Dear Hedgeye Nation,
In this new edition of Real Conversations, Hedgeye CEO Keith McCullough is joined by Annie Duke, former professional poker player, World Series of poker winner and author of How to Decide: Simple Tools for Making Better Choices.
Making high-probability bets, under conditions of uncertainty, is as relevant in poker as it is in financial markets. Don't miss this pro-to-pro conversation about decision making live on HedgeyeTV.
Please enjoy the entire hour-long interview below. We’ve also transcribed key excerpts from the eye-opening conversation between Annie and Keith for your convenience.
* * *
Duke: The work that I do, which is decision-making under uncertainty, people just relate to it no matter what they're doing. Everyone is trying to address different types of uncertainty. But in terms of what I'm talking about, it doesn't matter if you're fast-cycle and doing something like options trading, or if you're the complete opposite side of the spectrum in the slowest capital you could imagine, in something like seed or angel stage venture, which has an extremely long time horizon.
No matter which end of that spectrum you're on, you're feeling the effects of hidden information, you're feeling the effects of intervention of luck. When you're over on the seed or angel side, you're particularly feeling the effects of hidden information, because obviously you don't know very much about the company at that time. When you're over on the options side of the world, you're feeling more volatility and luck, and the fact that the world is stochastic.
In both ends, there's a dual influence, and you want to figure out how to create the best process given the amount of uncertainty that there is.
I also work quite a bit on the advisor side, because I think advisors really struggle to get their clients to behave in accordance with their stated long-term best interests. We know when people state their long term interests, they don't necessarily make choices in the short run that would match up with those long term goals.
McCullough: In my world, it's all about process; that's all I want to improve. And at the same time that I'm trying to improve my process, I'm trying to coach it. And that second part is very difficult because I'm trying to explain exactly what I'm doing. That's been my challenge, and you're books have really given me a specific set of avenues where I can get into some of these topics, like resulting.
Let's talk about stochastics; a lot of people in my world don't think stochastically, they think on a linear basis, or deterministically. I do think there's an 80/20 rule there - I think 80% of people think linearly, and I don't.
Duke: I think what goes along with that is that a majority of people have a high need for closure as well; those go hand in hand.
And what does a high need for closure mean - it means that you want to be able to close the loop and know for sure why something happened.
And you're going to get there much more easily if you sense the world from a deterministic place. That thinking works in the sense that if you do get a bad outcome, you think it must come from a bad process as well. That outcome was destined - whenever you made the decision, that outcome was determined. If that's the outcome that I had to have gotten, then obviously the process let me down.
That's a more deterministic way to look at the world, but you can also see how that satisfies that need for closure. Because then you always know "for sure" - you can just work backwards to close that loop.
But that's not the way the world works.
When you make a decision, what you're setting in place is not the outcome that you end up observing, but rather a set of possible outcomes. Each of those outcomes has some payoff associated with it, and some probability of occuring. You can create some broad groups; you can talk about the upside or downside, or get more granular and talk about the tails. When you're doing something like managing tail-risks, you're saying that this won't realize most of the time - in fact, it'll realize a very small percentage of the time.
What you're saying though is that when it does realize, the impact is so great, that I should want to figure out a strategy in order to blunt the effect if it does come to fruition.
But you need to get comfortable with the fact that whatever you're doing to try to manage that risk is going to feel unnecessary. You have to take a step back and look at it as an expectancy problem.