Neutral Momentum ETF Woes

The QuantShares U.S. Market Neutral Momentum Fund (MOM) has significantly underperformed its underline index in the last six years. Below are details and a performance comparison to similar strategies based on the S&P 500 stock universe.

A reader of this blog emailed me to ask my opinion about the significant losses of MOM in 2016 to the tune of -18.13%. I thought this is was interesting subject.

This is how MOM is structured, according to QuantShares:

“The Fund seeks performance results that correspond to the price and yield performance, before fees and expenses of the Dow Jones U.S. Thematic Market Neutral Momentum Index. The target index, which is compiled by Dow Jones Indexes, is equal weighted, dollar neutral, sector neutral and is not levered. The index rebalances monthly by identifying the highest momentum stocks as long positions and lowest momentum stocks as short positions, of approximately equal dollar amounts, within each sector.”

Then, according to S&P Dow Jones indices, this is how the underline index, U.S. Thematic Market Neutral Momentum, is structured:

The index is designed to measure the performance of a long/short strategy utilizing long positions in high-momentum companies and short positions in low-momentum companies. Momentum is calculated by ranking stocks by their 12-month historical total return, starting one month prior to reconstitution. The index is calculated using long and short indices as its basis. It is designed to be market- and sector-neutral.”

Below are performance data by year since 2012 for the fund and the index.

Year Fund Index Fund – Index
2012 3.84% 6% -2.16%
2013 3.80% 6.54% -2.74%
2014 -4.63% -2.25% -2.38%
2015 +13.55% +18.15% -4.60%
2016 -18.13% -18.09% -0.04%

Note that total return of the fund since 09/06/2011 is -7.6% as compared to 132% for SPY. My intention here is not to blast the ability of the fund to track the index despite the large divergences shown. Performance of the fund is affected by rebalancing, execution, liquidity constraints, fees, slippage, etc. However, the divergences are large and especially the “collapse” of performance in 2016 that was the main question in the email I received. The fact is that MOM fell 18% in a year that SPY rose 12%.

Some long-short equity market neutral strategies experienced losses in 2016. In my opinion, the main reason for the failure is the low quality of the features (attributes, predictors, etc.) employed for scoring securities and deciding which ones to buy and which ones to short. In this particular case, I believe that scoring based on 12-month rate of change is a naive way of doing things. There is simply no edge in using this and related scores since they are being extensively used.

In a recent article, we described a long/short strategy that used features calculated by our DLPAL software and the S&P 500 universe. Performance for 2016 was close to 7.8% net of fees.

Below are the results from using the same scoring mechanism as in the MOM ETF applied to the S&P 500 universe since 2015. Performance is net of $0.01 commission per share. The strategy holds up to 245 long stocks and up to 245 short stocks at any given time of equal dollar value. The return for 2016 was -5%, much better than the -18% realized by MOM. Click on the images to enlarge.

Hypothetical and real performance can diverge significantly as experience shows. I traded a long/short (but not market neutral) strategy during the 2008 bear market for a hedge fund and I am well aware of that. The performance of an index, and of an active strategy that attempts to track, it can diverge significantly. This answers the question about performance divergences. The main question about the performance collapse in 2016 can be attributed to a combination of a poor scoring mechanism and friction effects. Basically, the trade got crowded and also got more expensive. This tweet from Nashrullah summarizes well what is happening:

Someone uses a 12-month scoring mechanism, then someone else uses 11 months, then someone else 10, and so on, and eventually instability emerges in this simple rotation. Using strategies from public domain poses significant risks. There are still edges in the markets but they are not found in tracking any esoteric indexes and associated ETFs using naive scoring. Success requires a lot of hard work in identifying unexploited edges. Unfortunately, more emphasis is being placed on the fancy programming of worthless edges than on understanding how significant edges can be identified at a minimum data-mining bias. The bright side of this for those who are willing to put the hard work is that there are enough naive market participants to profit from. The markets do no reward anyone based on their degrees, university rank, or programming ability in some fancy language but based on their conviction to go beyond the obvious and into the realms of the unknown.

If you found this article interesting, I invite you follow this blog via any of the methods below.

Subscribe via RSS or Email, or follow us on Twitter

If you have any questions or comments, happy to connect on Twitter: @mikeharrisNY

Charting and backtesting program: Amibroker

Disclaimer

Technical and quantitative analysis of Dow-30 stocks and 30 popular ETFs is included in our Weekly Premium Report. Market signals for longer-term traders are offered by our premium Market Signals service. Mean-reversion signals for short-term SPY traders are provided in our Mean Reversion report.

Copyright Notice

This entry was posted in Quantitative trading and tagged , , , , . Bookmark the permalink.

Leave a Reply