Secrets of Success: Cash is king - and so are the dividends

Click to follow
The Independent Online

One of the first bits of investment advice I remember being given was that there is a useful rule of thumb about how attractive the stock market is at any time: when its average dividend yield is 3 per cent or less, the market is clearly a sell. When it is 5 per cent or more, it is time to buy.

One of the first bits of investment advice I remember being given was that there is a useful rule of thumb about how attractive the stock market is at any time: when its average dividend yield is 3 per cent or less, the market is clearly a sell. When it is 5 per cent or more, it is time to buy.

The rule has the benefit of being simple and straightforward, if not always hugely useful: there simply are not many periods when the market yield goes outside these parameters, or, at least, not enough of them to make it that useful in day-to-day investment.

With the help of the Global Investment Year Book, the ABN Amro/London Business School study of long-run market returns I referred to last week, it is possible to take a deeper look at how useful the dividend yield is as an indicator of the market level and prospects. It turns out, perhaps not surprisingly, to have been a good indicator, just as my old market sage insisted it should be.

What the data of the last 104 years shows clearly is that buying shares when the market yield is low, and selling or reducing equity exposure when it is high, has proved a successful approach. The graph below summarises the real (after inflation) returns delivered by the market over holding periods of from one to 15 years in two cases: one, when the starting market yield is high by historic standards and two, when it is low by historic standards.

The results show buying shares when the market yield is high delivers higher real returns over all periods than when the yield is low. The differential return - the gap between the two series - does narrow the further out in time you measure the returns, but it remains a substantial one. The high-yield cases appear to level out at about an 8 per cent real return on a five-year plus view, roughly double the average real return offered by equities over the last century.

The definition of high yield is based on whether the starting yield is high by historic experience at the time, not by comparison with the average over the whole measured period (1900-2003). This is to avoid the statistical pitfall of data mining, assuming the whole sample period is necessarily going to be typical of the future as well as the past. The truth is that we do not know. In fact, there is some evidence in the data that there has been a general slide in the level of dividend yields over the past 100 years. The UK market yield was 5 per cent in 1950, just as it was in 1900; but in 2000, admittedly an extreme case, being the top of the bull market, the dividend yield had fallen to an all-time low of barely 2 per cent. This pattern of falling yields was common to most other countries with stock markets.

There are reasons why this might be a trend that continues. One plausible one is that equity markets around the world are so much bigger and more diversified than they were 100 years ago that investors are justified in lowering the returns they require to justify the greater risk of holding shares rather than bonds or cash. Changing tax regimes and corporate fashion have also clearly had an impact on the proportion of earnings that companies pay out as dividends.

If there has been a permanent change in the level of dividend yield required by investors, looking forward from here, it might be reasonable to reduce the level of dividend yield at which you can say the market is either cheap or expensive. The level of dividend yield cannot be divorced from the yields on competing assets (and hence the level of inflation).You only have to look back to the final years of the 1990s bull market to see the absurd lengths to which the argument that "dividend yields should be permanently lower" can be taken.

The argument about dividend yields speaks to the issue of what your investment objectives and aptitude for risk are. Following the "sell below 3 per cent" rule would have taken you out of the stock market in 1998, and meant you missed the final 15 months of the bull market: but, just as importantly, it would have spared you the worst of the bear market.

Equally, the "buy at 5 per cent" rule would have forced you into the market through the second half of the 1970s and early 1980s, and again in 1990-91, both of which were outstanding times to have increased your equity exposure. The rules would also have worked very well in the earlier part of the century, when a market dividend yield of 5 per cent was a more frequent occurrence.

Reinvested dividend income has always accounted for the lion's share of the total return investors have obtained from shares over the very long term. Over the period 1900-2003 as a whole, the proportion is as high as 85 per cent.

And if you look at the dividend yield of the market today, the signal is giving mixed messages. For the FTSE 100 index, at 4.2 per cent, the omens are not that bad,in part reflecting the fact that large-cap stocks are now out of favour, but for the broader market indices (UK 3.1 per cent, the US market 1.9 per cent), the danger signals are again flashing for any passive long-term investor.

While there is a tactical case for retaining quite high equity market exposure, as the current market recovery looks to have the potential to run some way yet, there is clearly strategic risk in that position. For long-term investors, high-yielding quality stocks will remain a safer haven, as they almost always are for long-term investors.

Looking for credit card or current account deals? Search here