Momentum trading, a strategy centered on the idea that securities that have performed well in the past will continue to do so in the future, has gained popularity among traders seeking to capitalize on market trends. In this article, we enlisted the help of Marco Santanche, a quant strategist and author of the research newsletter Quant Evolution, to present a step-by-step guide for a better understanding of the components involved in momentum trading.
Define an Investment Universe
The first crucial step in momentum trading is to define your investment universe. This involves choosing the specific assets or securities you will be trading. The investment universe offers a range of options, with the most common ones including:
Stocks: The entire equity asset class opens up a multitude of possibilities. You can opt for individual stocks such as Nvidia or Microsoft, or concentrate on a group of companies such as FAANG stocks.
Read More: Factor Investing: A Deep Dive
ETFs: Exchange-traded funds offer a convenient option for capturing price movements across entire industries, sectors, or asset classes.
A Basket of ETFs: ETFs can also be strategically combined into a basket, providing a diversified approach to capitalize on market trends. As explained by Santanche, "For instance, we might select only one ETF per asset class (equity, fixed income, alternatives) and proceed by ranking each ETF in the basket based on their momentum, selecting the best-performing one. We then allocate a fixed percentage to the asset classes or conduct a portfolio optimization on the selected ETFs."
Select a Momentum Indicator
The next step is to choose an appropriate momentum indicator, and as Santanche points out, there are two categories to consider:
Time series momentum: This generally means using time series models to identify strong momentum in an asset. It normally involves considering the distribution of returns, employing noise removal techniques and addressing potential jumps.
Read More: Trend Following ETFs: A Deep Dive
Cross sectional momentum: In this approach, the ranking of all assets is considered in relation to the "cross-section" of returns. This allows for the identification of strong or weak momentum. The cross section can be identified in various ways, such as mean return, median return, a specific quantile of an estimated distribution for returns, and so forth.
Resist Finding Optimal Lookback Period in Backtesting
Moving on to the next phase, which is backtesting, it is essential to resist the temptation to identify an optimal lookback period.
"Traders may feel compelled to pinpoint an optimal lookback period during backtesting," cautions Santanche. "However, this would be a mistake. The issue lies in choosing a parameter based on multiple results over time solely due to its performance metrics (such as return, Sharpe Ratio, or others). This approach is widely recognized as problematic and falls under a statistical error known as data dredging. Essentially, it involves running numerous experiments and selecting the best-performing one from the pool of results."
"Nevertheless, the significance of the lookback period cannot be ignored. We need to formulate a hypothesis: why would one month, one day, or one year be the optimal horizon? We can potentially test the effectiveness on a validation set and conduct the backtest on a test set, but never overlap these universes," notes Santanche. "Even better if we produce synthetic data from the train set, or, as suggested by Graham Giller, a prominent Quant, backtest on correlated assets to avoid overfitting."
Building the Winning Portfolio: Selection Criteria
Once the investment universe and momentum indicator are established, the next step is crafting a winning portfolio. Momentum trading relies on identifying assets that exhibit strong upward or downward price trends. Establish clear selection criteria based on the momentum indicator. This systematic approach helps filter out noise and ensures that your portfolio aligns with the underlying momentum strategy.
“We need to select assets, either by filtering out low-to-negative momentum (in case of long-short, negative values can still be taken, with the emphasis on the absolute value) or by creating baskets,” says Santanche. “We might also examine the distribution to identify opportunities for exceptional returns while avoiding those that are merely average.”
Read More: Volatility ETFs: A Practitioner Guide
Alternatively, we can also take a cross-sectional approach, investing only if momentum is positive or trending. Regardless of the method chosen, this step is a fundamental precursor to portfolio construction.
Portfolio Management
As highlighted by Santanche, effective portfolio management is a critical aspect of the momentum trading process. While the signal itself may be effective, optimizing how we capitalize on the identified opportunities is equally vital. This optimization could take the form of a well-balanced portfolio (50% long - 50% short), an opportunistic approach (investing solely in relevant stocks based on their momentum), or a risk parity strategy (where each position contributes equivalently to the overall risk in the portfolio).
Furthermore, tailoring position sizes based on risk tolerance and portfolio objectives is instrumental in maintaining a disciplined and sustainable approach.
Trading Rules
The final and optional step involves defining trading rules. These rules may include setting stop losses, determining take-profit levels, or establishing rebalancing triggers for various reasons (such as event-driven factors like Federal Reserve decisions, performance-driven considerations like trailing stop losses, a combination of both, or none at all). This step ensures we are prepared for exceptional cases where our strategy may be over- or underperforming, or when the portfolio requires an "emergency intervention."
For a comprehensive exploration of quantitative strategy by Marco Santanche, visit Quant Evolution, a monthly research series on quant trading, ETFs and portfolio construction.
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.
