Brent Sheather: Leave market forecasts to the badgers

It's difficult to differentiate between luck and skill when picking market movements. Photo / Thinkstock

At this time of year it is customary for many stock market strategists to make forecasts as to where they think the market will go in the next 12 months.

Many of the participants know this activity is a bit of a waste of time but, nevertheless, have a go in the hope that one year they will get it right and be accorded guru status.

The "science" of forecasting stock market returns is usually based on a projection of profit growth for the year and then an estimation of how the stock market will value these earnings.

Both numbers are problematic even without accounting for the tendency of the finance sector to regard all news as good news - the Financial Times recently quoted a report by Goldman Sachs that showed that despite company profits falling one third of the time, not once have US or European analysts as a group predicted falling profits.

The Goldman Sachs report showed that the consensus earnings forecast was out by more than 20 percentage points one third of the time and out by more than 10 percentage points half of the time.

Even predicting the direction of profit growth is difficult. At the start of 2008 the consensus was that US corporate profits would rise 16 per cent when, in fact, they fell by 18 per cent. If this isn't bad enough remember that it is easier to predict profits than it is to predict what valuation multiple the stock market will apply to them.

As always it's difficult to differentiate between luck and skill. The Financial Times polled a number of experts early in 2011 as to how the UK stock market would perform in the year and also engaged the forecasting services of a badger living in one of the FT journalists' gardens. No surprises that the badger won.

So, faced with the fact that short-term stock market predictions are a waste of time, what is the probability of getting long-term projections correct? These numbers are not just of academic interest - retired investors or investors in a KiwiSaver fund need to calculate what level of savings is necessary to achieve a specific investment objective and these numbers obviously influence their asset allocation decision.

Where does one go for sensible, unbiased advice? Mr Badger is unavailable as he now works for a San Francisco-based hedge fund and is no longer talking to the press, but there are more conventional options.

Just recently, a group of luminaries from the US Chartered Financial Analysts Institute Research Foundation met to discuss future long-term US stock market returns. The following summarises some of the main points made in that meeting by two experts, Clifford Asness and Robert Arnott. Both of these men are successful hedge fund managers and academics. Their findings may be of help to longer-term investors.

First up was Clifford Asness, whom we last published in this column back in 2000 when he presciently warned Herald readers that the likely return from the US stock market in the next 10 years would be negative.

Asness ventured an opinion as to what long-term return from the US stock market was probable for someone investing today. He reckons that the US stock market will return 3.8 per cent a year after inflation.

If we add back 2.5 per cent for inflation, that is a 6.3 per cent return, but Asness adds a caveat: he notes that the stock market is expensive on the basis of long-term valuations and there could be some regression back to the average valuation. He does, however, think that stock market valuations in the past are too low compared to today because the market return is more attainable due to the advent of index funds.

In other words, investors today should put a higher value on the stock market because they can invest at lower risk. For local retail investors Asness' forecast 6.3 per cent a year return needs to be compared with the 5.0 per cent to 5.5 per cent available from long-term lower risk bonds locally, the higher fees frequently associated with investing in the world stock market via a financial adviser and Dr Cullen's ridiculous Fair Dividend Rate of Tax (FDR).

The impact of the latter two variables could reduce returns by 2 per cent to 3 per cent a year and 1.5 per cent a year respectively.

Next was a paper by Robert Arnott, former editor of the Investment Analysts Journal, who looked at various myths associated with forecasting long-term returns. Firstly he observed that all of the myths associated with forecasting long-term stock market returns rationalise higher returns.

"No one seems to construct a myth to explain why we should expect low returns."

We can explain this behaviour by the fact that the fee structures of most participants in the finance sector look much too high when compared with reasonable forecasts. If the finance industry is to get away with charging 1.5 per cent a year to manage a share portfolio then 1 per cent to monitor it, it needs to promote at least 8 per cent a year return expectations and preferably higher.

The first myth that Arnott takes an axe to is the fable that stocks will beat bonds for anyone willing to think long-term. Arnott says this myth lingers in spite of a 41-year interval in which the returns of long-dated US Government Bonds beat the US stock market. He notes that outside the US stocks are even more unreliable performers relative to bonds.

Arnott does say, however, that he expects shares to outperform bonds in the next 40 years, but he is even more pessimistic than Asness in that he expects real returns of just 2.5 per cent to 3 per cent a year, which means returns including inflation of 5 per cent to 5.5 per cent a year.

The big points for retail investors are that stocks are very volatile.They can and do go down for extended periods. They therefore cannot be relied upon to beat inflation over a 10- or even 20-year period. As Tim Bond from Barclays pointed out to Herald readers a few years back, the only reliable way of inflation-proofing one's assets is inflation index bonds.

The other implication of Arnott's research is that capital growth from stocks can be here today and gone tomorrow, despite many financial advisers' presentations to prospective investors including spreadsheets that assume regular doses of capital growth, which can be used to fund retirement expenditure. Nothing upsets retired investors like finding out their monthly drawings are no longer covered by returns.

The next myth that Arnott addresses is the one that dividends do not really matter. Locally, it's quite popular among financial advisers to stress dividends when picking stocks. However, the same advisers frequently also recommend international stocks that have dividend yields of around 2.5 per cent. The finance industry argues lower dividends indicate capital growth will be higher in the future.

Arnott says, in fact, there is a strong positive link between the size of the dividend yield and future total returns. In other words, lower dividend yields imply lower future returns. The other important part of this myth is that when companies retain more of their profits and pay out less in dividends, this leads to faster profit growth in the future.

Groundbreaking 2003 research by Arnott and Asness showed that, in fact, higher dividends are associated with higher profit growth in the future. The lesson here for Mum and Dad is that the New Zealand and Australian stock markets have much higher dividends than the world stock market. No surprise, then, that the Global Investment Returns Yearbook shows the Australian stock market was the best performing in their sample from 1900 to 2011.

As to which market will outperform in 2012, that is anybody's guess. Whatever will be, will be.

Brent Sheather is an Auckland-based authorised financial adviser and his adviser/disclosure statement is available on request and free of charge.