Oneof the first questions most investment advisers ask their prospective clients is "how much income do you need from your investments?"

If the answer is "not much", the adviser will usually make sure the investor's portfolio includes lots of shares, and that many of those shares are "growth" stocks - ones that produce little or nothing in the way of dividends.

On the face of it, this makes perfect sense. All other things being equal, companies which retain their earnings and reinvest the money in the business should be expected to reward shareholders with more capital growth, in the form of a rising share price.


Indeed, the 1990s bull market in US technology stocks was partly based on the strategy of keeping dividend payments low and reinvesting the money in the company. That approach has been associated with US and European shares, while this country's market has been seen as predominantly made up of "income" or "value" stocks.

In the mid to late 1990s, many local investment advisers swallowed the idea that their clients should invest overseas to share in the growth story, and switched some of those clients out of high-dividend local shares into growth stocks, either directly or via managed funds. Advisers and their clients were further encouraged by the fact that capital growth is usually tax free, while income is taxable.

A simple strategy but, according to a paper in the Financial Analysts Journal, simply wrong.

The article shows that if you want capital growth you should, perversely, buy shares with high dividends. A focus on higher-yielding shares provides better profit growth than buying low-yielding ones. This is yet one more paradox in the investment world, along with other apparent contradictions such as share prices rising on bad news and high-quality bonds falling in price when economic conditions improve.

Written by two US investment gurus, Robert D. Arnott and Clifford S. Asness, the article "investigates whether dividend policy, as observed in the payout ratio of the US equity market portfolio, forecasts future aggregate earnings growth.

"The historical evidence strongly suggests that expected future earnings growth is fastest when payout ratios are high and slowest when payout ratios are low. Our evidence thus contradicts the views of many who believe that substantial reinvestment of retained earnings will fuel faster future earnings growth.

"Rather, it is consistent with anecdotal tales about managers signalling their earnings expectations through dividends or engaging, at times, in inefficient empire building.

"Conversely, financing through share issuance and paying substantial dividends, although perhaps less tax efficient, may subject management to more scrutiny, reduce conflicts of interest, and thus curtail empire building."

Telecom may be a good example. A few years ago the directors reduced the dividend. They then blew the money on a technology-inspired spending spree in Australia, and Telecom's share price nosedived.

More recently, management has been making noises about a return to higher dividends and Telecom's share price is looking well supported by buyers.

Arnott and Asness conclude: "Unlike optimistic new-paradigm advocates, we found that low payout ratios historically precede low earnings growth. This relationship is statistically strong and robust.

"We found that the empirical facts conform to a world in which managers possess private information that causes them to pay out a large share of earnings when they are optimistic that dividend cuts will not be necessary and to pay out a small share when they are pessimistic, perhaps so that they can be confident of maintaining the dividend payouts.

"Alternatively, the facts also fit a world in which low payout ratios lead to, or come with, inefficient empire building and the funding of less-than-ideal projects and investments, leading to poor subsequent growth, whereas high payout ratios lead to more carefully chosen projects."

So what does it mean here?

In the recent past local investment experts have recommended share portfolios with 90 per cent in overseas shares and 10 per cent local.

Advisers who took that approach explained away the apparent cheapness of local shares, compared with US shares, by arguing that those US shares retained more of their profits and could therefore be expected to achieve greater share price growth.

The long-term performance of New Zealand shares, compared with those in the US, looks much more respectable given the huge difference in returns over the past five years. The good news is that our local sharemarket - and to a lesser extent the Australian market - is full of companies with relatively high payout ratios, unlike the US.

The dividend on New Zealand shares, at 5 or 6 per cent, is much more attractive than the 2 per cent or so typical of the US and other European markets (dividend imputation and lower prices also help).

Maybe a practical application of modern portfolio theory is that, while we do need to diversify our share portfolios globally, New Zealand shares merit a higher weighting than 10 per cent and Australian shares should compose much more of the "international" sector than the 3 per cent that the market's size implies.

* Brent Sheather is a Whakatane investment adviser.