Is there such a thing as the equity risk premium? I ask because the gap between the dividend yield on the All-Share index and long-dated index-linked gilt yields has been rising for most of the last 17 years, and many conventional explanations for this don’t make much sense.
Traditionally, we’ve thought of this gap as being equal to the risk premium on equities over gilts (the extra expected return on equities investors require to compensate them for their extra risk) minus expected divided growth. If expected growth is high, or if investors perceive little risk, investors will accept a low yield on equities. If not, they’ll require a higher one.
For much of the 1980s and 1990s, this yield gap oscillated around 0.5 percentage points. Since around 2002, though, it has trended ever upwards and is now almost six percentage points.
Some of this increase might be due to lower expectations for dividend growth. (Whether this is because of lower GDP growth or because growth might come from companies not listed on the market doesn’t matter for our purposes.) But only some. A reasonable expectation for real dividend growth in the 1980s and 1990s would have been around 2 per cent per year: actual growth between 1985 and 2000 was 1.5 per cent. Unless investors now expect dividends to fall sharply year after year, lower growth expectations can only explain a fraction of the higher dividend-index-linked yield gap.
Conventional thinking, therefore, says the higher gap must be mostly because investors think that equities are riskier now than they thought in the 1980s and 1990s.
And there are good reasons to believe this. Political risk is greater now: there are credible threats of nationalisation and protectionism. At near-zero interest rates conventional monetary policy can do little to mitigate recessions, which means these are a scarier prospect. Investors are now more aware of types of risk other than ordinary volatility, such as the danger of liquidity drying up or that the small chance of disaster will actually materialise. And it’s possible that memories of the 2008 crisis have had a scarring effect, making investors more risk averse even today.
Whilst there is some truth in these theories, they are inconsistent with two facts.
One is that the dividend-index-linked yield gap was rising in 2002-07, a time when it is (now) widely thought that investors were actually being too complacent about risk.
The other is that this gap is higher now than it was at the peak of the 2008 crisis. But the idea that investors are more nervous now than they were then just isn’t plausible. Scars are supposed to heal with time, but this one seems to have become more livid. Which is weird.
Perhaps, therefore, we should look for an alternative explanation for the rise in the dividend-index-linked yield gap. This begins from a simple fact – that the rise is almost entirely due to falling bond yields: the dividend yield hasn’t much changed since the 1980s and 1990s.
The question then is: why has the dividend yield not fallen as bond yields have done so?
For this to have happened, sufficient numbers of investors must have switched out of bonds and into equities. But there have been two barriers to them doing so.
One is that many of the investors who bought government bonds in the 1990s and 2000s did so because they were already exposed to lots of risk and wanted safe assets to mitigate those risks. I’m thinking here of Asian and Middle Eastern investors who have feared falling commodity prices, political unrest, fragile banking systems, the insecurity of their property rights and just ordinary cyclical risks to their domestic businesses. These investors don’t want to take on the risks associated with western equities, as these are correlated with the risks they already face. They want to lay risk off with bonds.
A second barrier is convention. To buy equities as bond yields fall requires us to believe that ultra-low dividend yields are sustainable – and, what is much more, that other investors will also believe they are. But history gives us reason to doubt this. For years and years, dividend yields have usually been in a 3-5 per cent range, and the tech crash taught us that sub-3 per cent yields are not sustainable. Investors have, therefore, been understandably reluctant to push yields down.
Which leads to a nice irony. In 1936 Maynard Keynes wrote that “the rate of interest is a highly conventional, rather than a highly psychological, phenomenon... Any level of interest which is accepted with sufficient conviction as likely to be durable will be durable.” This turns out not to have been true of interest rates. But it does seem to have been true of equity yields; yields of 3-5 per cent are accepted as durable and therefore are durable.
If all this is correct, or roughly so, then the dividend-index-linked yield gap has risen not because of heightened risk aversion but simply because the dividend yield has been stable for other reasons in the face of falling bond yields.
This might mean that there’s no such thing as the risk premium, as Eric Falkenstein at Pine River Capital Management has argued. Or it might mean the risk premium should be measured in a different way. Whichever it is, the dividend-index-linked yield gap might not be telling us very much.
