No investment strategy can be constructed without some set of rules for your strategy to adhere to and follow. As part of our Xplore Stock offering, we give our users access to approximately 45 factors that cover a wide spectrum of company financial data and price performance data. These factors allow users to filter their stocks by a strict set of rules and to automate your own investment strategy.

As standard our core Beginner, Established and Xplorer plan users can apply 3 filters. If you want to unlock more filters augment your core plan with Xpert.

We discussed previously the concept of style factors, which are ways to group stocks together that share common underlying properties. These common underlying properties are found by using individual factors. Factors are single pieces of information that can be used to partly represent a style factor. Each of our 45 factors can therefore naturally be grouped into one of the style factors.

For example, value is a style factor and price-to-earnings, price-to-book would be factors. Similarly, quality is a style factor and operation margin, return-on-assets and debt-to-equity would all be the factors associated with the overall quality style factor.

Filters operate very much like a set of criteria that must be obeyed and adhered to and they are an integral part of any investment strategy design. The more filters that are used in the design of your investment strategy, the more complex and constrained the strategy becomes. Using a lot of filters may mean the bar is set too high and as a result not enough stocks may obey all filters.

All filters chosen by a user must be satisfied - for example, if a stock satisfies 2 out of 3 filters then it still will not be included in your strategy as it still failed to pass the test on 1 filter.

We split our large set of factors into different buckets based on overall style factors. By grouping similar factors together this will give users a natural method of choosing their own filters from each style factor.

Our five style factor groupings are

Value

We provide 12 value factors that cover a broad spectrum of what typically represents a value stock. These are split into roughly three segments that focus on

• price based ratios
• enterprise and earnings based value metrics
• sales, cash flow and stock returns

Quality

We allow users to choose from 16 different quality factors that all work to formulate a strategy for quality stocks. As we discussed in our post on the quality style factor, there is no single universal theme that represents a quality company or stock. It really is more of a "the whole is greater than the sum of its parts" philosophy. A well-run company with a strong balance sheet, high quality earnings, robust capital structure represents a quality stock. Our range of 16 factors should cover this wide spectrum.

Dividend

We provide 3 dividend factors within Xplore, which should provide users with plenty of ways to build dividends into your investment strategy design.

Technical

For our technical grouping, we focus on 3 extremely popular and important concepts when it comes to investing, which are

• Momentum: do you like to go with the trend? or are you a natural contrarian? For example, users can use this factor to filter their stock choices to stocks that have recent positive momentum, or if you are more of a bargain-hunter to stocks that have had poor momentum recently.
• Volatility: this will be used by most investors to ensure that they are avoiding the riskiest stocks. One simple way to do this is to use it as a factor and filter out these high-risk stocks.
• Liquidity: we always calculate the total $ volume of each stock through time and then we compare every stock against each other. We then apply a scaling so every stock is given a liquidity score between 0 and 1, with 0 being the least liquid and 1 being the most liquid.

Example of the dividend and technical factors for Xplore users.

Financial

We have added four extra factors that some users may like to use when comparing financial stocks. These cover such items as interest income to total revenue, debt to market cap, interest income and deposits to assets ratio.

There are then three natural comparison criteria to apply to each filter:

• greater than
• less than
• between

Users then have two options for using filters on their chosen factors

1. Enter a fixed specific value and untick the percentile check-box. An example of this would be choosing the P/E factor and saying "less than 10", this will return all stocks that have a P/E less than 10. This method will also work for any % based factors such as dividend yield. Choosing "Greater than 2" and unticking the percentile box will return all stocks that have a dividend yield of greater than 2%.
2. Use the percentile method by ticking the check-box and entering a value. An example of this would be choosing the P/E factor and saying "less than 10". This will return all stocks with a P/E that is less than the 10th percentile of all stocks, which is very different from a fixed value of 10 for P/E.

For advanced users who really have an in-depth knowledge of company fundamental balance sheet items, hardcoded & fixed values like in option 1 above may make sense. For example, an advanced user might suggest they only want to search for companies that are less than a 1.5 P/B, which is a very explicit criteria.

For other users, the second percentile method may be easier to use and requires less knowledge. For example, filtering on the P/E factor using less than the 50th percentile just means that you only want the companies in the bottom 50% (the bottom half of all stocks), based on that P/E factor.

To show the importance of applying just the right amount of filters, here are examples of applying too many or too few.

In this first example we are going to open a new Xplore Stock strategy and choose the defaults of all the market caps and all sectors. Next we are now going to show an example of applying too many filters that are just too strict and no stocks actually satisfy all the filters.

Choosing too many filters, or filters that may be negatively related, mean it may be difficult for any stock to adhere to all of them.

First, we select stocks in the lowest 5th percentile for the price-to-earnings factor. We choose stocks with a price-to-book ratio of less than 1. For the final filter, we select the stocks with an Altman z-score greater than the 50 percentile (indicating good quality companies). When using percentiles we make sure that the check box is ticked. Finally, we apply a simple 6-month momentum ranking to order by strongest recent performance.

An example of choose some percentiles and factors that mean it is difficult for stocks to meet all these filters.

The results found below clearly indicate that the strategy in question employed filters that were just too strict to meet. Although the average universe size, after applying the filters and sorts, was 19 stocks, the strategy may not be able to form a portfolio of 15 stocks, at some point in time.

The strict filters and percentiles were only match for a brief period in 2021, otherwise this strategy was not a viable one as the filter criteria were too strict.

The next example we show is one where choices were made in terms of universe selection that severely impact the algorithm from forming a portfolio. The strategy chooses to only use large cap and only focus on some service sectors.

Forcing your universe and sector choices to be too narrow is risky and may mean your filters, even though they are simple, will mean your strategy design will not be successful.

The results from this portfolio speak for themselves and this is not really a viable strategy to invest in. Even though two simple factors are chosen, the market cap restriction and the sector choices shrink the universe so much that there are not enough stocks to satisfy the two filters.

Another strategy that is not investable given the very narrow universe and sector choices made. There is just a single year around 2017 where a portfolio was formed.

We hope this article helped with your introduction to filters within Xplore Stocks and explained some pitfalls of applying too many of them or with unrealistic parameters. Try it out and let us know how you got on!