Can relative momentum be used to beat the market? Here’s my 5-year experience with a simple ETF rotation strategy.

I’ve been active on the stock market for more than 20 years. In the first 15, I was mostly underperforming, trying all sorts of strategies for stock picking.

After endless learning and reading, in early 2021 I finalized a simple rules-based system **based on relative momentum**. It’s a rotation strategy, where every month it selects the 2 best performing ETFs from a predefined list of **15 wide sector- and factor-based ETFs** (no theme-based or narrow ETFs, no shorting, no leverage). The results have been amazing, to be honest…

**The core logic:**

* **Momentum is persistent:** Winners tend to keep winning in the medium term.
* **Low Correlation:** By rotating between different sectors and factors, you reduce the impact of a crash in one specific area.
* **Diversification:** By holding ETFs (no individual stocks) and splitting between 1 sector-focused and 1 factor-focused, you get smoother returns.
* **Zero Discretion:** The rules dictate the trade. No gut feelings or emotion.

I did a full backtest in 2021 going back to 2000. This showed an average return of \~16% with smaller drawdowns than the market. That of course made me skeptical, as it shouldn’t be possible according to most economic theory.

So I spent a long time trying to “break” this backtest to find an error. There’s no look-ahead bias, as it doesn’t have any future information available for each monthly decision. There should be no overfitting either, as **the only input to the system are the monthly historical prices** of the 15 ETFs (or rather indices, but there are ETFs available that track them).

I started out investing small amounts using this approach in 2021. As the results kept surprising me and **outperforming the market**, I gradually invested more, and in the past couple of years I’ve had 70-80% of my money invested this way.

Here are my results from the last **5 years of actively trading** this strategy (fees and taxes not included) compared to the MSCI World Index (in EUR):

|YEAR|STRATEGY|MSCI WORLD|DIFFERENCE|
|:-|:-|:-|:-|
|**2021**|38.03%|29.26%|8.77%|
|**2022**|10.16%|\-14.19%|24.35%|
|**2023**|24.54%|17.64%|6.89%|
|**2024**|33.14%|24.81%|8.33%|
|**2025**|11.31%|5.35%|5.96%|

I have a similar table with the full backtested and real results from 2000-2025, which shows a very consistent alpha (outperformance) compared to the market almost every year.

I should say that all the numbers I listed are measured in Euro (I live in Denmark) and without fees or taxes included. These may affect the results for people in other countries like the US.

**I’m curious to hear your input on this strategy**. Theoretically, this should not be possible. Do you think I’m missing something here? No strategy is perfect of course. Does anyone follow a similar approach? Or is this a strategy you would consider?

(I also have a full article with the details of how the strategy works and how it can be copied, including performance data and the full backtest, if anyone is curious.)