Today’s topic is risk management and how easy it is for well intentioned people to get over their skis and do really dumb things. For once, the really dumb things aren’t being done by insurance companies! No, these rubes work in the investment industry. Today, we discuss the fallacy of factor model investing!
And for a special bonus, as I realized the length of this post was getting out of control, I decided to break it up into two parts for your reading pleasure so expect Part II later in the week!
What Is Factor Investing?
Glad, you asked. Factor investing is a misguided approach to managing risk that mainly exists for marketing reasons (sort of like ESG investing).
In short, factor investing says portfolio managers should hedge all their risk to “factors” – such as industry exposure, market cap exposure, momentum, value vs. growth, etc. – and only produce returns that are “stock specific”.
In other words, don’t be like the retail yahoos running the market today being all long tech and short everything else like financials and energy…or like an ESG fund long tech and short financials and energy…or like the S&P 500 which is long technology and short financials and energy. You get the point.
No, factor investing says you should have 0 net exposure to tech and 0 net exposure to energy. In other words, every long in technology requires a short in technology.
But if only it were this simple…industry weighting is only one factor. There are about a dozen factors overall (depending on the model) and you need to minimize your risk to all of them, so you can’t be neutral tech by being long Netflix and short IBM cause now you have a momentum bet.
Why Do We Like Factor Investing?
Because it forces “discipline” and ascertains “skill”. In factor investing, the only way to generate positive returns (in theory – more on this in a minute) is to find two completely similar stocks and figure out which will outperform the other due to some company specific insight, e.g. one has an earnings surprise and the other a miss.
Most hedge funds, and increasingly most long only investors, now subscribe to a factor based approach and this is a problem (but more on that in a minute too).
Why did this approach become so popular? Marketing! It’s a very persuasive marketing message to potential investors to say “we’re not trying to time the market. We’re only focused on pairing one company off against the other”.
I should know. I worked at one of the first practitioners of factor investing before it became popular. Yes, I was an early adapter. Back then, it had an advantage beyond marketing…other people weren’t doing it.
This was very helpful as you could look for stock specific disparities that others ignored because they were making industry calls or unaware that what they thought was a fundamental bet was actually a bet on say value or large caps and they didn’t realize it.
But like any investment strategy that has success, people copy it – especially when the marketing pitch is so compelling. Soon, every hedge fund was looking at the same factors and buying the same stocks. Isn’t this great? They’ll all outperform with little risk!
When Factor Investing Became A Factor!
No, that isn’t how it works. When everyone copies each other and is making the same bets, that’s a really scary time to be an investor. Volatility increases because liquidity only goes in one direction (if everyone is long company A and short company B, there is no liquidity to sell A if there is bad news and it gets crushed).
Performance is not generated by identifying that company A is likely to do better than company B. No, performance is generated by being the first one to recognize A will outperform B and the first to exit once the new information is recognized in the stock price. In other words, your margin for error just got infinitely smaller.
This led to a new addition to the model. Firms created a “faux factor” called crowding! Yes, if too many people owned the stock you owned, now you had a new risk. The risk of popularity!
So, in order to remove your factor risk, you have to find enough stocks you like and don’t like that allow you to minimize your factor risk…but at the same time, you can’t like or not like the same ones your competitors like and don’t like!!!
When Factor Investing Broke
Maybe you’re saying that doesn’t sound so hard? No, it’s very hard. Why? Because rarely are there two stocks with identical factor characteristics, so you can never truly make a bet that clearly demonstrates your prowess at “stock picking”.
There are always trade offs being made where you might like a stock but can’t find ways to fit it into the portfolio because of the effect it has on your overall factor risk.
Because the universe of stocks that a) are in your wheelhouse (most firms using factor models have industry specialists thus restricting the breadth of their stock coverage) b) you have a compelling fundamental view on, c) aren’t owned by too many other people, and d) lower rather than raise your factor risk is small, you end up being limited to a very small number of choices.
This constraint of choice means any returns generated by stock picking skill are diminished by the limited opportunities to deploy capital. In other words, you’d be better off in a less constraining system that generated somewhat lower percentage returns but allowed for greater amounts of capital to be deployed.
In other, other words, you’d be better off finding the next new thing that nobody is doing yet, rather than continue to follow a factor investing approach.
There Is No Moral Value In Factor Investing
The biggest error I see with people who utilize factor investing is they treat it like a belief system, as if it is wisdom carried down from the gods!
There is nothing intrinsically good or bad about factor investing, just like there is nothing intrinsically good about value or growth investing. It is a style. It will work well at times and not so well at others.
As I have said, the reason it has persisted is because it proved popular when marketing funds. If that ever changes, factor investing will largely disappear. It is popular because of its popularity, not on its merit.
To say it a little differently, there is nothing wrong with making money by market timing. It is hard to do well, but it’s certainly worked well this year.
There is not even anything wrong with explicitly betting on factors. People who bet on growth over value have been persistently right for about a decade now and their returns dwarf any factor neutral approach.
Fund consultants think it is easier to show if a manager is good at factor neutral investing and harder to show they are good at betting explicitly on factors, but it is likely only a matter of time before someone figures out how to “prove” someone is good at selecting which factors to bet on and then that will be in vogue.
Why The Long Prelude to Part II?
Because I realized what I want to get to in Part II won’t be that relevant without the background provided above. Next time, I will explore where factor models break down beyond the issues highlighted above.
Specifically, I’ll address how the models themselves are used incorrectly and what risk management challenges that poses. As a teaser, I’ll leave you with we are amidst a period where the models are most likely to be significantly wrong.