Picking Out Actively Managed Funds that Predictably Outperform

There is an oft-repeated rule of thumb that past performance is no guide to future performance, that picking funds with recent above-average results usually leads to disappointment because their future performance tends to under-perform and revert to the mean. It certainly seems to be true in Canada (e.g. Richard Deaves in the book What Kind of Investor Are You? says there is no historical evidence of persistence in either under- or out-performance). Is that situation necessarily inevitable and immutable?

Along comes Super Fund Performance, a report from the Australian superannuation industry, to say "yes, it is possible to identify with considerable confidence which funds will continue to out- or under-perform". The key differentiator: whether the funds are retail for-profit offerings or not-for-profit (sponsored by public sector, corporate or industry). Not-for-profit has been winning, and by a huge amount - an average of 2% per year for the past 15 years!

Below is a fascinating chart from the study. Especially interesting is that the not-for-profits always did better than the retail for-profit funds. It looks like a sure-fire way for Australians to narrow down retirement fund selection. Too bad the not-for-profit option doesn't exist in Canada and it doesn't look like pension reform via the government's PRPP proposal will change that soon.....

Why the out-performance?

• lower fees/costs - the report doesn't state the difference but a quick search in one comparison website, the Canstar Superannuation Star Ratings, shows MER ranges from about 0.4% at the very low end to 2.5% at the extreme high end. The big not-for-profit funds look to be in the 0.6-0.8% Total Expense Ratio range according to another comparison site, Selecting Super.
• economies of scale, but only among the not-for-profit funds; retail funds don't get any more efficient with size, which the report says is due to governance factors - "... either economies of scale are not available to retail funds, or the benefits are not passed on to members". 
• embedded advice - the for-profit riposte is that their members get financial planning advice much more than clients of the not-for-profits, about twice as often according to Canstar. Investor Daily reporter Wouter Klijin says the not-for-profits are increasingly gearing up to provide advice, which he says will increase their costs and possibly/probably lessen their out-performance down the road. A big question is, of course, how useful the advice is, whether it is merely sales or client-interest-first, depending on the source. From which type of outfit would I want "advice"? That's a pretty easy call I'd say. 

Why Does Mutual Fund Performance not Persist? by Wolfgang Bessler, David Blake, Peter Lückoff and Ian Tonks provides another take on persistence using US data -

• Winner funds keep winning if, a) the fund investment manager stays on and b) there is NOT high fund inflow
• Losers start doing a lot better if, a) the investment fund manager is sacked and, b) there is a high fund outflow

The problem for the investor is of course, that what works on average may not happen for the specific fund selected. 

Equity Risk Premium and Likely Future Returns: Elroy Dimson's History Lesson

Highly recommended: Rethinking the Equity Risk Premium, Elroy Dimson's free (registration required) audio and slide presentation that says investors should expect no more than 3-3.5% annual real returns from equities over the riskless rate aka government bonds, which currently is more or less 0!

Noted academic and equity researcher Elroy Dimson (best-known as co-author along with Mike Staunton and Paul Marsh of the yearly investor treat called the Credit Suisse Global Investment Returns Yearbook, 2012 edition here) has been at it again, producing well-explained, substantial commentary on a supremely relevant topic - how much return have investors achieved around the world and what should one expect from equities in future?

Other notable points in the presentation:

• the equity risk premium has not been constant over time, or from country to country, though equities have substantially outperformed bonds over the past 112 years in every single one of the 19 countries for which they have data (slide 19)
• over the very long run (100 years or so) currency shifts between countries have barely mattered to equity returns for an international investor (using the USA investor as the example) ... though it has mattered a lot over short periods (slide 17 c. 21 minutes along)
• most (or all, in some countries) of long-term real equity return comes from reinvested dividends (slides 20 and 222)

There's more fascinating and potentially useful stuff on such topics as - do higher or lower GDP growth countries give better equity returns, do weak or strong currency countries give better equity returns, do high or low dividend markets foretell better equity returns? Here's one of Dimson's slides that reinforces the idea that "valuation matters", at least in the form of dividend yield (as this post on my other blog showed, the current Canadian yield around 3% is around the the historical average):

There's also a background paper with the same title (likely to be heavy reading; though I haven't yet read it, I'd bet it's worthwhile), also free from the CFA Institute.