The Dangerous Allure of Designer Risk Factors

Getting loaded on drugs is fraught with dangers.

Sure, you could have the best night of your life. But you might die. Even if you find a drug dealer who delivers discreetly, on time, and at the right price, can you really be sure of the quality?

As most drugs don’t come with a convenient list of ingredients, let’s face it: With the exception of pharmacy grads and chemists, who among us would really know what’s in the stuff? So one is faced with a choice: Trust the dealer or, in the immortal words of Nancy Reagan, “just say no.”

Institutional investors are facing a similar dilemma. The cheap highs promised by multifactor smart-beta strategies appeal to many, but how to assess the ingredients or the cooking methodology?

“Over the past few years, investors have reacquainted themselves with the idea of factors,” says Ana Harris, State Street Global Advisors’ EMEA head of equity portfolio strategists. Investors who want to ride stocks with positive earnings adjustments or good share-price growth can opt for products channeling the momentum factor. Or someone seeking exposure to companies that look cheap relative to their forthcoming earnings, sales, or competitors — but averse to paying an active stock picker — might buy a value strategy.

Harris’s firm built out its risk factor expertise and a product suite relatively early, and reaped in assets while other establishment giants played catch-up. But now everyone has their own brand of factor in a bottle.

The product explosion begat a factor explosion. Cynically, firms realized they needed to differentiate their momentum ETF from 50 others that had sprung up, so they blended a cocktail or mined some data, and stamped the new version “proprietary.” Optimistically, the surge in interest spurred factor research and development, and thus the discovery (versus invention) of persistent return anomalies.

Social sentiment, for example, could foretell company or sector performance from social media mentions, and whether these mentions were in a positive or negative context. In April 2016, Sprott Asset Management and ALPS Advisors launched the BUZZ Social Media Insights ETF, which “seeks to identify ‘social momentum,’ a factor that we believe serves as a leading indicator to future price momentum,” the index’s creator said at the time. The product has since rebranded as BUZZ US Sentiment Leaders ETF and now manages a total of $8.9 million.

Understanding of the underlying principles of individual factors is now at a level where most asset owners seem sufficiently comfortable to deploy at least some of their cash into factor-based investments. According to FTSE Russell’s 2017 survey on the topic, nearly half of asset owners reported some kind of smart-beta allocation, with value and low volatility the most common single-factor strategies.

For the addicts and enthusiasts out there, this is comforting news. Fund managers will happily still sell your favorite fix, and something more exotic to try out too. But beware. There are more dealers on the street and their market share is being challenged.

Fund managers know a price war is brewing. Conscious of losing their turf, providers have broadly embraced the Breaking Bad approach and are focusing on the product. The result is multifactor smart-beta strategies.

Dina Ting, Vice President, senior portfolio manager of the global ETFs group, says the multifactor market is still at an early stage, but notes her firm’s offering will have a “diversification benefit.”

A concern among academics and investors is that as these multifactor strategies become widespread — sold on promises of safer returns or higher returns, or safer, cheaper, and higher returns — their buyers will forge ahead without understanding how fund firms cooked up the recipe, or how to find out.

This is a big issue, according to Felix Goltz, head of applied research at the EDHEC-Risk Institute. He argues that investors should ask for three things: access to the underlying data used during testing, the live performance, and validation of the methodology. “There should be an assessment not just of back-test performance, but of how reliable the back-test is,” he says. “Is it based on factors and methodologies that are validated externally? Is it systematic? Is it consistent?”

In the U.K., the Pensions Regulator has already raised concerns about levels of competence among trustees and whether they deploy experts appropriately to assess portfolios. Nevertheless, some asset owners are already becoming hooked.

Multifactor, multi-asset strategies were the preferred mode of factor investing for 28 percent of investors in the Asia Pacific region, 47 percent of investors in Europe, and 52 percent of investors in North America, an Invesco study found this year. Respondents told researchers that they felt asset consultants were best placed to assess these strategies. Fund managers do typically offer support and education, but investors’ credulity only went so far. Just 13 percent saw their product pushers as appropriate and trustworthy advisers.

For the sake of the sector, argues one manager, providers ought to play fair. Aniket Das worked at investment consultancy Redington before joining Legal & General Investment Management in 2016, specializing in index and factor strategies. “We are trying to promote best habits in back-testing,” he says. “Asset managers should be able to produce some principles on back-testing. When we work with index providers, we want to know what formulations they looked at, how many revisions they have made.”

Das says he asks for these figures and guidelines because it is perfectly possible to run scenarios until you get the result you want — the practice of data mining. “How many methodologies did they come up with before?” Das asks. “Did they run this simulation with 20 different methodologies and just use the one that came out the best?”

Chasing that perfect high — an original factor as good as the first ones you ever tried — never ends with finding it.