Factor checklist:
4 requirements for a robust factor

Andrew Ang |13 set 2017

The capacity of factor strategies appears large 

Ever wondered why some investment ideas are defined as factors and some aren’t? Here’s a checklist for what makes a
true — or robust — factor.

Some cynics have grabbed their share of headlines lately by bemoaning the spread of the “factor zoo” or claiming that some factor strategies are poised to disappear. It’s true that factors get a fair amount of press. A reporter recently called me about an experiment in which she designed a hypothetical strategy based on a cat factor: holding stocks with the word “cat” in the name could apparently generate an impossibly astronomical return! I had warm, furry memories (maybe too much fur!) of our 17-pound, affectionate pet. Henry would always come when called, but we unfortunately had to give him away due to family allergies.

Like our cat, it turns out that the reporter’s approach was not a purr-fect fit. It did, however, highlight in a lighthearted way the proliferation of new factor-based products. So I figured it’s time to talk about how to sort through the noise to identify true factors — those broad and persistently rewarded drivers of returns.

Here’s a checklist to determine what’s a robust factor and what’s not.

1. Has it created value?

There are thousands of investment ideas. But in factor investing, we’re only interested in the ideas that have generated positive risk-adjusted returns over decades.

Academics began identifying factors as drivers of return in the 1930s, giving us decades worth of data showing long-term value creation. In particular, Ben Graham and David Dodd famously published the foundation of value investing, Security Analysis, in 1934. Academics began formally studying momentum — that trends tend to persist—in the 1960s. Today, the handful of equity factors that have delivered risk-adjusted returns higher than the broad market include the names you’d recognize: Value, momentum, quality, size and minimum volatility all have decades of history.1

You may ask, doesn’t that lead us to a world of analysts churning out factor ideas based on data mining and dodgy backtesting? Actually, no. That’s where our second checkbox comes in: economic intuition.

2. Is there an economic rationale?

Strong empirical data is not enough. Factors must be based on both strong economic intuition and academic evidence. Think of this as an economic smell test to see if a good analysis has gone bad: does it make sense that this idea should drive returns?

Here’s my rule of thumb from decades of researching, teaching and managing factor investing in my career: Don’t look at any data without also applying some intuition.

What kind of intuition? There are three main reasons why a factor may work:

  1. Rewarded risk – potentially earning higher long-term returns for taking on more risk.
  2. Structural impediment – leveraging opportunities created by market rules or investor constraints, such as restrictions on pension fund holdings.
  3. Investors’ biases – capitalizing on opportunities created by certain investor behaviors, such as extrapolating trends.

Let’s put this intuition to the test using the value factor. Does it make sense that discounted stocks could outperform pricier peers? Yes. Value stocks have tended to underperform during economic recessions, because those companies typically have factories, production lines and other fixed assets that are not easily redeployed.

Companies with high growth rates may also dazzle some investors into bidding up those shooting stars and overlooking more staid value stocks. When growth stocks flop, those undervalued stocks then outperform. So value is economically intuitive as a factor, based on rewarded economic risk and investors’ behavioral biases.

3. Is it diversifying?

The third checkbox is whether a factor provides something different to traditional market capitalization exposure. Does it add an additional source of returns? A factor that merely replicates an investment tracking the broad stock market, for example, doesn’t offer any diversification.

Ideally, a factor should have low correlation with other factors. As an example, we expect that the value factor should not move in exactly the same direction, by exactly the same amount and at exactly the same time as the momentum factor. This is because they capture different investment ideas.

4. Is it scalable?

The final test is whether the factor can be implemented efficiently and cost-effectively, ideally in a transparent manner. Here’s what I’d look for: reasonable trading costs, a sensible level of turnover and a high degree of capacity. Balancing returns, risks and costs is key to implementing a factor strategy efficiently, such as with smart beta ETFs.

It’s possible that a factor could check the first three boxes and fail this final test: It could be an idea that needs to be captured so quickly that it drives up transaction costs or it might target relatively illiquid securities.

Robust factors check all boxes

To be a broad and persistent driver of returns, a robust factor should check all four boxes on this checklist. It should have created long-term value, be based on economic intuition, add portfolio diversification and be able to be implemented efficiently and at low cost.

That cat strategy didn’t check a single box, and had it ever hit the trading floor, it probably would have been a cat-astrophe.

Andrew Ang
Andrew Ang
Head of Factor Investing Strategies
Andrew Ang, PhD, Managing Director, is Head of Factor Investing Strategies and leads BlackRock’s Factor-Based Strategies Group.