Smart security selection is a great friend of diversification. Another great friend of diversification is value. Not only has value shown to be a material and consistent performance catalyst for centuries, it makes sense. When you go to the store you want to get value for your dollar. It’s the same in the global store of stocks, you want to buy future cash flows and earnings and the best possible price now.
Our FundaQuant scoring process is unique, powerful, flexible and effective. We use it to create expected return forecasts for all stocks. These expected return forecasts are thus truly forward looking and not just based on historical price action. We can then, later let in a little price action that is designed to serve as a confirmation to out FundaQuant scores. Our research shows that this historical price influence (only when internally optimized) should account for no more than 15% of a forward-looking estimate.
FundaQuant thus represents the remaining 85%.
We succeed by predicting many things. Just to predict an expected return of one single stock we make over 40 predictions. This keeps the predictions grounded and is another way to prevent overconcentration. We only have to be more right than wrong over all 40 predictions.
Whenever we feel that we have discovered a reliable signal in the data we look to quantify the implication of that signal to the assets expected returns. Our scores are thus the composite of all the signals carefully proportioned to reflect the total idiosyncratic expected return for any asset.
Our current weights look something like this:
The score produced by FundaQuant is then literally passed to Gsphere where it combines with the assets potential diversification and risks to achieve the optimal combination.