Brent Sheather: Pay your money and take your choice

Andrea and Alan need some financial advice: they have a freehold home in Remuera, a section worth $350,000 with a $200,000 mortgage and $150,000 in the bank.

They wonder if they should:

1. Pay off the loan.

2. Borrow to put a house on the section and then rent it out.

3. Sell the section and invest the net proceeds and the cash in a balanced portfolio.

There is, of course, no right or wrong solution in this situation as the answer can depend on intangible factors. But if you are interested in reaching a logical conclusion, this involves estimating the future returns of each alternative.

Getting an idea of what an investment will return is fundamental to making a good decision, but many people have different ideas as to just how this should be done.

A favourite method is to look at past performance and assume that the trend will continue. For example, three or so years ago when property prices were buoyant Andrea and Alan may have opted for the borrow, build and rent business model because everything they read at that time had positive things to say about historic returns in the property market.

Focusing on short-to-medium-term trends is sometimes known as momentum investing and, while it is easier than other forms of analysis, there is no logical reason trends should continue.

Indeed, regulatory authorities warn investors not to pay attention to historic returns as they are at best irrelevant and at worst signal exactly the wrong course of action.

More exotic ways of estimating returns include estimating the worth of a stock, bond or property investment using a computer model and forecasting returns, costs and risk. I used to do this at Tasman Pulp & Paper in the early 1980s and, according to our forecasts at that time, the NZ dollar should now be worth about US6c.

One thing is certain: the high incidence of bad investments suggests forecasting returns is as much an art as a science. Even the experts can get it woefully wrong.

Economics and commonsense say that the return on an asset is equal to the income you receive on it plus the movement in its price over the period. For a bank deposit, the calculation is pretty straightforward - you are quoted an interest rate and you get your money back so the return is equal to the interest rate.

For assets whose prices move around a bit, such as property or shares and longer term bonds, you have to estimate what the income will be and what you can sell the asset for when you want to exit. But that is not the full story. Thirty-year Government bonds in the US in January 2008 had an initial yield of about 4 per cent yet produced a total return of more than 40 per cent in the subsequent 12 months.

While estimating the income is reasonably easy, guessing where the price will be in 10 years is not without risk. In fact, as US Defence Secretary Donald Rumsfeld was quoted as saying, "there are a number of unknown unknowns".

Economics can help us out, though, as it gives us a theoretical model to follow. The theory says that return (r) = dividend (d) + growth in profits (g).

Therefore, if you own a house in Remuera and expect the rent over one year will be 4 per cent of its value and the rent will increase at the rate of inflation (3 per cent), the numbers for d + g = r are: 4 + 3 = 7 per cent.

The equation r = d + g is pretty simple to use (or I wouldn't be using it), but its relevance takes a bit of understanding.

What underlies the assumption behind the relevance of (g) is that the value of the asset - a house in this case - will increase or decrease as the rent rises or falls. This follows from an economic theory which simply says something is worth the cash profits it will produce over its life.

So if we know that historically rents in Auckland or even nationally have risen at about the rate of inflation, and we know that most estimates of inflation are 2 to 3 per cent a year, then we can estimate that the return from that asset in the long run will be 4 + 3 = 7 per cent.

Of course, prices can go up or down randomly, tenants can default, floods can happen and so on. But at least the figure of 7 per cent gives a basis to compare a house as an investment with alternatives like bank deposits - ignoring risk.

Estimating (g) is generally where all the big problems of forecasting are concentrated. For example, in the year ended March 2000, the technology sector of the US stock market returned 48 per cent without the benefit of any dividends, largely because people thought that the g component of r = d + g was going to grow at a fantastic rate for a good many years. When analysts started winding down their estimates of g, prices duly collapsed.

The technology crash led many people to say the stock market was irrational, which may or may not have been right at the time.

However, it also points to the fact that (g)rowth is difficult to forecast and prone to exaggeration and that (d)ividends are a much more reliable component of (r)eturn.

Today, the dividend of the US stock market overall is about twice as high as it was in 2000 but still only about 2 per cent - so investors must be expecting a good deal of dividend growth if they are to get to a 10 per cent return.

In practice, however, g usually doesn't even rise as fast in the long run as economic growth. Robert Arnott, writing in the Journal of Portfolio Management, says earnings growth a share for US stocks in the 20th century averaged 4 per cent a year, of which 3 per cent was inflation.

So, realistically, with dividends only at 3 per cent, total returns from US shares may struggle to get past 6 to 7 per cent. However, this still looks pretty flash compared to 1 per cent in short-term US Government bonds.

A cynical New Zealand investor, with the benefit of hindsight, might wonder how Americans could be so silly as to price an asset (technology stocks) to yield virtually zero.

Of course, you had to be there at the time, but one major difficulty with taking an alternative view in such a case is that maybe the zero yield on the Nasdaq correctly reflected the likelihood of strong profit growth in the future. After all, the efficient market hypothesis (EMH) argues that stock markets are correct most of the time.

So if you couldn't justify technology prices, was it because the market was wrong or the fact that "you didn't get it?" After all the bubbles and crashes of recent years, maybe the EMH should be modified to say that markets are efficient except when they are not.

Taking a position away from the crowd is risky - some fund managers who couldn't see value in the Nasdaq lost their jobs when their performance didn't keep up with the indices. Of course, they were proved right eventually, but I doubt that they got their jobs back.

Certainly New Zealand stock market valuations look much more sensible. Dividends are about 5 per cent so that even if profits just grow at the rate of inflation (2 to 3 per cent) an 8-9 per cent return looks possible. R = d + g is, like all economic models, an abstraction of reality, but at least it gives us something to work with and it forces us to set out our assumptions as regards dividends and growth.

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