Marathon run to sharemarket finishing line

By BRENT SHEATHER

Whenever clients express doubts about the precipitous decline in the sharemarket, financial planners and fund managers invariably begin chanting the same mantra: "Long term, long term, buy and hold."

That's probably fair enough if you invested in the British or United States sharemarkets. But as we know, in the long run we are all dead, and new research from three professors at the London Business School shows that, even long term, with some sharemarkets you can still lose everything.

Furthermore, "long run" can turn out to be much longer than many investors are prepared to wait.

Much of the perception that shares were a sure bet long term was based on the pioneering work of Ibbotson Associates, who produced a definitive set of data measuring the returns on US shares, bonds and bills going all the way back to January 1926.

But, though many Americans do not realise it, not everyone lives in the US. So Elroy Dimson, Paul Marsh and Mike Staunton, of the London Business School, have done for the rest of the world what Ibbotson did for the US.

Their analysis highlights the fact that, compared with most countries, the long-term performance of the US sharemarket has been exemplary. But if you lived in St Petersburg in 1900, put all your savings into a Russian index fund and crossed your fingers, you would have lost the lot.

Dimson, Marsh and Staunton show that over the 103 years of data they collected, US shares have returned a positive real (after inflation) return over every 20-year period - 1900 to 1920, 1901 to 1921 and so on.

But in other markets - Japanese shares, for example - almost one-quarter of all the 20-year periods have been negative.

One needs to expand one's horizon to 50 years for returns from Japanese shares to have been consistently positive. Not many recently retired investors have 50 years to wait.

The three professors' analysis shows that in most countries investors have to give the market much longer than in the US before they can be guaranteed a positive return.

Of all the countries surveyed, Japan's experience is close to the median. If the non-surviving markets from 1900 (such as Russia and China) are included in the comparison, then Japan has done better than most.

"Over the last century, taken as a whole, common stocks have indeed been attractive investments," says the study.

"But they have not consistently been attractive over a horizon of just 20 years. In many countries, stocks have done well only over the exceptionally long run.

"Moreover, periods of underperformance can be more severe than has been experienced within the United States, and the duration of underperformance may, on past evidence, persist for several decades."

If investors have been assuming the US model of exemplary long term equity performance and lower volatility, but investing internationally, the net result may be that too large a proportion of their portfolios is invested in shares.

Anecdotal evidence suggests that many local Mum and Dad investors and their advisers are only now realising how long term their commitment to international shares is going to have to be to see a positive result.

The three professors have a cheerful word to say here, too - they calculate that, assuming an average return of 8 per cent a year, there is a 50 per cent chance that the British sharemarket will get back to its December 1999 peak by 2018.

Then again, there is a 50 per cent chance that it will take longer. Oh dear.

The trio also highlight the importance of dividends in total return calculations - a lesson here perhaps for investors in the various capital-guaranteed international share funds that are so popular at present, and which typically see all the dividends going to pay the fees.

Financial advisers and sharebrokers, perennially optimistic, often advocate financial plans stuffed full of shares by pointing to the apparent statistical improbability of a fourth straight down year in world stock markets.

Dimson, Marsh and Staunton will have none of this nonsense. Just because we have had three bad years in a row doesn't change the statistical probability of another occurring - about a 38 per cent chance, according to the good professors.

While four down years in a row is unlikely, the chance is, unfortunately, a fresh sample from the array of possible outcomes.

The professors provide an example to help financial advisers and fund managers understand this complicated scenario:

"An analogy would be the performance of a roulette wheel. Suppose you enter the casino at Monte Carlo, where one roulette outcome is as likely as another.

"Imagine that, as you join the table, the croupier whispers to you, 'Shhh. Let me help you: the last three balls landed on red'.

"Your odds of winning will not be improved. The roulette wheel has no memory: it cannot recall that it is now the turn of the ball to land on black. In the same way, markets flip between favourable and unfavourable returns."

A year ago, many advisers and market commentators confidently predicted a rebound because three successive down years have been rare, and that prediction turned out to be dead wrong. The 2000s are developing something of a reputation for remarkable events, terrorism and the looming global crisis in the defined benefit superannuation industry just two examples that spring to mind.

The study makes the point - fund managers please take note - that to properly understand the risks and returns of capital markets we need to look back a long way before we can make inferences.

Putting it bluntly, the periods need to be long enough to incorporate the good times and the bad. The last five or 10 or even 20 years simply won't do. So put away those charts unless you have a seriously long term time series.

The 103-year analysis helps to put the current bear market into perspective. In the three years from January 2000 to last December 31, world sharemarkets were down about 42 per cent. But in the 1929-32 Wall St crash, US investors lost 80 per cent of their shares' value, and in the 1973-74 bear market, British shares fell 71 per cent.

Relative to these two monsters, the current problems aren't the end of the world.

While it is comforting to know that things have been a lot worse than this, it still feels pretty bad and no wonder - the professors tell us that from 2000 to last year, worldwide sharemarket losses were about US$13 trillion ($23.4 trillion). That's 13, followed by 12 zeros.

So how do local shares compare with the US over the long term? The study doesn't include New Zealand in its analysis as the professors have had difficulty obtaining data from 1900-1925.

But a former Auckland University finance professor, Dr J. B. Chay, has constructed total return indexes of New Zealand shares, bond and bank bill prices from 1928 to last year.

Measured in New Zealand currency, local shares have returned an average 9.8 per cent a year, before inflation. This compares with 10.1 per cent a year for US shares from 1926 to last year.

United States share returns are further improved by exchange rate movements. Since 1952, which is as far back as I have exchange rate data, the New Zealand currency has depreciated against the US dollar an average of 2 per cent a year.

Using Dr Chay's data and the consumers price index, I calculated an index of real (after inflation) sharemarket returns over the period, following the professors' approach, to see what sort of time has been necessary to guarantee a positive outcome in real terms.

The good news is that, based on history, investing in New Zealand shares for a 20-year period will be enough to guarantee a positive outcome.

Even investing for a decade will usually produce positive results. Within the 64 10-year periods in our sample, shares produced a positive return 86 per cent of the time.

Keep the faith.

* Brent Sheather is a Whakatane investment adviser.