We’ve just been listening to Michael Mauboussin’s interview with Patrick O’Shaughnessy on his podcast, Invest Like The Best. During the interview Mauboussin provides some great insights into what investors can do if they’ve added a stock to their portfolio, and it goes down. Are they still right or are they wrong? Here’s an excerpt from that interview:
Mauboussin: The other thing I’ve been super excited about over the years. If I endeavour to be a long term oriented investor. I’ll just make this up. I’m looking at two or three year time horizons. It really is difficult if you’re descretionary because you put the stock in your portfolio. You’re long and it goes down. Am I right or am I wrong? That’s an interesting question. Part of that is to say, are there mechanisms to give us intermediate feedback that’s useful and timely. That’s why the work by Phil Tetlock and others on things like Brier Scores.
O’Shaughnessy: What’s a Brier Score?
Mauboussin: A Brier Score is a measure of the quality of your probabilistic forecast. Glen Brier who is a meteorologist. So it’s basically judging like you woke up in the morning and said there’s x percent chance of snow, or whatever is was. How accurate are those forecasts. So the way a Brier Score works is basically zero means youve nailed it. Everyday its sunny, you say sunny, everday its rainy, you say rainy.
A Brier Score scale can go to one or two. Lets say you score a two, that means you’re wrong about everything. Its a way of keeping track of probabilistic forecasts.
A couple of things that are really important about this. One is, whenever Brier Scores are kept and feedback is given back to decision makers. Meteorologists, people in the medical field. They get better at this. They get better calibrated. So the feedback makes people better. So you say, what does that have to do with investing?
Well you think about it. You have a thesis on a stock. What I should be able to do. Is layout the path, the thesis that I expect. Again its deviating from what the market believes. That’s a really crucial thing. And then I should be able to assign probabilistic forecasts to certain sign-posts. So I believe sales are going to be higher than the market believes. There’s x probability that sales will exceed this amount in this quarter. And I can give a probability to it. So now I’ve set myself up for a Briers Score.
Its a probabilistic forecast. It’s within a specified time period. We can agree on the outcome. And its important to our thesis. And now all of a sudden I’m giving you all of this intermediate feedback. And by the way we talked about Baysean updating before. It opens up your mind to say, “Hmmm, I thought this was going to happen. It didn’t. Let’s talk about whether we’re in the same place and the thesis is still unfolding correctly.”
That’s a technique that’s super-cool.
It’s this idea of intermediate feedback in a field where the feedback is really messy and noisy. Again, if you’re using two to three year horizons. This is a way to give you weigh-points along the way to make sure you’re doing things right.
For all the latest news and podcasts, join our free newsletter here.
Don’t forget to check out our FREE Large Cap 1000 – Stock Screener, here at The Acquirer’s Multiple: