Factor Spotlight
Factor University

Momentum - One Size Does Not Fit All (Sectors)

If you want to hedge momentum, just go short Info-Tech and long Energy, correct?

Actually, not so fast...

This strategy may have been somewhat effective in prior markets; however in today’s market, the efficacy of this type of hedging strategy has been rapidly unraveling.

We can see this clearly by looking at the Momentum exposure, using the Wolfe Research QES US Broad risk model of QQQ vs. XLE.


Additionally, the long Info-Tech/ short Energy trade is riddled with other systematic exposures, such as short Value or short Interest Rate Beta, that an investor may not find desirable.


As the charts above show, using sectors to bet on momentum can be a dangerous game to play. With sector plays and growth vs. value themes floating in and out of favor and the underlying macro environment increasingly driving the markets, a naïve momentum hedge may quickly push a portfolio to the wrong side of a market swing.

It’s essential to recognize that momentum transpires differently for different sectors. To illustrate this, we can use the Wolfe Research sector-specific risk models to understand how momentum performs within various sectors. These risk models are estimated within each respective sector and can capture the behavior of factors within a given sector group.


As the chart above shows, the return of the Momentum factor differs substantially based on the sector lens we’re using. From 2007 to the present, Momentum within Health Care had negative premia while the same factor within Energy returned over 200%! Luckily, we can use the sector-specific lens to create momentum hedges with more precision than the blunt objects of sector ETFs.

Design Methodology for Sector-Specific Momentum Hedge

To create sector-specific momentum hedges, we’ll follow a similar methodology as we’ve employed in the past to design factor-targeting baskets.

  1. Build a Tradable Universe
  2. Create Equal-Weighted, Market-Neutral, Factor-Targeting Portfolios
  3. Purify Factor-Targeting Portfolios

1) Build a Tradable Universe

Our starting universe includes all Russell 3000 securities with a market capitalization > $500 million and an ADTV > $5 million. Then, for each sector, we limited down further to only the names within each respective GICS sector classification.


Our filters yielded 6 distinct sector universes:

  • Wolfe US TMT = Information Technology and Communication Services
  • Wolfe US Health Care = Health Care
  • Wolfe US Financials = Financials
  • Wolfe US Consumer = Consumer Discretionary and Consumer Staples
  • Wolfe US Energy = Energy
  • Wolfe US Industrials = Industrials & Materials

2) Create Equal-Weighted, Market-Neutral, Factor-Targeting Portfolios

We created high-minus-low (HML) portfolios for each of the sector universes to get as much exposure to the Momentum factor in each sector model as possible.

The "high" side of the portfolio is an equal-weighted long book using the securities with exposure > 1; the "low" side of the portfolio is an equal-weighted short book using the securities with exposure < -1. We combined the long and short sides to form 6 final market-neutral portfolios targeting the Momentum factor. The final portfolios all had sector-specific Momentum exposure of about 3, on average, throughout the history of 2007-2021.

Below is an example of the exposure for Consumer Discretionary and Staples. We can see that the Momentum factor exposure is the highest, by far, among all of the factors.


3) Purify Factor-Targeting Portfolios

While the Momentum factor exposure is very high in all of our sector portfolios, we can also see in the screenshot above that other factor bets are inherent in the portfolios. To ensure that each of the sector Momentum themes is represented cleanly and do not pick up undesirable factor traits, we optimized each market-neutral portfolio using our SmartTradeTM technology to minimize total risk. We also included trading constraints so that the resulting portfolios are not too concentrated in any individual names. The resulting optimized portfolios form our final 6 sector Momentum-targeting hedges.

Why Create a “Pure” Basket?

To further highlight the importance of the optimization step in constructing the Momentum-targeting hedges, we’ll compare the risk profile of the equal-weighted portfolios from step 2 to the optimized portfolios in step 3. For each sector, we looked at the spread between the total risk of the equal-weighted and optimized portfolios; additionally, we looked at the spread in the percent contribution to total risk from all factors (“% Factor Risk”) as well as the individual Momentum factor (“% Momentum Risk”).

Since our goal with the optimization is to remove unwanted, non-Momentum factor risk, we expect to see lower total risk and a lower portion of overall risk being driven by all factors in the risk model. However, we still want the Momentum factor to be driving the portfolio, so we would also expect to see at least the same, if not a higher, portion of overall risk being driven by the Momentum factor. The table below confirms this exact assertion.

Momentum-Targeting Sector Portfolios - Equal-Weighted vs. Optimized Spreads



Our analysis shows that investors who rely on coarse approaches to Momentum factor hedging, such as simply going long one sector and short another, are likely to run into unwanted bets that can rock their portfolios. By leveraging sector-specific risk models and optimizations to purify the Momentum factor exposures, we can create more dependable baskets that can better meet factor hedging goals.

US & Global Market Summary

US Market: 05/10/21 - 05/14/21

Screen Shot 2021-05-15 at 2.48.13 PM.png
US Stock Market Cumulative Return: 5/10/2021 - 5/14/2021
  • The market rallied over the past couple days as investors shook off the initial shock of Wednesday’s CPI print. For the week, the S&P 500 fell by 1.4%, the Dow slipped 1.1%, and the Nasdaq saw a 2.3% decline.
  • April’s CPI saw its largest increase (+4.2%) since September 2008, eclipsing consensus expectations of +3.6%. Inflation fears rocked the market on Wednesday, but were somewhat assuaged as investors digested the Fed’s assertion that the spike was transitory and driven by base effects given low demand 1 year ago as the pandemic took hold.
  • Commodities were also front and center this week after the Colonial Pipeline (45% of East Coast fuel supply) was shut down by a ransomware attack. Cyclicals like energy and consumer staples gained on the news while growthy tech names suffered.
  • The BLS also reported that the Producer Price Index was up +0.6% from March, and up +6.2% YoY.

Normalized Factor Returns: Axioma US Equity Risk Model (AXUS4-MH)

Screen Shot 2021-05-15 at 2.09.43 PM.png
Methodology for normalized factor returns
  • Value exited Oversold territory with a sizable +0.46 standard deviation move. It continued to bounce off a recent trough of -1.76 SD below the mean on 5/4.
  • Volatility slowly headed back towards the mean after a recent nadir of -2.10 SD below the mean on 4/20. It remains Oversold at -1.31 SD below the mean.
  • Momentum continued its upward move, albeit at a slower pace. The factor remains Overbought at +1.12 Standard deviations above the mean.
  • The decline in Market Sensitivity continued as it headed deeper into Oversold territory at -1.89 SD below the mean.
  • Growth was again the week’s biggest loser, as it shed its Overbought label and landed at +0.53 SD above the mean.
  • US Total Risk (using the Russell 3000 as proxy) increased by 20 basis points this week.

Normalized Factor Returns: Axioma Worldwide Equity Risk Model (AXWW4-MH)

Screen Shot 2021-05-15 at 2.10.47 PM.png
Methodology for normalized factor returns
  • Value saw the largest positive move worldwide, exiting Oversold space after hitting a trough of -1.8 SD below the mean on 4/30.
  • Exchange Rate Sensitivity entered positive normalized territory after a +0.46 SD move.
  • Profitability remains an Extremely Overbought factor at +2.1 SD above the mean after a recent peak of +2.41 SD above the mean on 4/21.
  • Earnings Yield saw another significant decline, heading deeper into negative normalized space after a -0.42 standard deviation move.
  • Growth tumbled out of Overbought space after a -0.5 SD move, and now sits at +0.54 SD above the mean.
  • Global Risk (using the ACWI as proxy) increased by 3 basis points.


Related Insights
See All Insights

What Forces Are Impacting Your Performance? Find Out Now...

Schedule a Call