A personal finance columnist for the NZ Herald

Inside Money: Flight of the asset classes: see you later allocator

Photo / Thinkstock
Photo / Thinkstock

Any of your bog-standard investment processes today are centred on the idea of asset allocation: that is, your investments should be spread around a number of different assets according to some kind of logical plan.

The dominant idea that emerged after years of academics tinkering with formulae is termed strategic asset allocation (SAA). Under SAA, your investments would be split between asset classes (usually shares, bonds, cash and property) in proportion to the amount of risk you wanted to take where risk was the probability, based on historical records, that the asset class in question over a given time period would suffer a loss.

The SAA model assumed that over the long-term each asset class, whatever short-term performance fluctuations, would revert to historical type. Therefore, all you had to do was to determine your risk level, set up your asset allocation and periodically rebalance the actual investments (which tend to drift over time) so they stuck with the SAA plan.

If you hung around long enough, the SAA gospel went, all assets would give the returns suggested by their historical averages.

In recent years, however, high-level investment thought-leaders have begun to question the assumptions underpinning SAA, with its 'set and forget' model increasingly under threat. Some critics argue that while the SAA mean-reversion rule remains sound, unfortunately its 'long-term' might be a little too long for the average human life-span. Others dismiss the concept altogether, arguing investors are more or less in constant chaos management mode with history an unreliable guide to the future.

In practical terms, the SAA-rethink has led the development of several more active, acronym-expanding asset allocation models, such as dynamic asset allocation (DAA) advocated by global investment consultants Mercer.

A Mercer competitor, Russell Investments, has also argued that SAA is old hat. Graham Harman, Russell Investments head of capital markets research, presented the new-think at the recent Russell NZ conference, in a session titled 'Drop the pilot: do you need a strategic asset allocation?'.

Harman's original 'Drop the pilot' paper, written for an Australian audience, offers an historical overview of the SAA debate. It also embarks on an existential examination of the whole business.

"... the mere act of defining an asset... as an asset class, will give enhanced prominence to its diversification qualities and will lead quantitative asset optimisation models, in particular, to 'overweight' that 'asset class'."

Harman's paper considers the trend to what Russell calls adaptive asset allocation (AAA), where rather than relying on old labels investors can now forensically examine their investments along a broad range of factors.

"There is a hint, too, in this line of reasoning, that SAA models are technologically obsolete," the paper says.

"'Asset classes' are of no value in themselves, but historically have acted as a kind of primitive factor model and to that extent have served their purpose. But now that we actually have factor models, asset classes are going the way of the horse and buggy."

Get the news delivered straight to your inbox

Receive the day’s news, sport and entertainment in our daily email newsletter

A personal finance columnist for the NZ Herald

David is a freelance journalist who has covered the financial services business on both sides of the Tasman for over 15 years. He is the editor of industry website Investment News. David has edited magazines and websites for the financial advice, investment and superannuation industries.

Read more by David Chaplin

Have your say

1200 characters left

By and large our readers' comments are respectful and courteous. We're sure you'll fit in well.
View commenting guidelines.

Sort by
  • Oldest

© Copyright 2017, NZME. Publishing Limited

Assembled by: (static) on production bpcf05 at 29 May 2017 13:02:45 Processing Time: 303ms