What about 15 years of past returns? That's more like it. Not fool proof, but it is better.

20 years, even better.

I can tell from the anguished faces that investors can't wait that long to tell whether they have a dud adviser, or not. By the end of 20 years it is too late! The horse has bolted.

What can we use to judge performance?

The past 15 to 20 years returns is one place to go, but even then, there can be misleading signals in those years.

There was a time recently when the best performing share markets in the world over 20 years were South Africa and Australia. But, things change, cycles come and go, and lo and behold, for the next ten years or so they weren't. Damn.

I have always thought that the best way to pick an adviser, a fund or any investment for that matter is by having an investment philosophy that makes basic sense.

Some proof that the philosophy works in practice is also important. But don't expect it to be top of the pops every year. A successful approach may only be above average half the time over the next 20 years.

Mercer have just released some great research. They looked at their top eight international fund managers over the past 15 years. They had, on average, added 2.7% per annum above the middle manager. A good result.

However, strikingly, despite their stellar long-term track records, these funds, on average, underperformed in nearly six of the 15 calendar years. An investor in these funds would have been experiencing below-average returns 40% of the time, and those influenced by one-, three- or even five-year performances may have been inclined to give these funds a wide berth at numerous points over those 15 years.

There are, unfortunately, a lot of clever people who just can't understand these 'home truths' about investing. They think that we are just being parochial with a vested interest in holding onto our clients when we defend under-performance. They may not have experienced their share of over-performance yet if they are new.

We do have a vested interest in keeping our clients. Therein lies the problem. We have a vested interest and we are right in saying that past returns are not a good guide to future returns. Both when they are great, and when they are not-so-great.

Investors who chase recent past returns run the risk of getting it doubly wrong, jumping ship at exactly the wrong time. See how it works below -

'Hopeful' starts with adviser A and jumps ship after five years to adviser B because B's returns are better:

• Adviser A averages returns of 6% per annum for five years
• Adviser B averages returns of 8% per annum for five years

Only to find the table turns for the next five years:

• Adviser A averages returns of 8% per annum for the next five years
• Adviser B averages returns of 6% per annum for the next five years

If 'Hopeful' had stayed with adviser A or adviser B for the whole ten years, they would have returned a total of 97% or 7% per annum. Because 'Hopeful' jumped ship they only received 79% or 6% per annum.

If they had, in fact, a million-dollar portfolio it cost 'Hopeful' $175,000 by jumping ship half way through the ten year period.

What if adviser B had continued their run of good luck? It could have worked for 'Hopeful'. Sure. It was a gamble. But the Mercer study showed that there was no predictive power in any five-year return for their sample and this has been confirmed in every bit of independent research we have ever seen on the matter.

Returns from one five-year period do not correlate with the returns of the next five years. The connection between those two lots of five year periods is random.

Why do fund managers or advisers fail to replicate their past returns into the future? The answer: investment strategies don't work over every time period. They only work on average over longer periods of time. Investment cycles come and go. An active manager might try and pick them ahead of time but that is not something you can rely on.

Sad, I know.

Choose a strategy very carefully and stick with it, through thick and thin.

Keep asking great questions ...

P.S. Disclosure – our investment strategy of picking small over large companies and value over growth companies, including emerging markets and not over-weighting New Zealand shares is currently out of favour. Large growth companies and New Zealand shares are flavour of the month. Emerging market shares, not so.

Next fortnight catch the historic rise of value investor Julian Robinson who thought that “value investing was dead” despite being a deep value investor himself. He pulled out at just the wrong time. A great lesson for staying the course.