Quintile Performance Analysis: A Plain English Explainer


Posted by Vident Financial on 8/22/22 4:37 PM

We do a lot of research and writing about how various factors such as valuation, inflation, economic growth, and earnings quality relate to investment returns. One of the tools we use to do this is “quintiles.” Quintile analysis is a particular form of bucketing, which simply means to divide your data into various “buckets” from highest to lowest. So, imagine that you take the data for 3,000 companies (which, by the way, is roughly the size of the universe that Vident begins with when constructing its domestic equity index, the Vident Core U.S. Stock IndexTM). Each company has a Gross Margin at a given point in time, Gross Margin is one way of measuring profits. Then you take the 3,000 companies and divide them into 1/5ths. The first bucket is filled with the 600 companies (600 is 1/5th of 3,000) with the highest Gross Margin, the second bucket is filled with the 600 companies with the second highest Gross Margin, and so on.

The "factor," the thing that you think might be the cause of some effect, is used to sort the companies into these buckets. These factors are often described by analysts as X factors. A nice way to remember this is that these factors are the eXplainers, the factors which you think might explain why something happens. Typically, what we are trying to explain is returns on an investment - that's usually called a Y factor, and a nice way to remember that is that it is the reason whY you are doing your analysis in the first place. If you are trying to design indices with higher returns, if that's whY you are doing the work, you will look at various factors (like Gross Margins) to see if they eXplain what happens.

Below you will find a recent example. The bars from left to right are the quintiles of companies sorted in 5 buckets from highest Gross Margins to lowest Gross Margins. The height of the bar is the average return in the year after you sorted by Gross Margins. For the time and data this analysis covered, the 600 highest Gross Margin companies had an average return of 17.21% the next year.

(Source: Bowyer Research, 1-1-1993 - 12-31-2021. Performance is gross of fees as VCUSX does incur any management fees. Returns include the reinvestment of dividends. Past performance is not indicative of future results. You cannot invest directly in an index or any of the shown Gross Margins quintiles.)

Why look at the next year? Because that’s the way things work in the real world. You know the Gross Margins now, not a year from now. You make an allocation decision based on the Gross Margins now in the hopes of getting higher returns in the future. So, in this type of analysis, the Y factor lagged behind the X factor.

Another thing we often do is to average the top 2 and the bottom 2 quintiles. This also fits a likely real-world scenario: Yes, an investor might just buy the companies in the best quintile, but the problem with that is that it might be too limiting to shrink your whole universe of eligible stocks down by 80%. So, in portfolio construction, often what's done instead is to have higher weightings (a bigger proportion of your index) in the best quintile and then somewhat higher (but not quite as high) in the second best, and the lowest weighting in the worst quintile and the second lowest in the second worst quintile. This way you can still have a large number of stocks to choose from to be diversified.

We often create imaginary portfolios this way, with one made up of the 2 best quintiles but double weight the best, and another portfolio made up of the 2 worst quintiles, but with half the weighting in the worst, and then compare how well the 2 imaginary portfolios would have done in the past. We call this a 'performance differential', and if you read our analysis, you've seen that phrase a lot: it's kind of a 'go-to' tool for us.

This approach does not prove anything about the future. A factor can work very well for 10 or 20 years, and quintile analysis can show that, and then someone can use that factor in the real world, and it can underperform for them, because past patterns and correlations and performance differential are simply not a guarantee about the future.



The opinions expressed herein are those of Vident Financial at the time of publication and are subject to change. This document does not constitute advice or a recommendation or offer to sell or a solicitation to deal in any security or financial product. It is provided for information purposes only and on the understanding that the recipient has sufficient knowledge and experience to be able to understand and make their own evaluation of the information described herein, any risks associated therewith and any related legal, tax, accounting or other material considerations. Recipients should not rely on this material in making any future investment decision.

Investors cannot invest directly in an index. Indexes are not managed and do not reflect management fees and transaction costs that are associated with some investments. Past performance does not guarantee future results.