Factor investing: what actually drives returns
A factor is a rule that historically paid. The danger is that once everyone knows the rule, the payment stops.
Understand what a factor fund is actually buying on your behalf.
How Factor investing: what actually drives returns works, in one picture
The same argument as the text, as a chain. Each step is what makes the next one possible.
What a 2% fee costs over thirty years
Both lines earn the same 8%. One pays 0.07% a year, the other pays 2%. The gap is not a rounding error, it is most of the point of the exercise.
- 1
The evidence: some characteristics have paid, persistently
Cheap stocks (value), smaller ones (size), recent winners (momentum), profitable and stable ones (quality), and less volatile ones. Across many decades and many markets, portfolios tilted this way have on average beaten the market.
- 2
Two explanations, and they imply opposite futures
Either the premium is compensation for real risk, in which case it should persist because someone must be paid to bear that risk. Or it is a behavioural mistake, in which case it should erode as more people exploit it.
Which one you believe determines whether you expect the last thirty years to repeat.
Value spent well over a decade underperforming badly. Anyone who tells you a factor works reliably has not lived through that.
- 3
The pain is the price of admission
Every factor has long, brutal stretches of underperformance. That is arguably why the premium survives: most investors cannot endure it, abandon the strategy at the bottom, and hand the return to whoever could.
- 4
And check what you are actually being sold
'Smart beta' funds vary enormously in how they define a factor, and the fees can be several times a plain tracker's. A quality tilt at 0.6% has to beat a market tracker at 0.07% before you are ahead.
You can name the factor your fund tilts to, and the longest stretch it has ever underperformed.
Read next
A factor is a rule that historically paid. The danger is that once everyone knows the rule, the payment stops.
Moat scores for 1,100+ companies, with the returns on capital that either back the score up or quietly contradict it.