Increasing The Probability Of Success - Part 1
Probabilities are used in calculating probability-weighted returns. Since probability is the most subjective part of the probabilty-weighted return, this article discusses multiple approaches that can be used for different firms.
I was discussing with a new client how analysts should approach probabilities. Probabilities are used in calculating probability-weighted returns by multiplying them by the client’s scenarios of price forecasts to come up with a probability-weighted return.
The probability piece is the most subjective part of the probability-weighted return (see our “False Precision” blog post that discusses why it is important to set probabilities), so we came up with several approaches to see what fit best for their firm. I thought I’d share them with anyone that may be struggling with probabilities:
1. Fixed Probabilities (Distribution)
Analysts come up with price targets that match the part of the forecast distribution associated with the probabilities. In this example, all positions have a “fixed” 20%/60%/20% probability framework. The goal is to come up with price targets that match those buckets (i.e. what is the 20% risk price target?).
This method pulls price targets associated that reflect the probability-weighted outcomes associated with a broad range of outcomes associated with different probability “buckets”. An analyst would iterate the assumptions in their financial model to estimate the extreme outcomes (two 20% probability buckets at the end) and the higher probability outcomes (60% probability bucket in the middle). The result is a price target that blends the possible outcomes in each bucket by their associated probability. Another way to think of this is a cumulative probability distribution.
For example, the analyst may associate-5% sales growth and 10% EBITDA margins as the 20% cumulative probability outcome, 25% sales growth and 40% EBITDA margins as the 80% cumulative probability outcome, and 60% growth and 55% margins as the 99% cumulative probability. There would be many other points in between (represented by the green dots) where the analyst would apply different assumptions in their model.
The benefits of this method are that the probabilities are fixed and require no subjective assessment. This method also allows for highly-sensitive models with extreme outcomes to be reflected in the resultant probability-weighted return. The downside of this method is that it is time-intensive and allows no flexibility in the probabilities.