Brent Sheather: Back to the future returns

During the past four weeks, we have looked at the tendency of investors to assume past rates of growth will continue. Under this rule of thumb, all you need to do is buy last year's hot market. However, thanks to the good work of two London-based Deutsche Bank analysts, we have seen that that is often a trap - high historic returns from international bonds and international shares very likely imply lower returns in the future.

The main reasons are:

* First, that the dividend yield, a big component of total returns, declines as prices rise.

* Second, valuations tend to return to their average level. Periods of high growth are typically followed by subdued returns. Whereas the historic return from United States shares since 1900 is about 9.8 per cent a year, the future is not nearly so bright - even 6 per cent a year looks optimistic.

So much for international markets - what about local bonds and shares? Given present prices and historic averages what return can we expect in the long run from New Zealand assets and how does this compare with the US data?

One of the biggest problems of trying to do this exercise for NZ bonds and shares is the difficulty of getting reliable long-term information. However, a few years back, several Auckland University professors put together a database of local inflation, cash, bond and share returns from 1928. This was updated recently by a local fund manager, Peter Urbani.

While we have historic performance and dividend figures back to 1928, other historic valuation data like average price-earnings multiples only goes back to 1988. But it is better than no historic perspective at all. The performance data is presented in the table.

First up, we look at the local sharemarket. Back in December 1929, the dividend yield or income of the NZ stockmarket was 5 per cent and, since then, total returns have averaged 9.6 per cent a year. This is well above inflation, which has averaged 5.1 per cent, but only 2 per cent a year better than short-term deposits.

In the US, shares at 11 per cent pa, have returned more than twice the 4.7 per cent pa from bank deposits. But is this outperformance by US equities a reason for bullishness or caution? We know that, in the long run, returns from shares (r) is equal to the dividend yield (d) plus the rate of dividend growth (g), plus or minus any change in the overall valuation* of the market.

Local shares have underperformed US shares in the period but a big part of that outperformance is actually due to US shares becoming more expensive rather than any economic miracle or new paradigm. Back in 1928, US shares sold at about 13x profits, today that ratio is above 20x. It's a similar story with dividends: Back in 1928 US shares delivered an income of about 5 per cent; today the dividend yield is less than 2 per cent.

New Zealand, however, appears to be different - dividends in 1929 were 5 per cent and, today, they are even higher at around 6 per cent to 7 per cent. Add in long-term earnings per share growth of inflation plus 1 per cent and a continuation of 10 per cent pa returns seems achievable.

Unlike the US market, long-term price/earnings data isn't available for NZ shares. However, data is available from 1988 and, for what it is worth, the PE multiple over this period has averaged 15.2x. The local market is now selling at around 14.2x.

So on both these bases, NZ shares look a better bet whereas the US sharemarket looks overvalued from an historic perspective, implying downside if valuations eventually regress back down to the average.

In the bond market, too, history suggests New Zealand doesn't suffer from the same degree of overvaluation common in the US. Although NZ bonds have beaten US bonds since 1928, this return has been largely due to higher inflation rates locally rather than a capital gain as a result of lower bond yields. (This higher inflation has caused the kiwi dollar to weaken against the US dollar since 1972).

In the past 25 years, however, US bonds have roared ahead as their interest rates have fallen further in percentage terms. For most of those 25 years, US bonds have been in a roaring bull market returning 10.4 per cent pa. You could get an income of 12.4 per cent on 30-year US government bonds in 1980. Today, those yields are down to 4.55 per cent and they have been as low as 4.16 per cent.

So someone buying 30-year treasuries in 1980 received, and indeed is still receiving, 12.4 per cent pa on their capital despite the yield today being only 4.55 per cent. Those lucky investors have benefited from the high yield and the fact that, because interest rates have fallen, their bond is worth more if they sold it. The present low yields on US bonds are the reason the asset class has done so well in the past but, at the same time, they constrain performance in the future.

In January 1980, 10-year NZ bonds were yielding 13.4 per cent and it took more than 10 years for yields to fall back into single figures. Total return in the past 25 years has averaged about 11 per cent pa. Today, income is harder to find - 10-year government bonds only yield 5.7 per cent.

Local investors buying 10-year bonds in 1980 have long since given up their initial double-digit returns, so barring a return to hyper inflation, bond yields would have to fall dramatically to get returns above 10 per cent pa over the next 10 years.

The outperformance of US bonds during the past 25 years is due to the fact that their interest rates have fallen further and they have more longer-term bonds in their universe. Both factors are likely to constrain returns in the future but even NZ bonds are unlikely to repeat their historic performance in the next decade.

* Valuation here refers to the price to earnings ratio (PE). If Telecom's share price was $5 and EPS were 50c its PE multiple would be 10x.