractible preferred shares as debt for audited balance sheet purposes.
He then wrote a follow-up paper examining the effects of this change, titled Management Reaction to Mandatory Accounting Changes: The Canadian Preferred Shares Case:
The new Canadian accounting standards for financial instruments require that retractable preferred shares be classified as debt, thus negatively affecting the debt/equity ratio. Previous research, most of which has examined the impact of a change in American accounting standards affecting the determination of earnings, indicates that firms with such shares will act to mitigate the negative impact of the accounting change on their financial statements. Specifically, firms are likely to: a) reduce the amount of retractable preferred shares outstanding, and/or b) reduce the amount of other liabilities, and/or c) increase the amount of equity outstanding.
I test these predictions using data on firms required to file information on their preferred shares with Canadian securities commissions. Evidence based on a sample of 34 such firms indicates that they did indeed reduce the amounts of both retractable preferred shares and the amounts of other liabilities and issued additional common shares. Surprisingly, smaller firms did not make greater reductions (as a proportion of total assets) than larger firms.
He makes one statement about market efficiency that I consider worthy of comment:
The economic consequences of the replacement of retractable preferred shares by other sources of financing, primarily common shares, is not clear. If retractable preferred shares became a popular financial instrument during the 1970s and 1980s because they allowed firms to issue debt in the guise of equity, then their disappearance due to the standard change having stripped this disguise from them is desirable because it promotes informational efficiency.
This is of interest because it may be flipped upside down and used as an attack on the Efficient Market Hypothesis. The mandated change in classification is cosmetic only; no information is supplied to the marketplace by an auditor’s opinion as to whether a particular instrument is best placed in the shareholders’ equity section of the balance sheet or not. The prospectus contains all the necessary information for investors to judge the nature of the instrument for themselves and is not changed by this opinion.
Thus, any change in management or market behaviour due to such a change is evidence that the market’s efficiency has changed; and if the market’s efficiency can change, then the EMH does not hold.
In this particular case, I take the view that the change increased the efficiency of the market, by allowing bozos (who equate the term “research” with “looking stuff up in Compustat”) to more closely match the judgement of those who have actually examined the data and thus reduces the rewards for market professionals to think about what they are doing.
This supports the hypothesis that informtion has a value – it ain’t free!
Of interest in the paper was Table 1 “Number of publicly listed Canadian companies with outstanding retractable preferred shares”, which documents a nearly monotonic decline from 1988 (114 companies) to 1997 (32 companies). With the benefit of over a decade’s worth of additional data, we can now see that retractibles from Operating Companies (the HIMIPref™ subindex “OpRet”) has dwindled to virtual insignificance.