Big, juicy dividend yields abound in today's stock market. The Datastream UK total market index is yielding 4.3 per cent, well above its average of 3.1 per cent over the past decade. Of 36 industry groups, 25 have a yield greater than their 10-year average, compared with a mere eight this time last year. If you're an investor who likes high-income assets, this may well seem like an excellent opportunity to fill up your portfolio with shares.
As tempting as high dividend yields may seem, it's worth remembering that they're high for a reason. Conventional finance theory says you only get super-sized rewards for taking on super-sized risks. The main risk in the case of many high-yielding shares is that the companies will reduce or even scrap their dividend payout. This, in turn, could cause share price weakness, giving investors a double whammy of reduced income and capital losses.
Thankfully, dividends are much more stable than either corporate earnings or share prices. Companies are almost always reluctant to reduce their payout to shareholders unless they really have to. A lowered or missed dividend sends out an extremely negative signal about management's view of a company's prospects. However, there have been times when reduced dividends were quite common. By studying these periods, we can give ourselves a better chance of dodging the unkindest cut of all.
The lessons of history
Since 1965, there have been four periods when dividends on the Datastream Total UK stock market have gone down in absolute terms. The three most significant episodes were all associated with periods of economic turbulence, such as the recession of the early 1990s and the sharp slowdown after the millennium. However, this does not tell the full story. Investors can also end up out of pocket when dividends fail to increase in line with inflation.
Once we adjust the UK stock market's dividend history for inflation, we learn that dividends can go down for extended periods (see 'GDP moves dividends' chart above). Even though the market's dividends rose by 55 per cent between 1966 and 1976, that was not enough to keep up with the massive increase in the cost of living over that period. As a result, the purchasing power of the whole market's dividend fell by 38 per cent during that era.
As well as the torrid time for dividends in the first part of the 1970s, there was also a rough patch in the early 1980s. This episode too was associated with a deep recession. But rising prices across the rest of the economy were also a problem at that time. The lesson for shareholders is clear: we should fear the effects of inflation on our dividends as much as we do the ravages of an economic downturn.
The outlook for dividends today
Economic slowdown and inflation pose a twin threat to our dividends today. The UK economy is cooling rapidly and may well lapse into recession, if it hasn't done so already. At the same time, there are worrying signs that the Bank of England has lost control of inflation. Soaring energy and food costs are feeding through into other areas. Because of the downturn, however, it cannot risk the interest-rate hikes that are needed to tame prices.
There is a worrying parallel here with the 1970s. Having remained subdued for many years, inflation returned with a vengeance in that era. It also occurred alongside economic weakness, confounding those who believed that there was a trade-off between the two phenomena. The reappearance of upward price pressure has taken many people by surprise, but there is still a strong belief that it will subside before long. If that turns out to be wrong, investors should brace themselves for possible dividend cuts, in real terms and perhaps in absolute terms, too.
For the moment at least, British dividends seem well backed up by profits. Dividend cover – the ratio of earnings to dividends – stands at around 2.3 times (see 'Dividend Cover for UK Plc' chart, above). That's comfortably above the average of about 1.9 times since 1965. This is reflected in healthy dividend cover across most stock market sectors. Thirty-one out of 36 sectors have cover greater than their 10-year average level.
While this may sound encouraging, it doesn't necessarily tell us much about the future. These dividend cover figures are based on trailing – rather than forecast – earnings and dividends. In other words, they reflect what we already know: that earnings have gone up a lot in recent years and that they amply covered dividends already paid out.
This is a crucial point when talking about dividends. By nature, they provide a backward-looking view of the world. Total dividends for the UK stock market are at an all-time high. But that has little relevance to the outlook. With the world economy coming off the boil, earnings seem set to fall, thereby reducing companies' ability to pay dividends comfortably.
Just as dividends are at an all-time high even as the stock market is falling, the market will pick up well before dividends do. The long dividend recession of the 1960s and 1970s continued until 1976, well after both the economy and the FTSE had bottomed. More recently, dividends only bottomed in 2004 – a whole year after the stock market had reached its nadir.
The hunt for stable dividends
Although it may not help us much to look at the recent record of dividends, the long-term history is worth considering. Examining previous episodes where dividends have fallen overall can identify which industries are most at risk and which will cope better.
I went through the past 43 years of British stock market data looking for the sectors with the best and worst records of maintaining dividends. As well as looking at actual dividend cuts, I examined effective cuts, in other words when a company fails to raise its payments to shareholders in line with retail price inflation.
The favourite statistical technique in investment analysis for measuring surprises is standard deviation. This figure tells us how widely scattered a set of numbers is around its average. The problem with standard deviation is that it makes no distinction between nasty surprises – such as a dividend cut – and nice ones – such as a bigger-than-average increase. For this reason, I have measured sectors' dividend growth over time using the semi-variance. This statistic considers only changes in dividends below a certain threshold, in this case 0 per cent.
The results of my investigation are shown in the following table. It's too big to display online, but can be viewed as a PDF by (it will open in a new window, and you can download Acrobat software for free here). It certainly provides some food for thought. Over time, the best sector – the one whose dividends have had the lowest downside deviation – has been banking. And during every single dividend recession – defined as those periods when the wider stock market's inflation-adjusted dividends fell – the banks' dividends have fared better than the index's. Household goods & home construction, real estate and general retailers also come out well.
Other steady dividend payers include more classic defensive industries. The gas, water and multiutilities score highly, although this sector only goes back to 1986. Of those areas of the market that do have a complete history going back to 1966, food producers, pharmaceuticals & biotechnology and tobacco all come out well. Food retailers and personal goods also get a look in.
The bottom end of the list is made up of more economically sensitive and risky industries. Oil equipment and services is the riskiest, followed by the two IT-related areas of software and hardware. At various times, these industries have scrapped their dividends altogether, accounting for the large downside deviation figures here. There is a big group that have always produced negative real dividend growth when the wider market has done so, including travel & leisure, oil & gas production, media and chemicals.
