Think more than 3,000 ETFs is a lot? Think even bigger because the industry is booming, with another 431 funds coming online last year. That asks a lot of advisors trying to find the signal in the ETF noise, the focus of the ETF U: Sorting Through a Growing ETF Universe panel Sunday at Exchange: An ETF Experience, where panelists including Bloomberg Senior ETF Analyst Eric Balchunas, Factset ETF Research Director Elisabeth Kashner, and CFRA Head of ETF Data and Analytics Aniket Ullal all shared their thoughts.
In such a vast world of ETFs, it’s critical to parse the data into digestible bites – that’s how Factset approaches its ETF work, according to Kashner, describing each ETF with a fit score, a letter grade, and a story on their site with metrics like efficiency, tradability, and fit. She added that misunderstanding and choosing the wrong ETF invites tracking errors, trading spreads, and performance gaps.
“We’ve had years where the difference between the top and bottom ETF in a category is 60%. So getting that right really made a difference,” Kasher said.
Factset allows advisors to compare strategies with an efficiency analysis, as well, in which, for example, the and the can be compared based on data points like median tracking difference, fund closure risk, derivative use, and even average spreads. PXI, for example, has a near 25 basis point average spread, while VHD has just a three basis point spread, making it somewhat cheaper.
To Balchunas, there are five factors in the ETF Due Diligence checklist to consider: exposure, cost, liquidity, risk, and scoring systems. To Balchunas, it’s important to look at ETFs and score them based on risk factors that in some ways reflect movie ratings, from G to rated R, but at Bloomberg is called the ETF Traffic Light System, rating strategies from red, to yellow, to green. He listed the risks contributing to those ratings, including leverage, potential future roll costs, lack of liquidity, and more.
“All the commodity ETFs sound so innocent, but those roll costs will eat you alive. You’ll have made the right bet but made no money,” Balchunas said, comparing at least one strategy to Gremlins, the cute buy monstrous creatures from the 1984 film.
What are some of the advantages of such a scoring system? It covers all ETFs, Balchunas said, adding speed and a recognition that most people don’t actually read prospectuses. Looking at so many different ETFs, the fields in the Bloomberg Intelligence system can empower investors to see measures like purity, in which revenue comes from pure play sources, or ESG, measured at Bloomberg Intelligence with ETFs scored with an E, an S, or a G.
Closing out the talk, Ullal addressed CFRA’s own research process, examining ETFs in a broader context. Considering model portfolios, sector, industry& thematic factors, and market comment, CFRA then looks closely at each given firm.
“What we do is take the entire ETF universe globally and classify it in granular ways,” Ullal said. “Once we’ve done that, we can get into other specific ETF comparisons and ETF level.”
CFRA uses machine learning to help produce the firm’s star ratings, a one to a five-star system that is fed by information like sector performance and expenses versus other ETFs. To build the machine learning model, CFRA analysts gather ratings of the underlying stocks and add forensic trends on top of earnings score and manager track record on top of industry trends. The star model is updated on a monthly basis, with the firm having moderated the ratings, so they don’t fluctuate wildly thanks to a smoothing algorithm.
“Over time, it’s getting smarter; it’s learning through the data we’ve already compiled,” Ullal said. “We really try to make sure the data’s diverse, and it really represents what’s happening in the market.”
For more coverage of the Exchange conference, please visit VettaFi | ETFdb.