Traditionally, the big danger has been cyclical risk - the tendency for value stocks to suffer in recessions. For example, in 2007 high yielders included mortgage lenders and housebuilders, who collapsed during the great financial crisis.
However, cyclical risk doesn't seem a big danger for value stocks now. High yielders include utilities, tobacco and telecoms stocks, which few would consider cyclical.
Nevertheless, as Tesco and Sainsbury have helpfully reminded us, there's still risk in high yielders. One of these risks is growth risk.
To see this, remember that a high yield can mean one of only two things. Either the market regards the share as unusually risky; and/or it has worse growth prospects than other stocks. These are the only reasons why investors would demand a higher yield on some stocks than others.
But here's the problem. Investors' expectations for growth are often wrong, simply because corporate growth is hugely unpredictable. "Firm growth appears to be an idiosyncratic and fundamentally random process," concluded Alex Coad of the University of Sussex in one survey. And US economists have found that "there is no persistence in long-term earnings growth beyond chance".
Usually, this unpredictability of growth is good for value stocks: if we can't predict corporate growth, why not buy the highest yielders because a low-yielding stock will often be on a premium for no good reason? For this reason, value stocks have - on average - outperformed around the world over the long run.
However, sometimes this unpredictability has a downside for value stocks. If growth is unpredictable then occasionally even low growth expectations won't be low enough. In these cases, even high-yielding stocks will prove to be overpriced - as investors in Tesco and Sainsbury have discovered.
Herein, though, lies a comfort. This risk to value stocks is idiosyncratic. Of course, individual high-yielding stocks or even sectors might fall. But there's no reason, other than luck, why they should do so at the same time as other value stocks do so. Growth risks should therefore have low correlations across high-yielding stocks, and have low correlations with the market generally. This means such risk should be diversifiable; across a basket of stocks, it should largely wash out. For example, our no-thought high-yield portfolio (which comprises the 20 highest yielders) has only slightly underperformed the market in the last three months despite containing Sainsbury and Morrison (and De La Rue) because other high yielding stocks, such as Vodafone and Phoenix, have held up okay.
This poses the question: what would count as a systematic risk to value investors? I suspect, given the current composition of value stocks, that the answer is sentiment risk. If investors become significantly more optimistic, they might actually move out of tobacco and utilities and into 'growth' stocks, causing value to underperform.
