Australian equity returns restated down

Ian Rogers
Method wins the equity data race    (Image: RBA Museum)
A new and "hand-collected" dataset on listed companies from 1917 to 1979 by a team of patient researchers at the Reserve Bank of Australia forms the basis for findings of contrarian, conservative measures on returns from the Australian equity market.

Thomas Mathews, from the RBA's Domestic Markets Department, is the lead and sole author named in "A History of Australian Equities", a Research Discussion Paper published yesterday.

Mathews found that:
  • dividends for the early 20th century were lower than previously believed;
  • the realised returns on equities has averaged about 4 percentage points above that on government bonds since 1917, "somewhat lower than previous estimates";
  • price-to-earnings ratios are currently almost exactly at their very long-run average, in contrast with the experience of some other countries.
On average around 65 per cent of listed company earnings were paid back to investors in the form of dividends from 1917 until the present, although this has varied over time and in the case of large banks in particular in recent years, these ratios have been much greater.

The dividend yield on a market capitalisation-weighted portfolio of stocks "is about
200 basis points lower than Lamberton's estimates," a reference to "the most widely-used source of historical Australian stock market data calculated in 1958 at the request of the Sydney Stock Exchange" by economist Donald Lamberton.

Mathews most discordant finding may be that "the lower dividend yield in the RBA's equities dataset compared to other sources implies a lower total return to equity - and therefore a lower equity risk premium".

"If we add the dividend yields onto capital gains calculated in Lamberton's share price series, this results in an average realised return on equity of around 4 per cent in excess of that 10-year government bonds."

This is roughly a percentage point lower than estimates from three Bond University analysts produced in 2008, "which themselves are … on the lower end of Australian ERP estimates".

Mathews concludes "the realised equity return relative to safe assets declined substantially in recent decades (and, relative to long-term bonds, briefly turned negative), as declines in interest rates granted bondholders large capital gains."

And in a nod to the oligopolistic character of Australian capitalism, Mathews highlights that today more than half of the modern stock exchange (by market capitalisation) is comprised of financials, particularly banks, and resources companies, particularly miners.

This turns out to be nothing novel.

"Strikingly, this was also the case 100 years ago, and in fact the relative proportions are very similar.

"Not only is the aggregate composition by sector the same as it was 100 years ago, but in many cases the exact same companies (albeit following some merger and acquisition activity) still dominate the exchange," Mathews said.