2017 was another year in which astute sector pickers would have comfortably outperformed the S&P 500.
Technology led the way, up by more than 30%, while energy languished with a loss of more than 5%, as the S&P 500 closed in on a 20% gain for the year.
Given this wide dispersion of sector returns, though, it would be helpful to have a more systematic process for moving between them. A forthcoming paper by Golam Sarwar and colleagues in the Journal of Asset Management may offer just that.
Sarwar and his fellow authors propose rotating between sectors based on five-factor alpha. Using the Fama-French five-factor model – market risk, value, size, profitability and investment – they calculated the rolling 36-month alpha for each sector between January 1967 and December 2014.
The trading trigger was simple: They bought only the sectors with positive five-factor alpha over the preceding three years, rebalancing monthly. ‘We do not claim that our choice of the alpha estimation period or the rebalancing frequency is optimal; it is simply used to explore the benefits of sector rotation based on five-factor alphas,’ they noted.
Rotate for dizzying returns
Nevertheless, those benefits were considerable. If this methodology is applied to State Street’s suite of SPDR sector ETFs between January 1999 and December 2014, these funds handsomely beat the S&P 500.
A long-only portfolio of the sectors with positive alpha returned an annual average of 5.5% over this period; the equivalent gain from buying and holding the S&P 500 was a mere 2.1% per year. The long-only portfolio was also slightly less volatile than the broader index.
Interestingly, when Sarwar and his colleagues based their trading on Fama and French’s three-factor model instead – which excludes the controversial profitability and investment factors that have been greeted with skepticism from some academics – the results were weaker. That three-factor portfolio returned just 3.8% a year –still ahead of the S&P 500, but by a smaller margin.
The most impressive outperformance, however, came with an extra innovation: selling out of equities altogether in recessions. From the start of 1999 to the end of 2014, there were 26 recessionary months and 166 expansionary months, according to the National Bureau of Economic Research. Sarwar and his associates bought the sectors with positive alpha as usual in the up months, but switched the entire portfolio into one-month Treasury bills during the downturns.
Doing so boosted returns hugely. The recession-proof portfolio returned an annualized 9.2% through the review period, almost five times better than the S&P 500 and with a Sharpe ratio more than six times better thanks to its lower volatility.
Trying to implement the strategy on a long/short basis, however, proved unsuccessful. Being long the positive-alpha sectors and short the negative ones delivered an annual average return of just 0.5% – well below that of the S&P 500, albeit with less than half the volatility.
‘This can infer some conclusions about portfolio persistence: positive alphas utilized in the long-only strategies are more likely to lead to future positive alphas, while negative alphas are not a good predictor of future negative alphas, leading to poor performance of the long/short strategies that utilize them,’ the authors commented.
It is worth noting that Sarwar and his colleagues’ long-only results from sector-picking may actually understate the potential profits.
They used the SPDR sector ETFs for their study, which carry an expense ratio of 0.14%. They also estimated that their average round-trip transaction cost when trading, including the bid/ask spreads on the ETFs, would be 25 basis points.
The SPDR ETFs, though, are not the cheapest sector products. Fidelity has a range of 11 sector ETFs with net expense ratios of just 0.084%, and Vanguard’s equivalent ETFs cost 0.1% each. iShares, on the other hand, has a wider selection of sector ETFs – including the likes of medical devices and regional banks – which charge between 0.44% and 0.48%. It also has multi-factor sector ETFs for 35 basis points.
In any case, the approach identified by Sarwar and his colleagues – essentially applying alpha momentum to sectors – is intriguing and easily executable. So much so, it may even lead to its own ETF. After all, there is already the Dorsey Wright Sector Momentum ETF, but this focuses on price rather than alpha momentum and also buys stocks from within sectors rather than the whole sector. Launched in January 2017, it has returned 19.7% since inception, compared with 18.8% from the S&P 500. Promising signs indeed.