In the previous article, we learned the basics of mutual fund screening and how can we use it to find suitable mutual funds. In this article, we will understand how to build a mutual fund screen by taking a relatable real-life example.
Building your first screen is super simple.
You can start from a template screen and edit the filter/query or start building a brand-new screen by adding your first query. We have highlighted the results in the images in yellow.
Let’s work with an example. Suppose an investor wants to identify large-cap mutual funds that have generated the highest returns with good consistency and minimum drawdown.
Now we will run through the steps required to create this screen.
Some users prefer a granular approach where they can apply the query, evaluate the list, and iterate on the query. This offers more flexibility in query building. In this user guide, we follow this approach.
The starting list will be all the mutual funds available.
As we can see, we have 1175 different fund options (direct plans in growth options of open-ended schemes). This includes all the fund categories like equity funds, debt funds, hybrid funds, etc.
Our first step is to narrow down the universe to funds in the large-cap category. This can be done by selecting the MF category from the basic filter section. From the dropdown, we will select Equity: Large-cap. As index funds that follow broader indices also invest in large-cap stocks, they will be included in this list.
As you can see below, after adding this rule, we have filtered down our selection universe to 96 mutual funds (from available 1175 funds).
Now that we have 96 funds on our list, we can add another filter that will help us identify top performers.
For that, we will use the 3-year alpha as a parameter. We will shortlist funds that are in the top 25% based on 3-year alpha.
Our categorization system makes it easy for users to find the metric they want. Suppose our investor decides to use category percentile rank instead of alpha. She can see a detailed description of category percentile rank including the benefits and risks involved in using the metric.
With alpha as the selected metric, the user can build the query (filter) through an easy-to-use and intuitive query builder.
Users can create an “absolute” comparison query (e.g., alpha less than 1%) or a relative comparison query (alpha is in the top 25 percentile). The latter will pick the top 25% of funds (from the original list) with the highest alpha. In our case, applying this query will filter down the list to 16 funds with the highest alpha.
Note 1: In our screening system, whenever we screen based on percentile, metric values are always sorted in ascending order internally (low to high). So, if you want lower values (as is the case with drawdowns) you should select “bottom x percentile” and if you want to screen higher values (as will be the case with alpha) you should select “top y percentile”.
We have picked 16 funds (top 25 percentile) with the highest alpha. We can now proceed to apply our next filter – funds with good consistency.
We have quarterly consistency and yearly consistency. You can use any one of them. We will use yearly consistency. For quarterly consistency, we use the past 12 quarters (3 years) and for yearly consistency, we use data from the last 5 years. We will select those funds that are in the top 30% based on the yearly consistency.
The fund count after this step is 8.
We need to apply one final filter to our filtered list from Step 3. We need to pick funds with low drawdowns. We apply the same procedure by selecting the drawdown metric and applying the relative (percentage-based) filter. Here we have selected funds that are in the bottom 25% based on the drawdown.
After applying this filter, we have only 2 funds.
That’s it. We have just built our first screen.
Note 2: In our screening system, whenever we screen based on percentile, the sequencing of the rules matters. If you change the sequencing, the results will also change. Below are the results of 2 queries with the same rules but different sequencing.
As you can see, the rules are the same and the sequencing is different. But both screens show different results. This happens because of internal sorting that happens when we use relative (percentile-based) queries.
In the next article, we will look at some examples of building a mutual fund screen.
