A Guide to Trading Systems

Lesson -> Study of Momentum Portfolio

16.1: The concept of Momentum

I am certain you have been exposed to market jargons if you have worked in the market. Many of us become accustomed to these jargons, and we end up using them without understanding their true meaning. I have been guilty of using a few jargons and not understanding their true meaning. I suspect that many of you who read this will be.

One example of such jargon is momentum. We all have heard the term momentum in our daily discussions about the markets. But what is momentum? And how can it be measured?

When traders are asked about momentum, they loosely refer to the speed with which markets move. While this is true to a certain extent, it's not enough. We should not limit our understanding of momentum to that.

"Momentum" is a term in physics that refers to how much motion an object has. This definition can be viewed in the context stock markets. However, you will need to replace "object" by stocks or the index.

Simply stated, momentum refers to the rate at which the returns of a stock or index change. The momentum is defined as the rate at which returns change in a stock or index. A high rate of return change is considered high. Conversely, a low rate of return change is considered low.

This brings us to the obvious question: What is the rate at which returns change?

Rate of change in return is the rate at which return has changed between two periods. Let's keep this discussion to the end-of-day rate of change in return. In this context, the rate at which the stock's daily return changes is called the rate of change in returns.

This example will help you understand the concept better.

(IMAGe 1)

Below is the table that shows you the closing stock price for a stock over 6 days.Two things are there-

  • Prices are rising on a daily basis
  • On a daily basis, the percentage change is 0.5% to higher

Another example:

There are two things you should be aware of:

  • Prices are rising on a daily basis
  • Daily, the percentage change is 1.5% to 2.5%

Two questions are posed by me regarding the behavior of these stocks.

  • Which stock has the highest rate of daily return?
  • Which stock has greater momentum?

These questions can be answered by looking at the absolute or percentage change in Rupee value.

If you take the absolute Rupee changes into consideration, it is obvious that the Stock A change is greater than the Stock B change. This is not the correct way to view the daily return change. In absolute Rupee terms stock within the 2000-3000 range will always have a greater change than Stock A.

Therefore, absolute Rupee changes will not be sufficient. We need to consider the percentage change. We can see that Stock B's daily percentage change is greater and we can conclude that Stock B enjoys a greater momentum.

Consider this another scenario:
(IMAGe 3).

Stock A has trended upwards on a daily basis while stock B has been a disappointment for most of the time, except for the past two days. Both stocks have produced similar results if you look at the percentage change in the seven-day period. Which stock has the best momentum, given this?

Stock A clearly shows consistency in daily returns and a strong uptrend. Therefore, it can be considered to have continuity when showcasing momentum.

What if I decided to measure momentum differently? What if we looked at the return over a 7-day basis instead of daily returns? Both Stock A and Stock B would be considered momentum stocks if we did that.

My point is that traders tend to view momentum in terms daily returns. However, this is not the only way to see momentum. The momentum strategy that we'll discuss in this chapter actually looks at momentum over a longer timeframe and not just on a daily basis.Later we will discuss this.

You should now have an idea of what momentum is and that it can be measured in terms not only of daily returns, but also over longer time periods. High-frequency traders actually measure momentum on an hourly or minutely basis.

16.2- Strategies of Momentum

The momentum strategy is one of the most common trading strategies. To identify potential opportunities, traders use a variety of methods to measure momentum. All strategies have the same core concept: identify momentum and ride it.

Momentum strategies can also be created on a stock basis. The idea is to compare momentum across all stocks and trade those that have the most momentum. You should note that momentum can go either way, meaning that a trader who follows a single stock momentum strategy will have both long-term and short-term trading opportunities.

Trader also create sector-specific momentum strategies and establish sector-specific trades. This is done by looking at momentum in specific indices to find the sector that exhibits strong momentum. After identifying the sector, look for stocks in the sector that exhibit the strongest momentum.

Portfolios can also apply momentum. This is the idea of creating a portfolio with say "n" stocks, each stock in which to showcase momentum. This strategy is great because it's not just a simple momentum strategy, but also provides safety and diversification.

One strategy we will be discussing is the creation of a stock basket, or a portfolio that consists of 10 momentum stocks. The portfolio is kept until momentum lasts, then it is re-balanced.

16.3- Understanding Momentum Portfolio

Before we get into the strategy, let me remind you of a few points.

  • This is a brief overview of how to set up a momentum portfolio. This is not the only method to create a momentum portfolio.
  • To implement this strategy, or any other momentum strategy, you will need to be able to program. You don't have to be a programmer like me. Find a person who can provide some help to you.
  • This strategy must be tested as with any other.

Here's a guideline for building a "Momentum Portfolio" based on the above.

Step 1 Define your stock universe

There are approximately 4000 stocks listed on BSE and 1800 on NSE, as you may be aware. TCS, a highly valued company, and all Z-category stocks on BSE are included. These companies are at the extreme end of the spectrum. Do you need to follow all these stocks in order to build a momentum portfolio of stocks?

It's not possible, it would be a waste.

You need to filter out stocks and create a 'tracking universe. We will select stocks from the tracking universe to make our momentum portfolio. The momentum portfolio will be an ever-growing subset of this tracking universe.

The tracking universe is a collection or malls you love. You might find yourself going to 2-3 malls every day out of 100 malls in your area. Your entire wardrobe can be purchased from these 2-3 malls (see portfolio). These malls are the best place to start tracking your 100-plus options in your area.

It can be either the Nifty 50 stocks, or the BSE 500 stock. The momentum portfolio will always consist of a subset either the Nifty 50 stocks or the BSE 500 stock. It is a good idea to keep the BSE 500 stocks in your tracking universe. However, if you are feeling adventurous, you can create your own tracking universe.

You can create custom creation on any parameter. For example, I could filter the 1800 stocks on NSE to remove stocks with a market capital of at least 1000Crs. This filter will reduce the list to a smaller and more manageable size. You may also add additional criteria, such as that the stock price should not exceed 2000. And so on.

This is just a random list of filter ideas. The custom creation tools allow you to filter out the noise and create a tracking universe that meets your requirements.

Finally, based on my experience, you should have between 150-200 stocks in the tracking universe if your goal is to create a momentum portfolio with 12-15 stocks.

Step 2 Setup the data

Once your tracking universe has been set up, you can now move on to the 2 step. This step will require you to obtain the closing prices for all stocks in your tracking universe. You must ensure that your data is accurate and updated for corporate actions such as splits, bonus issues, special dividends and other corporate actions. A trading strategy is only possible with clean data. You can find a variety of data sources, including the NSE/BSE websites, where you can get the data for free.

What is the lookback period? What number of historical data points is required? This strategy requires only 1-year of historical data points. If, for example, I have 2 nd February 2019, then I would need data points from 1 to stMarch 2018 to 2 _ndMarch 2019.

You can also update the last 1-year data set on a daily basis. This means that the closing prices for the day are recorded.

Step 3 Calculate returns

This step is crucial to the strategy. We calculate the returns for all stocks in the tracking universe. You may already know that we calculate the return in order to determine the momentum of each stock.

As we have discussed in the chapter, it is possible to calculate the returns at any frequency. This could be daily, weekly, monthly or even annually. For the sake of this discussion we will keep the annual returns in mind. However, you are free to add your own spin to the strategy and calculate the returns at any time frequency that you choose. You could calculate half-yearly, monthly or fortnightly returns instead of yearly.

At this point, you will need a tracking universe that includes between 150 and 200 stocks. Each stock should have historical data that spans at least one year. You will also need to calculate the annual return for each stock in your tracking universe.

This sample universe contains just 10 stocks to help you better understand the concept.
(IMAGe 4).
The tracking universe includes the data for the past 365 days. Also, the 1-year returns can be calculated.
(IMAGe 5).

It is easy to see how returns are calculated if you look at ABB.

Return = [ending value/starting value]-1

= [1244.55/1435.55]-1


It's quite simple, I think.

Step 4: Rank the returns

After the returns have been calculated, it is necessary to rank them from highest to lowest. Asian paints, for example, have a return rate of 25.87% which is the highest on the list. The rank of Asian paints is therefore 1. HDFC Bank is second, and will be awarded the 2 rank. Infosys's rate of return is -35.98%. On the list this one is the lowest rank. And so on.

This is the portfolio's'return ranking.
(IMAGe 7).

If you're wondering why most stocks have negative returns, it's because stocks react to deep corrections. I wish I could have chosen to discuss this strategy from a better place.

What does this ranking tell you?

The ranking is a reordering of our tracking universe that gives us a list stock from the highest to the lowest return stock. This list shows that Asian Paints is the top performer in terms of returns over the past 12 months. Infy is the worst.

Step 5: Create your portfolio

A typical tracking universe would contain between 150 and 200 stocks. We can reorder the tracking universe using the information from the previous step. We are now ready to build a momentum portfolio with the reorganized tracking universe.

Remember that momentum is the rate at which the return of an investment changes. This return is measured annually.

A good momentum portfolio contains about 10-12 stocks. Personally, I am comfortable with a portfolio that contains up to 15 stocks. Let's assume we are creating a 12-stock momentum portfolio.

The momentum portfolio consists of the top 12 stocks within the reordered monitoring universe. We buy all stocks from rank 1 through 12. As an example, suppose I wanted to create a 5-stock momentum portfolio. It would have -

  • Asian Paints
  • HDFC Bank
  • Biocon
  • ACC
  • Ultratech

The portfolio would not include the rest of the stocks, but they will remain in the tracking universe.

You may be wondering what the logic is behind selecting this subset within the tracking universe.

This is important to understand: If the stock has performed well over the past 12 months in terms of return generated, it means that the stock has good momentum over the specified time period. This momentum is expected to continue into the 13 th month, which will mean that the stock will continue to earn higher returns. If you buy such stocks, you will benefit from the expected momentum.

This is clearly a claim. Although I don't have any data backing this claim, I have used the exact same technique for several years with good success. This strategy is very easy to test back, and I recommend it.

In the old days, my trading partner (and I) were encouraged to create this momentum portfolio by reading this Economist article. Before you implement this strategy, it is important to read the article.

The idea is to purchase all momentum stocks in equal amounts once the portfolio stocks have been identified. When the capital is available at Rs.200,000/– and there are 12 stocks within the portfolio, it is possible to buy Rs.16666/– of each stock (200,000/12). This will create a momentum portfolio that is equally weighted. You can adjust the weights to create an unbalanced portfolio. This is fine. However, you must have a good reason to do so. Backtested results should be your reason.

Here are some ideas for experimenting with skewed portfolios.

  • 50% capital allocation for the top five momentum stocks (ranks 1-5), and 50% for the remaining 7 stocks
  • The top 3 stocks receive 40%, while the rest of 9 stocks get 60%
  • If you're a contrarian who expects lower-rank stocks to perform better that the higher-rank stocks, then you should allocate more to the last 5 stocks

And so on. The ideal approach to capital allocation should be based on your backtesting process. This also means that you will need to backtest different capital allocation methods to determine which one works best for you.

Step 6 - Balance the portfolio

We have so far created a tracking universe and calculated 12-month returns. Then we ranked stocks in terms the 12-month returns. Finally, we created a momentum portfolio through the purchase of top 12 stocks. The 12-month performance was used to build the momentum portfolio. We hope it will continue to show the same results for the 13 months.

These assumptions are not very common.

  • The portfolio is created on the first trading day in the month.
  • This means that all number crunching takes place on the last day in a month, after the market closes.
  • After the portfolio has been created and purchased, you can keep the stocks until the end of the month


The question arises here that what happens at the end?

You can then re-run your ranking engine at the end of each month to find the top 10 or 12, which have performed well in the past 12 months. Please note that we always consider the most recent 12 months of data.

We now buy stocks from ranks 1-12, just as we did last month. My experience shows that only a small number of stocks in the original portfolio would have changed their positions. Based on the list, you will sell stocks that no longer belong in your portfolio and purchase the new stocks that have been included in the momentum portfolio. You essentially rebalance your portfolio at the end each month.

And so on.

16.4 - Variations in Momentum Portfolio

Before closing this chapter and this module, I would like to mention a few variations of this strategy.

The portfolio is held for one month and the returns are calculated on a 12-month basis. You don't need to adhere to this.Various options are available such as-

  • Calculate the return, rank the stocks according to their monthly performance, and keep the portfolio for the entire month
  • Calculate the return, rank the stocks according to fortnightly performance, and keep the portfolio for 15 consecutive days
  • You can rank on a weekly basis, and then hold it for one week
  • Calculate daily and create an intraday momentum portfolio

You can see that there are many options and your imagination is only limited. The momentum portfolio is price-based, as we've seen. You can also build a fundamental-based momentum strategy. Here are some ideas:

  • Create a tracking universe of fundamentally sound stocks
  • Notice the difference in quarterly sales (% wise).
  • The quarterly sales of the stock companies are used to rank them. The company with the greatest sales jump is ranked one.
  • The top 10-12 stocks to buy
  • Rebalancing at the end the quarter

This can be done on any fundamental parameter, such as EPS growth, profit margin or EBITDA margin. These strategies are very attractive because the data is readily available so backtesting is much easier.

16.5 Warning

The price-based momentum strategy is only effective when the market trend is up. The momentum strategy fails when markets become turbulent. When markets fall, the momentum portfolio is more vulnerable than the market.

It is crucial that you understand the strategy's response to market cycles in order to make this portfolio a success. It was hard for me to learn it. This strategy worked well in 2009 and 2010, but it was not as successful in 2011. Do your research (backtesting) before you implement this strategy.

Let me assure you, after all that, a price-based momentum strategy can provide you with great returns if it is implemented in the right cycle.



  • Momentum can be defined as the rate at which there is a change in return.Over any time period it can be calculated.
  • A price-based momentum portfolio is made up of stocks that have shown the most momentum over the specified time period.
  • It is important to carefully populate the tracking universe. BSE 500 is an excellent tracking universe
  • Calculate the returns on the tracking universe
  • List the stocks according to their highest-to-lowest return
  • The top 12 to 15 stocks are the momentum portfolio.
  • It is expected that momentum will continue throughout the holding period
  • Backtesting equal weighted portfolios can make it difficult to determine the best asset allocation strategy.
  • Momentum can also be measured using fundamental data - sales growth, EBITDA margins and net profit margin, etc.
  • Price-based momentum is most effective in an upward trending market, and not in a sideways market or down market.