Defensives win again

This is peculiar. In the last three years the FTSE 350 has risen by 13.9 per cent. You would expect low-beta stocks to have underperformed and high-beta ones to have outperformed during this time. But, in fact, the opposite has happened.

This is consistent with long-term evidence from around the world, which shows that defensives - on average - do better than they should. This could be because investors are prone to over-optimism and so pay too much for 'exciting' growth or recovery stocks, and too little for dull defensives, causing these to deliver nice returns to the investor smart enough to buy them.

This seems to overturn the folk wisdom that 'you've got to speculate to accumulate'. And something else challenges this 'wisdom', too - the poor performance of stocks with idiosyncratic volatility. Unlike the performance of defensives, though, this fact is at least consistent with conventional economic theory. This says that investors should not get paid for taking on risk that can be diversified away. And it has been true in the last five years.

However, we should not entirely dismiss the folk wisdom. In the last 12 months, high-beta stocks have done exactly what you would expect, rising 18.3 per cent against the 350's 12.9 per cent. This reminds us that market risk matters, and you are sometimes - though not always - rewarded for taking it on.

One curious feature of our benchmark portfolios just recently has been the great performance of value stocks; our portfolio of the FTSE 350’s 20 highest yielders on 30 June has risen 11.4 per cent in price terms in the last three months. This partially reverses months of appalling performance since 2007.

Greggs has contributed to the success of the low-risk portfolio, which has beaten the FTSE 350.

Explanations

There are two possible reasons why this has happened. One is simply that value stocks tend to expose investors to more cyclical risk than most - which is why they did so terribly in 2008 - and hopes of economic recovery in the last few weeks have caused cyclical risk to pay off well. It might be no accident that value stocks have done well at the same time that small stocks, which are also cyclical, have also risen.

The other explanation is that investors have begun to seek out 'undervalued' stocks, having avoided them for a long time. Loser stocks - the 20 worst performers in the three years to June - have also done well in the last three months, consistent with investors looking to buy recovery plays.

What's more puzzling, however, is the performance of momentum stocks. These (which I define as the previous quarter's 20 best performers) have underperformed the market in the last 12 months, reversing some of their fantastic earlier gains.

This might, of course, be just bad luck: we should never expect even the best investment strategies to always pay off well. But it might warn us of a problem with strategies based upon behavioural finance. There is always the danger that investors will wise up to the mistakes that have caused a set of stocks to be underpriced - in this case, their habit of underreacting to good news - with the result that profits from the strategy will disappear.

This danger doesn't just apply to momentum investing. It also applies to our defensive portfolio. It could be that its nice gains are not evidence of the virtue of defensive investment but rather of the exact opposite - that investors have cottoned onto the case for defensives, and in buying them have bid up their prices to a level from which future returns will be ordinary. The problem is that we just can't tell which it is.

In one respect, however, momentum is still working. Our negative momentum portfolio - the worst performers in the previous quarter - did badly in the third quarter and has underperformed the market in the last 12 months (since we started it). Recent winners might not keep winning, but recent losers carry on losing. In this sense, the old advice to 'cut your losers' is good advice.

And there's one reason to think this might remain the case. It's the short sales constraint. Anyone can buy a stock, but not everyone can sell; many fund managers are unable to sell stocks short, and retail investors are often loath to do so. This means there’s an asymmetry between positive momentum and negative momentum. It's more likely that people will buy shares to take advantage of positive momentum - and thus bid away their gains - than sell shares to eliminate the negative momentum anomaly. We might, therefore, expect the latter to persist even if the positive momentum strategy stops working.

But it's not at all certain that the positive momentum effect has gone for good. Evolutionary finance predicts that investment strategies wax and wane. It works like this. First, an anomaly emerges which some investors - through luck or judgment - buy into and make good profits. Then, others learn of this strategy and copy them. Initially, their buying adds to the profits which earlier investors enjoy on that strategy. But, eventually, it drives prices up and so eliminates those profits. This causes the people who used that strategy to exit the market. But, eventually, their lack of buying causes the shares subject to that anomaly to become underpriced, and so the process starts all over again.

This could be just what happened to value investing. Its huge losses in 2008 caused value investors to leave the market, and their lack of buying caused value stocks to become cheap earlier this year, with the result that value investing has paid off again recently. Maybe momentum's time will come again, therefore.

All of this might sound rather inconclusive. In one sense, this is the point of our benchmark portfolios. Their purpose is to map the investment terrain, to show which risks and which behavioural anomalies pay off and which do not. But the thing about the investment terrain is that it changes. A fixed map - such as 'buy value' or 'buy defensives' - will therefore be right only sometimes.

A further purpose of these portfolios is to be a benchmark against which we can measure whether judgment-based investing works. These portfolios do not use any thought, which means that you should outperform them if you can pick good stocks by using skill. So, for example, if you think of yourself as a defensive investor, your judgment-based stock picks should have beaten our defensive portfolio. Ditto if you're a value or momentum investor.

And herein lies a little bit of good news for judgment investors. In the last 12 months, most of these strategies have underperformed the market ' and the one exception to this, the high-beta portfolio, is only to be expected. This might be a hint that, in stock-picking, brains can be useful. This has not always been the case: in 2007-09, for example, our no-thought momentum and defensives portfolios did better than most judgment-based approaches. Whether it remains the case, though, is something that only time will tell.

Performance of benchmark portfolios