Entering text into the input field will update the search result below

Momentum Strategies With Leveraged Funds

Mar. 08, 2021 7:25 PM ETDBB, IEF, MDY, MIDU, MVV, QLD, QQQ, SPXL, SSO, TLT, TQQQ, UUP, XLP, XLY, SPY59 Comments


  • Leveraged funds may be successfully included in portfolios managed with momentum strategies.
  • Leveraged funds may generate great returns with moderate risk if used during market risk-on periods.
  • We compare the performance of three portfolios, two with leveraged funds, the third with the corresponding non-leveraged funds.

Momentum Strategy

In this article we apply a modified version of our dual-momentum strategy to a portfolio made up of three ETFs based on the following US market indexes: S&P500, S&P 400 and Nasdaq 100. We investigate the basic portfolio based on the stocks in the indexes without leverage, as well as portfolios with 2X and 3X leverage. The list of ETFs is given below.

  • SPDR S&P 500 ETF Trust (SPY)
  • ProShares Ultra S&P 500 (SSO)
  • Direxion Daily S&P 500 Bull 3X Shares (SPXL)
  • Invesco QQQ Trust (QQQ)
  • ProShares Ultra QQQ (QLD)
  • ProShares UltraPro QQQ (TQQQ)
  • SPDR S&P MIDCAP 400 ETF Trust (MDY)
  • ProShares Ultra MidCap 400 (MVV)
  • Direxion Daily Mid Cap Bull 3X Shares (MIDU)

We shall use a dual-momentum strategy with two regimes: risk-on and risk-off. During risk-on periods, the strategy allocates to the top two funds based on their relative strength over the previous three months. During risk-off periods, the strategy invests all the funds in one or two Treasury bonds funds.

  • iShares 7-10 Year Treasury Bond ETF (IEF)
  • iShares 20+ Year Treasury Bond ETF (TLT)

A topic that received much attention in this work was the choice of factors associated with the transitions between risk-on and risk-off states of the equity markets. Here is the list of funds used to generate signals for market state:

  • Invesco DB Base Metals Fund (DBB)
  • Invesco DB US Dollar Index Fund (UUP)
  • Consumer Discretionary Select Sector SPDR Fund (XLY)
  • Consumer Staples Select Sector SPDR Fund (XLP)

In our latest article we used the pair DBB-UUP as an absolute momentum indicator. In that application the return of DBB was compared with the return of the US dollar. In this application we add a new pair: XLY-XLP. The condition that must be satisfied for switching to risk-off allocation is that both pairs indicate

This article was written by

Toma Hentea profile picture

Toma Hentea, PhD. is a University Professor Emeritus with extensive research experience in developing quantitative system models with application to financial markets. His market exposure is varied in accordance to the state of leading US and global economic indicators. Options and leverage are used to enhance the performance of his investments.

Toma leads the investing group Adaptive Momentum Investing, where he applies computational algorithms to evaluate the market state (risk on/off). He uses adaptive momentum strategies to construct robust stock and ETF quant portfolios, aiming at achieving double digit annual returns. Learn more.

Analyst’s Disclosure: I am/we are long SPY, MDY. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Seeking Alpha's Disclosure: Past performance is no guarantee of future results. No recommendation or advice is being given as to whether any investment is suitable for a particular investor. Any views or opinions expressed above may not reflect those of Seeking Alpha as a whole. Seeking Alpha is not a licensed securities dealer, broker or US investment adviser or investment bank. Our analysts are third party authors that include both professional investors and individual investors who may not be licensed or certified by any institute or regulatory body.

Recommended For You

Comments (59)

@Toma Hentea if i look back with 62 days range using DBB,UUP,XLY,XLP, the flag is off now.
Toma Hentea profile picture
@zarkavin You are right. The conditions for risk-off are marginally met with an evaluation period of 62 days. They are not met for 77 days my algo uses now. Besides, all other factors indicate a risk-on state.
10 Apr. 2021
@drftr @Market Trends Investor

I’m using the 3x ETFs of SPY,DJI, NXD and TLT. Than using dual momentum of 3 month. Right away, I’m getting 2010 until today around 60% CAGR. At a first step I would like to improve CAGR and then later deal with DD. What would be the max CAGR one could roughly get/shot at before tackling DD ?

Not sure if you aware, but one quantconnect a group of people is working on a similar algo. See “the In & out strategy”, using different relations to switch on and of, but this might be a bit over fitted and not sure if the relations they use are holding.
Very hard to say because when starting with 2010 from the top of my head you would already have missed 3 excellent years IF you had the timing right (and adding inverse ETFs). Starting from 2000 I ended up a bit over 90% before starting to tone it down to currently 58%. So maybe 75-80%?

I think we shouldn't "pollute" Toma's thread further though as this discussion probably is not relevant to most readers, but feel free to PM me all the same.

10 Apr. 2021
I did two regressions:
1) SPY rolling weekly percentage change vs VIX absolute values.
Found not really good thresholds
2) SPY rolling weekly percentage change vs VIX weekly percentage change.
This gives really a good regression and I can find easily 2 levels to decide on 1x, 2x, 3x leverage.
Could you please comment on 1) the absolute VIX value. Does not make sense to me to decide on fixed VIX values. I probably did not got your point.
One other point, did you check the VIX term structure? (Vix divide by its 3month future) It looks as well promising.
littletiger101 profile picture
@Loger I had similar finding as yours. asked the same question to drftr and got no answer
@Toma Hentea thx for another great angle using dual momentum model. I was trying to follow this model manually or using some simple python backtesting tool too. One thing I noticed is that, based on the new signal model, the check is done on daily basis, so the risk on/off flag could flip quite frequently. e.g. during 2017-2018, the flag could cycle on/off during the same week. Is this also you observation? If it is case, I am wonder if this flag is too sensitive to tell a real trend? thx.
Toma Hentea profile picture
@zarkavin All indicators that are computed daily, may cross a threshold multiple times. One possible solution is to compute it only weekly or monthly. Another solution that I use is to impose a delay to the switching process. The switching from risk-off to risk-on is done the next day at market open, but the switching from risk-off to risk on is done only after the risk-off was confirmed for two days, so it is done only the third trading day at open. That reduces drastically the jitter and the system switches only about five times annually on average. The last time it had a switching signal was in May 2020.
@Toma Hentea thx for clarification, 5 times switch sounds a lot better. I will try to replay based on ur inputs.
Toma, have you modelled how sensitive this approach is to the day in the month you select to do the rebalancing (I assume you are using the first day of the month)? I saw another study on Dual Momentum that showed an alarming variation in the outcome dependent on what day of the month re-balancing took place. In fact the first of the month produced the best outcome, with the variation from top performer to bottom perfomer being quite spread out. Ideally you would not want a strategy like this one being so reliant on the day of the month changes are made.
I always wondered how that could be UNLESS you move from RiskOn to RiskOff or vice versa.

When markets are falling, say -2%, you sell your current holding for a -2% loss while you buy the new holding for a -2% discount. And in rising markets you sell your current holding with a gain of a certain percentage while you buy the one for a higher price of the same magnitude. In short: It doesn't matter at all on what day or even what time you rebalance as long as you can immediately switch from one asset to the other, assuming that your chance of always selling for a worse percentage than you buy back for is about 50%. One day you lose a bit, the other day you have luck on your side. Maybe investors that can prove they always lose money because of this should be short? ;-)

I'd be interested in reading a scientific reason for something that in my uneducated mind shouldn't exist, although I'm very aware of the research that's been out for decades.

Thanks for your feedback. I will reflect on your points. In the meantime I found the study I was referring to: allocatesmartly.com/...
Toma Hentea profile picture
@Steve2222 The trade may happen any day of the month, without any restrictions. This is important to avoid large corrections.
Sorry if this was in the article, but how often are you evaluating DBB-UUP and XLY-XLP? Is it every day, every week, every month, or every quarter?Very interesting article!
Thank you for these articles explaining you very useful methodologies. One question: regarding this method and the recent one involving Dow stocks, do you have available the data on how much time was spent risk-on (in the market) versus risk-off during the backtest period.
Keep up the very good work.
Toma Hentea profile picture
In the article involving Dow stocks the strategy was in risk-on 61% of the time. In the current article with three broad market ETFs the strategy was in risk-on for 43% of the time.
@Toma Hentea Thank you. So on average the two methods are in the market only about half the time - makes the returns even more impressive.
DynamiteDen profile picture
Although I'm still taking advantage of this bull market, I keep having this awful feeling we are in for a crash. So I was wondering what would happen if you were to use inverse ETF's, or even a model where you used both 2x Bull and 2x Bear ETF's, in which case what might your model risk-off trigger & investments be :)
Toma Hentea profile picture
Your questions are excellent, but I have not done any research on combining bull and bear leveraged funds. My reasoning is that the equity markets are asymmetric, much longer time in uptrends than in downtrends. I prefer going to short term bonds or cash when in risk-off. Using inverse ETFs is harder to manage.
I'm using 20% inverse equities combined with 80% treasuries during RiskOff and echo Toma's comment as inverse equities need very good timing to avoid losing out twice at a turnaround:

Not only would you miss being long equities but it bites you in the back by adding a loss of a similar percentage. That hurts twice. And both for your strategies' results as in your head because you were double wrong. The benefit that bonds have over inverse equities is that they CAN (and often DO) go up together with equities, while inverse equities never have that possibility by design when the market is doing the opposite of what you expected.

littletiger101 profile picture
@drftr thank you for sharing. You have described your strategy is to move asset from leveraged to non-leveraged. can you explain when is risk-on or risk-off based on VIX? TIA
Curious if anyone has looked at sentiment indicators to assess risk on vs risk off conditions? Seems like sentiment, and maybe some sort of 'smart money' flow indicator would be good to include to assess conditions as well. Great analysis!
Toma Hentea profile picture
I check some sentiment indicators and have tried including them in simulations. The improvements I get are not significant. I need to do more research in that direction.
ScottHB profile picture
@Toma Hentea
Momentum is far and away the best sentiment indicator ever identified. Further research is virtually certain to prove fruitless.
Here is a simpler idea with the SPXL that works. Dollar cost average by purchasing a set dollar amount each week. I downloaded all the SPXL daily prices since inception. Over any 5 year period, it outperforms the SPY by a lot.
Market Trends Investor profile picture
Are the equity ETFs equal weighted on buy signals? I would suggest substituting the S&P 500 ETF proxie allocations with the Dow 30 proxies. The Dow is less correlated with the Nasdaq 100 ETF proxies. Try DIA (1x), DDM (2x) and UDOW (3x). You may be pleasantly surprised at the results.
Toma Hentea profile picture
@Market Trends Investor
During risk-on periods the strategy is invested equally in the top two funds.
I did try with DOW instead of SPY. The performance was slightly worse with annual gains lower by 1% to 2%.
Market Trends Investor profile picture
@Toma Hentea Were drawdowns less? Higher Sharpe?
Toma Hentea profile picture
@Market Trends Investor
Drawdowns are the same, Sharpe is higher with SPY.
Augustus profile picture
Thank you for the valuable article with your research.
Villi Grdovich profile picture
Any momentum portfolio goes through the floor in a market downturn, and i would think that a leveraged momentum portfolio is an even tougher proposition in any downturn.

So identifying a risk off with almost no error in real time is essential. I note the comments below and am quite keen to independently investigate how well that works because if it does, i would regard that as a breakthrough in momentum portfolio management.
@Villi Grdovich
Try a weekly rebalance if you do so, Villi. This to reduce the chance you're either under- or over-trading. Going daily may capture far too much whipsaw, while going monthly makes your strategy way too slow to respond to nasty downturns.

SBSisHim profile picture
Thanks for this Toma. Any chance of using proxy mutual funds for the signal assets to extend these backtests back further than 2008. I am interested how this strategy performed in the dot.com meltdown and the ranging markets of 2005, 2006.
Toma Hentea profile picture
@Frank Lind
Unfortunately, the system I use for backtesting does not have access to mutual funds historical data. I consider using a commodity index and the US dollar exchange rates instead. Will write a new comment when I get some results.
Nice one, Toma...

For future exploration:

A trick I use in my own leveraged strategy is using both absolute and relative VIX numbers (compared weekly) to add or reduce leverage of a normally double leveraged portfolio. Works like a charm, is probably good enough to increase by 50-ish percent but the biggest advantage is that it leads to an enormous reduction of maximum drawdowns, making risk-adjusted returns explode to the upside.

@drftr so if spot VIX is higher week-over-week and also above a certain cutoff you reduce leverage?
I would suggest using either/or as some absolute values can be very low while the relative number has advanced way too much to be called healthy.

At the other end of the spectrum this means you can use relative change only to decide to move to triple leverage when volatility decreases, say 10 or 15%, as that tells you a lot about where investors the market will be going. Obviously they allocate money accordingly; no matter whether the absolute VIX is at 60 or 20.

@drftr Really interesting idea - my interpretation is that you are looking at both the VIX level, and the VIX trend. So if the level is high, but falling, you would keep leverage minimal. If the level is low and falling, you would increase leverage. And if the level is low, but rising, you would decrease leverage. Is this generally accurate?
Disagree with this article? Submit your own. To report a factual error in this article, . Your feedback matters to us!

Related Stocks

SymbolLast Price% Chg
Invesco DB Base Metals Fund ETF
iShares 7-10 Year Treasury Bond ETF
Direxion Daily Mid Cap Bull 3X Shares ETF
ProShares Ultra MidCap400 ETF

Related Analysis

To ensure this doesn’t happen in the future, please enable Javascript and cookies in your browser.
Is this happening to you frequently? Please report it on our feedback forum.
If you have an ad-blocker enabled you may be blocked from proceeding. Please disable your ad-blocker and refresh.