Distinguishing Alpha Factors from Risk Factors in Portfolio Design
Summary
The document explores why portfolio investors distinguish factors that forecast returns from factors used to describe risk exposures. Its central correction is that bearing risk does not guarantee a return premium: some exposures may have a researched premium, while others may not. The response uses high-beta stocks as a counterexample to the assumption that greater risk always earns more, and points to low-volatility research as evidence that the relationship can run in the other direction.
It also emphasizes that factor labels depend partly on the investor’s framework. Value or momentum may be treated as return drivers by one manager and as risks to monitor or diversify by another. The discussion connects this choice to crowding and to whether a manager’s returns offer anything beyond exposure available through factor indices. It does not explain the mechanics of the factor alignment problem or provide empirical tests, so it offers conceptual guidance rather than a procedure for resolving alignment in an optimization model.
Key ideas
- Taking on a risk exposure does not guarantee compensation in the form of a return premium.
- A factor can be classified as a return driver or a risk exposure depending on the investor’s framework.
- High-beta and low-volatility examples challenge the idea that more risk always earns more return.
- Factor crowding and readily available index exposures can affect how investors assess a manager’s returns.
- The response does not specify a method for detecting or correcting factor alignment problems.
Tags
Full text
# Factor alignment problem in portfolio optimization # Factor alignment problem in portfolio optimization I'm taking a course 'AI for trading' in udacity, and there is a part I really want to make sure. The lecture keeps teaching me that there are alpha factors (driver of return) and risk factors (driver of variance) and it also says that alpha factors can turn into risk factors. I've heard that there is a joke between quant 'Your alpha factors are my risk factors' As I study more about factors, it seems that there is no difference between alpha factors and risk factors. Because, by taking risk factors, I mean by making ourselves exposed to those certain risks, we are getting risk premium which is 'alpha' itself. It seems that risk factor and alpha factor are like 2 sides of a coin. It can be flipped anytime. Momentum, Value, Quality factors which were thought to be alpha factors are labeled as 'risk factor' by BARRA. I cannot understand why they keep distinguishing alpha factors from risk factors. However, it seems that what I'm confused about now also made other people confused in the past. I read the paper 'Do risk factors eat alpha factors?' by JH Lee and D Stefek in The Journal of Portfolio Management, Summer 2008 and it discussed the potential problem we can get when we use similar factor both as alpha factor and risk factor. And the problem was called 'Factor Alignment Problem (FAP)' there. Is this a problem which should be dealt seriously when we try to divide factor into alpha and risk factors? How does this all procedure work in real world? I wonder if quants even really use the concept 'alpha factors' or they just think the only factor is 'risk factor'. Moreover, do they concern about 'Factor Alignment Problem'? As an undergraduate student from Korea, sorry if my grammar is bad and hard to read. But I really want to make it clear. Thank you ## Answer by Dhruv Mahajan (score 1) https://quant.stackexchange.com/a/61244 Your argument breaks in this line Because, by taking risk factors, I mean by making ourselves exposed to those certain risks, we are getting risk premium which is 'alpha' itself. Market won’t always give you a premium over risk free rate for taking any kind of risk. Some factors have a researched premium attached to them, some don’t. A perfect example of this is High Beta Stocks, it was a long held belief that stocks with higher beta aka higher risk, will in the long run provide you with a premium over the market. But that’s not true always, there has been quite a few studies on premium of the Low volatility factor, which essentially says the opposite. Also I believe the treatment of risk vs return factors is on the discretion of the investor. AQR uses value as a return driver, but some other hedge funds treat it as risk, I think it comes down to the investor whether he wants to diversify away from the more “crowded” factors. Also if you’re charging 2/20 and your portfolio is mostly driven by the common factors like value or momentum, why would anyone pay you fees when they can get the same exposure from factor indices
Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.