Why does momentum work?
- Created:
- 12 November 2007
- Updated:
- 25 January 2008
- Written by:
- Chris Dillow
Momentum investing is still working well. Since 28 September (the last trading day in September), our momentum portfolio (the 20 best performing FTSE 350 stocks in the previous calendar quarter) has returned over 5 per cent despite a dip in the general market. The momentum portfolio has now returned 82.1 per cent in the last two years, compared to just 27.7 per cent for the FTSE 350.
But why does momentum investing work? Behavioural finance says it's because of a mix of under- and over-reaction. If a dull stock announces good news, investors initially might under-react to it, as they insufficiently update their Bayesian priors about the stock. Only gradually will their scepticism wane. As it does, good returns lead to further good returns; this gives us post-earnings announcement drift.
Better still, under-reaction eventually gives way to over-reaction. If a stock rises a long way, investors jump onto the bandwagon, causing it to rise further.
Sadly, though, there are problems with this story. One is: why is it that investors are only sometimes irrational? Momentum investing failed for most of 2005. Why were investors smart then, but stupid since?
Also, if investors over-react, you'd expect them to sell bad stocks too much, causing these to eventually become under-priced and so bounce back. But this hasn't happened recently. Our loser portfolio (the 20 stocks in the FTSE 350 with the worst three-year returns) has consistently under-performed the market.
Perhaps, then, momentum profits aren't due to investor irrationality at all, but rather are just a reward for taking risk. The fact that momentum did badly in May 2006 and this summer should alert us to this possibility.
Researchers suspect there might be three types of risk.
One, says Tim Johnson of the London Business School, is growth rate risk. A given amount of bad news about a stock's growth prospects hurts fast-growing stocks harder than slower-growing ones; remember how tech stocks collapsed in 2001? And because momentum stocks tend to be fast-growers - which is why their prices have risen so much - they'll carry more of this risk than stocks generally.
Another risk, says Paul Zurek of the University of Pennsylvania, is that momentum stocks' dividends are sensitive to consumer spending. This means they are likely to do especially badly if we get bad news about future economic growth that causes us to cut our spending.
And other researchers - such as Lakshmanan Shivakumar at the London Business School and Lu Zhang of the University of Michigan - believe momentum stocks are correlated with macroeconomic conditions and so do badly if the economy falters, although some others are sceptical of this.
Although we can't be sure of the precise dangers in momentum investing, it's highly likely that risk, rather than irrationality, explains at least some of their profits.
Q4 Momentum portfolio, performance to date
| Constituent |
Price on 28/9 |
Price on 8/11 |
% gain/loss |
| AMEC |
1929.46 |
2019.3 |
4.7% |
| Amlin |
712.71 |
704.6 |
-1.1% |
| Antofagasta |
243210.3 |
249749.1 |
2.7% |
| Autonomy |
31.44 |
35.2 |
12.0% |
| BAE Systems |
3879.08 |
4044.11 |
4.3% |
| Beazley |
283.71 |
249.6 |
-12.0% |
| BHP Billiton |
1131.67 |
1070.89 |
-5.4% |
| Cairn Energy |
1068.41 |
1237.16 |
15.8% |
| Ferrexpo |
185.71 |
177.86 |
-4.2% |
| IG |
332.72 |
358.86 |
7.9% |
| London SE |
642.57 |
686.37 |
6.8% |
| Micro Focus |
240.69 |
263.28 |
9.4% |
| National Express |
1200.94 |
1202.89 |
0.2% |
| Randgold |
336.97 |
360.72 |
7.0% |
| Stagecoach |
1602.15 |
1729.06 |
7.9% |
| Talvivaara |
115.8 |
120 |
3.6% |
| Tullow |
3805.35 |
4221.67 |
10.9% |
| Vedanta |
548.41 |
595.68 |
8.6% |
| Wellstream |
254.84 |
297.97 |
16.9% |
| Wood, John |
215.71 |
233.68 |
8.3% |
| Average gain |
|
|
5.3% |
| FTSE350 |
3970.44 |
3936.50 |
0.0% |