This is Part 5 of our series on position sizing. The rest of the series may be found here – PART 1, PART 2, PART 3, PART 4, Part 6, Part 7.

As Sean has discussed in Part 1, Burton Malkiel provided guidelines for how many stocks to own, and Modern Portfolio Theory provides a framework for portfolio construction under an assumption that markets are efficient, and risk is defined as stock price volatility. In contrast, the field of active equity portfolio management has not developed any widely-agreed upon approach to position sizing. In Part 4 we showed how we’re able to incorporate qualitative judgements into our quantitative portfolio construction process. We know that human judgement is required to make long-term forecasts about companies and yet we know that decision-making and forecasting is rife with cognitive biases.

One solution is to equal-weight all holdings, but we believe a better system is one that combines qualitative and quantitative factors to weight positions based on a combination of their return potential and the likelihood that a positive outcome occurs. This system converts our qualitative judgments into an algorithmic position-sizing framework. We’ve crafted a process that we believe retains most of the value of qualitative assessment while stripping out as much of the biases as possible.

This system is made up of our conviction (as discussed here and here) and our assessment of the company’s discount to fair value. In isolation, each of these components only tell part of the story. For example, say you have two companies that are available for investment and they come out to have the same discount to fair value. If you only looked at discount to fair value, you might weight them the same. But, if you had a lot more faith in one company’s management, moat, or your ability to forecast the company’s results, you’d want to weight one more than the other.

The merging of these factors creates a matrix with various targets throughout each intersection. Here is a sample illustration of this matrix:

There is also a third dimension to this weighting scheme – the research stage (as discussed here). This measures the extent of research we have completed on the company. Since we’re still learning about the company in the earlier stages, we apply an adjustment to the weighting scheme to take into consideration that there’s greater uncertainty around our conviction and discount to fair value related to information we may not yet know that you don’t know.

This creates a portfolio that is tilted toward our highest conviction and highest discount to fair value names. This strategy of “putting your money where your mouth is,” can pay off. Through some great work at AlphaTheory, they’ve shown that managers that apply a disciplined weighting process to their portfolio tend to more heavily weight securities that outperform. According to their research, “Top 5 positions went up an average of 12.1% while the portfolio as a whole went up 8.4% (for shorts, went down 8.4%).  That’s 50% better!”

We’re not advocating for a 5-stock portfolio. The prudent investor remembers that the first rule in managing money is to limit permanent loss of capital. Investors do not have perfect insight into what will happen with a company and its stock price, so holding a portfolio of 20-25 stocks helps mitigate that risk. We discussed this in more detail here.

And, while the framework is algorithmic, that doesn’t mean we blindly follow it. We monitor our model position sizes constantly and regularly update each company’s ratings. From time to time, our intuition conflicts with the output of our model. When this occurs, we dig deep into what is happening and why. Sometimes, for example, while our intuition might be telling us we shouldn’t be trimming a stock that the model is calling for, after further assessment we recognize that we’re just suffering from a bias of wanting to hold a “winning” stock and so we go with the model’s recommendation and trim the position. Occasionally, we recognize that our intuition is correctly recognizing that some aspect of our rating has gone stale and we update our conviction. This continual examination of our process can lead us to slightly edit our model from time to time. We know that our model is not perfect and so we’re open to refining it if the evidence is strong that we should make a change.

In the next post in this series we’ll review our process for letting stocks drift from target and how we decide when to trade between companies in the portfolio.

Please click here for Part 4 or here for Part 6

Please read important disclosures HERE.

For more information about positions owned by Ensemble Capital on behalf of clients as well as additional disclosure information related to this post, please CLICK HERE.

While we do not accept public comments on this blog for compliance reasons, we encourage readers to contact us with their thoughts.

Past performance is no guarantee of future results. All investments in securities carry risks, including the risk of losing one’s entire investment. The opinions expressed within this blog post are as of the date of publication and are provided for informational purposes only. Content will not be updated after publication and should not be considered current after the publication date. All opinions are subject to change without notice and due to changes in the market or economic conditions may not necessarily come to pass. Nothing contained herein should be construed as a comprehensive statement of the matters discussed, considered investment, financial, legal, or tax advice, or a recommendation to buy or sell any securities, and no investment decision should be made based solely on any information provided herein. Links to third party content are included for convenience only, we do not endorse, sponsor, or recommend any of the third parties or their websites and do not guarantee the adequacy of information contained within their websites. Please follow the link above for additional disclosure information.