Where would you expect to see high returns on equity?

Last week I posted some data showing the pre-tax rate of return on equity (ROE) for different sectors of New Zealand business, based on the latest Annual Enterprise Survey (AES) from Stats, and taken back over the past five years. Going by hits on the post, there was a lot of interest - partly, I think, because profitability is an interesting and important concept, and partly because nobody else seemed to be mining the rich seam of data in the AES, or not in public at least, so the results were new to a lot of people.

One of the conclusions I came to was that some sectors seemed to be achieving rates of profitability that looked rather high for the kinds of activity they're in, with wholesaling, retailing and construction, in particular, earning what looked like high rates of return for what looked like relatively workaday industries (although recently high ROEs in housebuilding were more explicable, given the very large post-earthquakes demand for scarce housebuilding resources). It's possible that there are subcurrents in the data that are exaggerating the ROEs being earned: for example, some industries don't need much capital invested in them, so any profits at all get compared with a small investment, giving you a large ROE. Or returns to human capital are being misattributed to physical or financial capital. But overall it still looked to me as if some industries seemed to be earning quite generous profits, given what they do.

That, however, was based on a rather subjective view of the relative riskiness of each sector of business. And it seemed reasonable to do that, at least for some sectors: without doing any sophisticated analysis at all, I'd have rated the more infrastructural activities like electricity, water, gas, transport, and warehousing as relatively low risk, everyday activities that would be consistent with earning modest ROEs, and indeed that's exactly what the AES data show. But for all I know there's more risk in some sectors than amateur navel-gazers might guess from the outside, and higher ROEs might well be appropriate compensation for those real risks.

Which was why I was interested to come across this guest post, 'The Industries Plagued by the Most Uncertainty', on the Harvard Business Review blog site. The three authors came up with one of these 2 x 2 tables, with an index of technological uncertainty along the horizontal axis and an index of demand uncertainty on the vertical axis. Here are the results: they're on American data, but I don't think that makes much difference, though we obviously don't have some of the industries that the States does (such as aircraft manufacture, or big pharma).

This seems to me to provide quite a nice anchor for the ROEs you might expect to see in an industry: it may not cover absolutely everything that an equity investor might expect to be compensated for, but it certainly captures two of the major kinds of risks, In the bottom left, you'd expect lower ROEs, since there isn't a lot of demand or technology risk that investors need to be compensated for, and you'd expect higher ROEs in the top right corner, where both risks are high. And when you look in detail the results, they seem commonsensical. The utilities, for example, feature where you'd expect them (bottom left), as do the high tech sectors (top right).

The bottom line is that I'm still left with some of the same conundrums as before. Why, for example does wholesaling, which on this analysis is one of the least risky business activities (and which you might have guessed was, without ever seeing this analysis), earn an ROE in New Zealand in 20-22% territory? Twice the return that manufacturing earns?

And it's not just industry sectors earning more than you'd think they ought - there are also some strange examples of industries earning less than you'd think they should be. Agriculture on this analysis is reasonably risky - it squeaks into the top right quadrant - but in New Zealand it earns a pitiful, pre-tax, 5% return on equity in recent years.

So there are some real puzzles here. And even for those of us who reckon that markets in general are a pretty good way of getting the most out of our resources and best delivering what people want, you find yourself wondering if something isn't working out the way it should.  On face value, these patterns of profitability don't sit comfortably with the view that competition will deal to excessive profitability, or (consequently or independently) that capital is being allocated to its most productive use.