This is an old paper (1997) but interesting none-the-less.
Alistair Murdoch of the Deparment of Accounting and Finance, University of Manitoba wrote a paper Are Preferred Shares Debt or Equity?: Some Canadian Evidence with the abstract:
I test the appropriateness of new accounting standards that would treat some types of preferred shares as debt rather than equity. I develop a new model to examine whether capital markets view the (systematic) risk of preferred shares to be more like the risk of debt or more like the risk of common equity. The proposed model is compared to a traditional model tested on 1986 to 1994 data of thirty-nine companies that trade on the Toronto Stock Exchange and issued preferred shares during that period. Debt, retractable preferred shares and p lain vanilla preferred shares appear to be substantially less risky than common shares. Other types of preferred shares are more risky than debt; some appear to be more risky than common shares. These results support the view that some types of preferred shares should be classified as liabilities.
He reasons that:
A firm that increases its debt/(common) equity ratio increases the risk of the (common) equity and may increase the risk of the debt. If preferred shares are viewed as debt, then issuing preferred shares should have the same effect as increasing the debt/(common) equity ratio. If they are viewed as equity, then their issue should have the same effect as decreasing the debt/(common) equity ratio. Finally, if they are some intermediate form of financing, they may have no effect on the risk of the (common) equity.
…
… I confirm that debt is less risky than common equity. I find that retractable preferred shares and plain vanilla shares are also less risky than common equity and that in most of my tests their level of risk is not statistically significantly different from that of debt. Other kinds of preferred shares are more risky than debt. Convertible preferred shares are as risky as common shares, while preferred shares that are convertible at market or that are both retractible and convertible appear more risky than common shares
Risk is defined as Beta in the Capital Asset Pricing Model. The “Asset Beta” of each company is decomposed into the equity beta and terms representing the contributions of debt and the various flavours of preferred share, where:
A firm’s asset beta is a measure of the relation between the firm’s return on assets and the market return. Beta is usually estimated by regressing the firm’s return on the market return over some period during which beta is assumed to be constant.
After performing the regressions, the author concludes:
This paper has attempted to determine empirically whether during the past decade the Canadian market has viewed preferred shares as being more like debt or more like equity. It reports that retractable preferred shares have been viewed much like debt which supports the recent change in the accounting classification of these securities
An interesting paper in that it attempts to classify preferred shares according to the effect of their issuance on the Beta of the common, rather than considering the market price of the preferred shares themselves.
[…] PrefBlog Canadian Preferred Shares – Data and Discussion « Are Preferred Shares Debt or Equity? […]
It would be interesting to have those regressions re-run with the experience of late 2008 added into the mix. I suspect that the paper’s conclusions regarding straight preferreds could not be sustained, although plain vanilla debt was hardly treated kindly either.
Such an extension of the time period would be especially interesting in light of the IMF finding that balance sheet leverage (which I believe uses Tangible Common Equity as the divisor) was strongly correlated with market returns on common stock during the Credit Crunch and Treasury’s new-found emphasis on TCE.
The Crunch will be providing fodder for academics for a long time to come!