An introduction to harvesting risk premia
Why most so-called alpha was never alpha — and how ordinary investors can capture the risks that actually pay, while skipping the ones that don't.
Most bond investors already live by this idea, even if they never name it. A 10-year Treasury yields more than cash; a corporate bond yields more than a Treasury. Those extra slivers of yield aren't free money — they're a risk premium, the compensation collected for holding something riskier than the safe alternative. Stretch the maturity and you earn a term premium; lend to companies instead of governments and you earn a credit premium.
What fewer people notice is that the stock market is built the same way. Beneath the headline "the market went up," equity returns break down into a handful of distinct risks, each paying its own premium. Learning to see them — and to collect the ones that pay while sidestepping the ones that don't — is most of what systematic investing is about.
What a risk premium actually is
A risk premium is the reward for bearing a discomfort other investors would rather avoid. Someone has to hold the volatile asset, the illiquid one, the one that hurts most in bad times — and on average, over long stretches, they tend to be paid for it.
That word average matters. A premium is not a guarantee; it's a tendency that shows up over years and cycles, not days. In the short run it can vanish or invert — which is part of why it pays at all. If the reward were smooth and certain, it would be competed away.
The slow disappearance of alpha
For decades, a manager who beat the market was assumed to have skill — alpha, in the jargon. Then researchers began decomposing those returns, and the story changed.
It started with the Capital Asset Pricing Model in the 1960s, which split a portfolio's return in two: beta, its exposure to the broad market, and alpha, whatever was left over and credited to skill. Buy high-octane stocks in a rising market and you would outperform — but that was beta, not genius.
Then, in the early 1990s, Eugene Fama and Kenneth French showed there was more than one beta. Cheap "value" stocks and small companies had historically out-earned the market by margins too large and too persistent to be luck. Others followed: momentum, quality, low volatility, profitability. Each was a distinct, documented driver of returns — a risk premium hiding inside what people had been calling skill. (Fama shared the 2013 Nobel for this body of empirical work.)
The uncomfortable implication: a great deal of the "alpha" investors paid handsomely for was really factor exposure in disguise — premia that could be captured systematically, without the star manager or the star fee. And genuine alpha — being right when the crowd is wrong — turns out to be scarce and getting scarcer. Year after year, once costs are counted, the majority of active funds trail their benchmarks. Alpha is shrinking; the risk premia underneath it are not.
The premia hiding inside the market
A few of the best-documented equity premia, in plain terms:
- Value — cheaper stocks (relative to fundamentals) tending to beat expensive ones over time.
- Size — smaller companies carrying higher long-run returns, and higher risk, than large ones.
- Momentum — recent winners tending to keep winning for a while, and losers to keep losing.
- Quality / profitability — durable, profitable businesses tending to outperform fragile ones.
- Low volatility — calmer stocks delivering surprisingly competitive returns for the risk they carry.
Why do they persist? It varies. Some look like pure compensation for risk — you get paid because the asset really does hurt at the worst moments. Others appear to survive for behavioural reasons: people chase stories, overreact, and crowd into the familiar, leaving a repeatable pattern behind. The label matters less than the test worth applying to all of them: is there a clear, durable reason this should keep paying, or did it merely happen to work in the past?
Why this is good news for ordinary investors
Here's the cheering part for anyone without a hedge fund: capturing these premia no longer requires finding a once-in-a-generation manager. Because they come from rules — cheapness, trend, profitability — they can be harvested systematically and transparently. That is the idea behind so-called smart beta, or factor investing: rules-based portfolios that deliberately tilt toward a premium instead of hoping a manager stumbles into it.
It's the same lineage we sit in — systematic, rules-based, premium-driven — and the opposite of paying up for vanishing alpha.
The catch: not all risk pays
But there's a trap, and it's the part we care most about.
The naïve way to "do factors" is to buy a handful of off-the-shelf factor funds and bolt them together. Two problems follow. First, overlap: the same stocks show up across several funds, so a portfolio that looks diversified is quietly making one concentrated bet. Second, and more important, unrewarded risk.
Not every risk carries a premium. Interest-rate sensitivity, oil-price sensitivity, an accidental tilt toward one sector or country — these can swing a portfolio around violently, yet over the long run the market does not pay you for holding them. They are pure volatility with no paycheck attached. Left unmanaged — as a basket of single-factor funds tends to leave them — they can swamp the very premia you were trying to earn.
The discipline, then, is twofold: harvest the risks that are compensated, and deliberately neutralise the ones that aren't. Get paid for what pays; refuse to carry what doesn't.
Where this meets our work
This is the soil Risk Harvest grows in. We treat returns as something to be built from distinct, well-understood premia rather than conjured by prediction — combining genuinely different drivers so they diversify one another, sizing everything to a steady level of risk, and working to strip out the macro and sector exposures that add noise without adding return.
Same family as smart beta, in other words — just combined with more care, and with a stubborn refusal to take on risk that doesn't pay. We're still learning and still researching; but that principle — only bear the risks you're compensated for — is the one we keep coming back to.