Such an approach has been shown to work nicely for the US stock market. In a 2010 paper, Mebane Faber, a money manager and researcher tested the effects of investing in the strongest of the ten broad industry groupings between 1928 and 2009. (Relative Strength Strategies for Investing – Mebane T. Faber, Cambria Investment Management.) He found that the resulting portfolios of up to three sectors led to greater absolute returns and also outperformance of buy-and-hold investing in seven years out of ten.
UK market by sector
The UK stock market can be split into 10 broad categories - oil & gas, basic materials, industrials, consumer goods, healthcare, consumer services, utilities, financials and technology. The chart above shows their current weightings, while the table shows their performances over time I've therefore back-tested various strategies of buying every month into the sector or pair of sector groupings that have done best just lately. As well as learning whether this simple momentum approach worked or not, I wanted to understand the risks involved and whether the returns were consistent over time.
|1965-2014||Oil & Gas -||Basic Mats||Industrials||Consumer Gds||Health Care||Consumer Svs||Telecom||Utilities||Financials||Technology||Market|
More return, less risk
Had you bought the best-performing two UK industry groupings of the last rolling three month period every single month between 1965 and early 2014, and then sold or rebalanced one month later, you would have made a total return of 20.1 per cent a year. Buying and holding the Datastream UK Total Market index - a look-alike of the FTSE All-Share index would have generated a mere 13 per cent annualised return. Put slightly differently, a £1,000 investment in 1965 would have grown to £7.66m with this strategy, compared to £384,451 for buy-and-hold.
Is three months the best period for measuring recent momentum? In the case of UK sectors, it has been. But I also tested how you’d have done buying the winner of the last one, six, nine and twelve months. They all produced market-beating returns in the high teens, although none quite as strong as for buying the winner of the last quarter. This isn’t to say there’s any particular significance to three months, although some other studies have also settled on this as the best timeframe.
You might think that sticking everything into just two areas of the stock market each month would be horribly risky. However, you'd be wrong: buying the leading two sectors incurred not much more volatility than buying the whole market. And this strategy suffered less severe losses at its very worst moment, too. Whereas a buy-and-hold investor would have seen their index holdings tumble 63 per cent at one point, a momentum investor's worst drawdown was 51.2 per cent.
Which sectors would you have ended up investing in most with this strategy? The accompanying chart shows how often each sector was represented in the momentum portfolio, compared to its average weighting in the market over time. Most notably, the technology sector was the most frequent constituent of the portfolio, whereas it has historically been the lowest-weighted sector in the UK market. Consumer goods and industrials also figured far more than their long-run average weightings.
Momentum portfolio appearances vs weighting
When momentum works best
Over time, my proposed strategy would have beaten passive investment in the index in around 55 per cent of months and in some three-quarters of calendar years. So, not only would this approach have beaten the market, it would have done so for most of the time. By contrast, some other outperforming strategies win less than half the time, which can make them offputting to professional managers, who need to be seen to be winning for as much of the time as possible, or face the sack.
While the sector momentum strategy has done better than buy-and-hold overall, there are clearly times when it has done especially well or badly. The accompanying chart tries to capture this by showing how many months out of the last six momentum has beaten buy-and-hold. Times of momentum strength are therefore when you get five or six months out of six where momentum wins, and times of weakness are therefore where it has outperformed in two or fewer of the last six months.
When momentum wins
As we might expect, sector momentum has worked well in some early-stage bull markets such as recoveries from the massive slumps of 1974, 2000 and 2007. But it has also done well during later-stage uptrends such as in 1999 and 2006, as well as in mid-stage rallies such as 1984, 1996 or 2012. That said, sector momentum doesn't automatically do well when the UK market is heading higher. It had rotten spells in 1985, 1986 and in 2003-4.
Sector momentum outperformance isn't merely a bull-market phenomenon, though. The approach had good streaks during dark days of late 2008 and early 2009, as well as in the horrendous meltdown of 1974. However, it did conspicuously badly during other shakeouts such as those of 1969 and early 2001.
Every loser wins?
How would you have done if you’d systematically invested instead in the worst-performing recent UK sectors? Had you bought the two worst performing groups of the last quarter every month, you’d have made just 7.5 per cent a year, versus 20.1 per cent for the best two and 13 per cent for the market as a whole. Or, to put it another way, a £1000 investment in 1965 would have become £34,817 by the end of January this year, a tiny fraction of what you’d have made from buying recent winners, or even from buy-and-hold.
However, this is far from saying that buying losers never has its day in the sun. In fact, this approach has beaten buy-and-hold investing in 17 calendar years since 1968. It did especially well in the recovery from the disaster of 1974, when it returned 200 per cent, compared to 151 per cent from buy-and-hold, as well as three years in a row coming out of the dot.com collapse of the early 2000s.
The gains to be had from a UK sector momentum strategy clearly seem to be worth going for. But are they actually achievable in practice?
In the US, investing along the broad sectoral lines suggested here has become possible thanks to the introduction instruments like the SPDR family of exchange traded funds. Unfortunately, no such super-sector trackers exist yet for the UK market. In due course, as ETFs gain even more in popularity here, this could well change.
Another consideration is the frequency of the trading involved. Buying the top two sectors each month involves making twenty-four round trips a year, either to purchase a new sector or to rebalance the portfolio if the constituents are remaining the same. This is plainly a costly exercise, although historic outperformance would have more than made up for the additional dealing costs. As a future project, therefore, I am going to experiment with other holding periods, such as three months.
The weighting of the portfolio is another potential criticism. Equally weighting a portfolio often tends to bias the returns upwards, but is unrealistic in some people’s eyes. An alternative would be to use market weightings when determining the holdings.
While my strategy is not easily investible, it offers a good way to track the performance of momentum as a style and also for further research into the characteristics