Michael Mauboussin’s Paper – BIN There, Done That

During their recent episode of the VALUE: After Hours Podcast, Taylor, Brewster, and Carlisle discussed Michael Mauboussin’s Paper – BIN There, Done That. Here’s an excerpt from the discussion:

Jake Taylor:
Yeah, let’s eat our veggies upfront. Now this paper, if you’re making any kind of decisions, discretionary in an investment context, this paper I would call must read. Really, the gist of it is, is that we all suffer from biases and everyone …

Tobias Carlisle:
Speak for yourself, man.

Jake Taylor:
Yeah, well, it’s true. Most of us have read about them, but it turns out that that almost does nothing. Just knowing about them does nothing to protect against them. You actually have to bake it into your process, how to control for it. What’s really fascinating is that the study that Mauboussin is largely quoting from is that they break it down into biases, information, and noise. What would you guess was the most important factor of those three things?

Bill Brewster:
Say what?

Tobias Carlisle:
Information to noise? I don’t know.

Jake Taylor:
Let’s …

Bill Brewster:
I don’t think you want to reduce noise and bias. I don’t think I’m adding anything to this conversation.

Tobias Carlisle:
Can you just …

Jake Taylor:
Maybe I’ll …

Bill Brewster:
I’m going to back to the corner now.

Tobias Carlisle:
Let’s define with that. Can you define what they are? Let’s talk about what they are.

Jake Taylor:
Yeah, sure. Bias would be defined is when you use rules of thumb or mental shortcuts that can lead you astray.

Tobias Carlisle:
The heuristics.

Jake Taylor:
Heuristics, yup. Actually, Mauboussin has done this one himself where he’s had 10,000 subjects take this test, where it turned out that on average, we had a 70% confidence on our guesses, but we were right 60% of the time. In his dataset, we’re 10 clicks overconfident, which is probably what you would expect at a minimum. Information is a subset of signals that are used relative to the full amount of information that’s available to make the decision. You think about if you have unique information, if you update your views faster than other people, and then the waiting of components of information can make a difference.

Jake Taylor:
Basically, the quality of your information and how do you interact with it. Then noise is actually chance variability in judgments. The funny study that they cite for this is Money magazine did this thing where they interviewed 50 accountants. They gave them the financials of a theoretical family of four that made $120,000 a year. They said, what’s this family’s tax bill? They came up with the range varied from $10,000 to $21,000.

Tobias Carlisle:
Woah.

Jake Taylor:
The range then tells you what the noise is. It basically like a standard deviation of those things. Based on the same inputs, tells you how much noise there is. I’ll cut to the punch line. It turns out that 25% of the errors can be explained by bias, 25% by information difference, and 50% is actually from noise. Noise cancellation, being mindful of noise turns out to be hugely important.

Tobias Carlisle:
How do you avoid it?

Tobias Carlisle:
Get the pros headphones on.

Jake Taylor:
Great question.

Bill Brewster:
Three minus new air pods, dude.

Jake Taylor:
[crosstalk 00:06:45]

Tobias Carlisle:
That was weak.

Bill Brewster:
The air pod pros, come on.

Jake Taylor:
Three ways to reduce noise according to Mauboussin. Number one, combine judgments. That means independent and errors offsetting. This is basically like wisdom of the crowds. That’s one way to reduce noise. Use algorithms. Set rules beforehand, but do think a checklist fall under this category. Then number three, they call it mediating [crosstalk 00:07:19].

Tobias Carlisle:
Just got sucked into the matrix then. You got to [crosstalk 00:07:20] again.

Bill Brewster:
Right into the bet. The checklist matrix.

Tobias Carlisle:
That I want you to know it’s something that they don’t want you to tell us.

Bill Brewster:
This is the really good stuff.

Jake Taylor:
All right, number two, use algorithms, which means setting rules and then following checklist. Number three is what they call mediating assessment protocols. What do you think about that is you basically define what attributes you think are the most important for making the decision. Then you gather facts and then you assign some score to them. It’s almost like an algorithm, but run through a human processing. Those are the ways that Mauboussin talked about to reducing noise.

Tobias Carlisle:
Yeah, I love that.

Bill Brewster:
I feel like one of the thoughts that I had when you were talking is rather than trying to be so precise in what you think fair value is for something, widening your fair value estimate a lot. I know it’s really unsatisfactory for people to hear that you think something is worth … I don’t know, say, it’s $7 to $14 billion. That is not precise at all. What it does say is I think if it’s less than seven materially, it’s a good buy. If it’s north of 14 materially, it’s probably a sell and anywhere in there. It’s I think embracing some of the imprecision, is a smart takeaway.

Jake Taylor:
You’re not going to sell any newsletters with that attitude.

Bill Brewster:
Yeah, I know. I can’t pound the table on my … I’m not in that business, which is helpful.

Tobias Carlisle:
I think that it’s a pretty good explanation of those. If you find it trading at 3, and then 7 to 14, that’s a good deal. Let’s just finish trading at 10 and your range is 7 to 14, that’s not very helpful.

Bill Brewster:
Yeah, then you stay away or hold, right?

Tobias Carlisle:
It’s a pause, but most positions of that way, anyway. Most things is 7 to 14 and it’s trailing at 10. That describes 90% of the positions that stocks that are out there at any given time. Maybe not now.

Jake Taylor:
Maybe.

Bill Brewster:
Yeah, now it’s been a little nuts.

Tobias Carlisle:
I really loved that approach. I build that into my process. I have very strict screening, very strict update, without side information, build portfolios in a particular way. I 100% agree with that and adhere to it as strictly as I possibly can because I think that I’m guilty of all of those bias and so on. I’m too competitive about it. It makes me to get the juices flowing too much. I got to wind it back and get a little bit cooler when I’m doing the trading and putting positions on.

Jake Taylor:
Yeah. Bill, how about you? I think you’re maybe the little more discretionary, little more qualitative.

Bill Brewster:
I certainly am now. I tell you, my brain reacted a weekend ago or so by trying to go to what I perceive like quality at a discount and a lot more diversified than I normally run. Then at the bottom of the portfolio, this is not investment advice. I don’t know how many different ways to say it. I picked up Sabre. It’s tiny. It’s really tiny. I think it’s a debt special situation. Their credit agreement indicates to me that the lending group understands that these type of events happen and that they might be willing to amend. Given my background, I think it’s possible that they amend. If they do, I think it’s a three to four X.

Tobias Carlisle:
What is it, Bill? Can you just give us a little more background of that?

Bill Brewster:
Yeah. They’re basically like the back end, what certain people see when they’re booking seats and stuff like that. Basically, it’s a bet on hotels and airlines coming back at some point. I think right now, it’s a debt bet more than anything. I think it’s an amendment issue. I don’t even think business quality. I don’t know that it’s a business I’m trying to own for the really long term, but I view it as a dead special situation.

Jake Taylor:
Is it traveling the streets and SaaSing the sheet?

Bill Brewster:
It’s definitely not SaaS. They like to say they’re SaaS. It’s like super cyclical. If bookings aren’t happening, it’s not getting revenues to the extent that they write. It’s not a true recurring business, but anyway. I don’t know, that’s what I found.

You can find out more about the VALUE: After Hours Podcast here – VALUE: After Hours Podcast. You can also listen to the podcast on your favorite podcast platforms here:

Apple Podcasts