Advertisements
algos ETF Markets passive investing quant

One Bank’s Quants Think They’ve Found The Key To Outsmarting ETFs

"We conjecture that market participants do not realize that passive investors are buying or selling securities."

The contention that the epochal active-to-passive shift distorts markets and has a deleterious effect on price discovery is of course not new. But as the trend continues, the debate about unintended consequences continues to fester.

Here’s Deutsche Bank describing the current state of the markets in the context of passive ownership:

The shift to passive transmitted through stock-level ownership by passive vehicles is currently at historic levels  Passive investors own $5 trillion or 21% of the market-cap in S&P 500 stocks and $575 Bn or 23% of Russell 2000 stocks. This includes ETFs, mutual funds, and benchmark tracking investment mandates.

Passive

What’s virtually unassailable is the notion that a passive, low-cost S&P 500 index fund is a critical building block when it comes to constructing a portfolio. In the same vein, there is no question that most investors are better served by simply allocating passively using vehicles that charge what amounts to token fees as opposed to trying to outperform by either picking stocks themselves or else picking a manager who then attempts to best benchmarks.

The problem is that past a certain point, “too much” passive creates distortions and the proliferation of “innovations” like smart-beta have helped to create something akin to what Howard Marks has variously characterized as a “perpetual motion machine”. When a handful of names become synonymous with multiple factors used to construct non-index passive products, ever more money gets funneled into those same names. Invariably, some of those names are heavily-weighted at the index level, which means index products which passively track the benchmarks indiscriminately buy those same names. This self-referential dynamic optimizes around itself. That’s the “perpetual motion machine” and the “machine” bit is perhaps more apt than Howard Marks realized. After all, it amounts to less-evolved robots (index passive strategies) interacting with their more sophisticated upgrades (non-index passive products) on the way to pushing markets around.

“The more a stock is held in non-index passive vehicles receiving inflows (ceteris paribus, or everything else being equal), the more likely it is to appreciate relative to one that’s not and stocks like Amazon that are held in a large number of smart-beta funds of a variety of types are likely to appreciate relative to stocks that are held in none or just a few”, Marks wrote, in his latest missive on the subject, before telling you, in plain English, what this means:

For a stock to be added to index or smart-beta funds is an artificial form of increased popularity, and it’s relative popularity that determines the relative prices of stocks in the short run.

Here’s a look at the growth trajectory of smart-beta products:

SmartBeta

Meanwhile, active managers find it harder and harder to outperform. Needless to say, it’s difficult to deliver alpha against benchmarks that only rise and when you throw in the central bank “put” and buybacks, you end up with a scenario where active managers are better off simply not selling. Last year, Wells Fargo described a “sellers’ strike” among clients amid an environment where passive investing was serving as “QE for U.S. stocks“:

Currently, the shift to Passive is coinciding with ever higher equity prices with many equity indices trading at or close to all-time highs. As QE seemed to exaggerate trends in the fixed income markets, so it appears that Passive equity flows are exaggerating stock movements. Recently we’ve observed consistent net inflows to Passive Equity funds, which have morphed into a type of Black Hole. Money goes in and stocks never come out (as least for now).

If active managers sell, they underperform. So they simply stop selling. But by not selling, they join the perpetual motion machine. That only makes that machine stronger and thereby makes it even more impossible for any one active manager to go against the grain and sell. The result: selling has become anathema. It’s a logical impossibility.

This is “just a function of index dynamics, historic flows and the scope of Tech buybacks,” Wells Fargo’s Chris Harvey wrote, in a note dated July 15. “These formulaic market buyers and dynamics won’t last forever but they are too strong to ignore in our view”, he added.

Harvey was making a near-term prognostication in that piece, but the dynamics he’s describing are the same as those described by Marks.

It’s against that backdrop that Deutsche Bank is out with a sweeping new study of ETFs. While there’s a ton of analysis in the piece including an overview of the industry (cited above), the particularly interesting bit comes when the bank’s quants describe a strategy that seeks to benefit from distortions created by passive flows. Here’s the rationale:

We highlight a trading strategy based on the idea that ETF flows affect the price of securities, and prices subsequently reverse afterward. Specifically, we conjecture that market participants do not realize that passive investors are buying or selling securities. Hence, an “information-less” buy order by a passive investor due to flows into an index or an index reconstitution is construed the same way as an “information-led” buy order from an active investor. Thus, market participants overreact to passive flows and eventually learn there is little price discovery associated with passive flows and prices revert.

Ok, so the methodology here is pretty straightforward, although it doesn’t come across that way in the note. Basically, they look at ETFs flows from 2006 to 2017, look at the impact of those flows on stock performance in a given month, and then build a portfolio that’s long the names experiencing the largest outflows and short the names with the largest inflows.

Well, guess what? It turns out that the strategy backtests pretty well. Here’s Deutsche to explain:

Figure 28 reports equal-weighted annualized buy-and-hold returns for decile portfolios formed on ETF flow. Portfolio returns are nearly monotonic going from left-to-right – annualized buy-and-hold returns decrease by 2.5% moving from decile 1 (lowest ETF flow) to decile 9. The highest decile which reflect those stocks that are most purchased by passive investors have abysmal returns that are close to ½ that of other firms in the sample. Figure 29 reports the growth in wealth for the long, short, and long / short portfolios. As we show, $1 invested in those firms with the most negative ETF flow in January 2006 grows to $3.48 at the end of December 2017, compared to $1.40 for those firms with the highest ETF flow. We construct a hedge strategy that goes long stocks with low ETF flow and short stocks with high ETF flow that generates high risk-adjusted performance (Sharpe ratio of 1.28)

DB1

Figures 30 and 31 show the IC decay and turnover of the long/short strategy. The IC decay plot suggests the signal explains return differences for the next five months. While the IC is not particularly high when compared to the long/short portfolio performance, this is not surprising as much of the return difference is due to the substantial underperformance of the highest decile based on ETF flow. The turnover chart indicates a moderate level of turnover (out of 400%), with an average holding period of four months.

DB2

Again, this is a lengthy note and the analysts – led by Alex Chao and Ronnie Shah – pretty clearly went to a good bit of trouble when it comes to being rigorous, so there’s not much point in trying to pick this apart or otherwise poke holes in it unless you’re prepared to buckle down and do it right.

Admittedly, I’m biased in favor of this type of research because I am squarely in the camp that believes it simply isn’t possible for these types of epochal shifts to take place without exploitable anomalies popping up somewhere. It’s just a matter of finding those anomalies. Chao and Shah look like they might have discovered one.

What I would add is that the notion of active investors chasing down the rabbit hole and thereby magnifying moves created by passive flows because they (the active investors) are unable to discern the difference between “information-less” buy orders by passive due to flows into an index or an index reconstitution versus “information-led” buy orders by other active investors, is amusing because it represents, in some ways, a kind of mirror image of the adverse selection problem for HFTs as described by Goldman earlier this year.

Anyway, the overarching point here is that market structure is evolving so fast that we can’t keep up with the myriad implications of that evolution.

For humans, it’s essentially a race against time to identify exploitable, alpha-generating distortions like the anomaly Deutsche Bank claims to have discovered in the note cited above. If we (carbon-based, active managers) lose that race, well then we’ll go extinct.

The ultimate irony in this particular case is that the opportunity identified by Deutsche Bank involves exploiting human error (i.e., the tendency for active managers to amplify moves catalyzed by passive flows, because they can’t differentiate between “information-less” buy orders and fundamentals-based buying).

And because it was discovered by quants, you can bet it will eventually be rolled up into an algorithm that will then be unleashed into the market to capitalize on the distortions created by the interaction of human beings with the “stupider”, less evolved version of itself.

 

Advertisements

2 comments on “One Bank’s Quants Think They’ve Found The Key To Outsmarting ETFs

  1. Error404

    Excellent piece.

  2. Mr. Lucky

    So logically the next thing is to base an ETF on the anomaly? Coming soon to a dealer near you.

Speak On It

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Skip to toolbar