During their latest episode of the VALUE: After Hours Podcast, Carbonneau, Forehand, Taylor, and Carlisle discuss What Impact Will AI Have On Investing? Here’s an excerpt from the episode:
Jack: Yeah. No, this is part of why I became a quant investor, because I’m so bad at trying to analyze this stuff and I’m like, “I might as well try to find some models that work over time and just try to follow them,” although they can still be hard to stick with. But yeah, part of it is, like, my recognition that I’m not very good at analyzing these kinds of things.
Justin: Don’t worry. We can just throw into ChatGPT and they’ll give us all of our answers and we’ll be all set. [chuckles]
Tobias: That’s so neat.
Jake: I was going to say we are changing the name of this podcast to AI: After Hours.
Justin: There you go.
Justin: All right. You guys are going to blow out.
Jake: A hot– [crosstalk]
Tobias: Value: After AI. Maybe that’s what we call it.
Jack: [chuckles] Okay. That’s actually an important thing to think about though what value is going to look like after AI.
Tobias: Yeah. What do you need–? The thing I like about value is that it picks up a whole lot of stuff. Like, it gets energy when energy is cheap. It gets the home builders when they’re relatively cheap. And then, it seems to work out– I think that continues to happen, doesn’t it? Value, it only likes it if nobody else likes it. That’s what everybody’s worried about, AI is going to get smart and pick this stuff up, but doesn’t it then definitionally not fall into value?
Jack: Yeah, I would say. And also, on the side of actually picking stocks, I’m not sure AI changes things all that much. I don’t think there’s going to be more alpha available in the market because AI is present. So, maybe it becomes like AI is competing other AIs to pick stocks or something like that. But I don’t know if it fundamentally changes like the way investing works. You’re still going to have periods where you struggle. You’re still going to have periods where your strategy doesn’t work. In terms of picking stocks, I’m sure on the high frequency side, it makes a big difference. But I’m not sure in the type of stuff we’re doing that it really makes that much of a difference. I don’t know what you guys think.
Jake: Well, I think about where do the three advantages come from. You have a data advantage, you have analytical advantage, or you have a behavioral advantage. Which of those three vectors is AI going to radically change compared to what’s happening today? It’s not clear to me that any of those are really that have a lot of juice in them. The datasets are pretty well–
Tobias: Data mined.
Jake: Yeah, everyone’s pretty well mined the shit out of them, [Tobias laughs] present company included, certainly. And then analytically speaking, I’m not sure what’s AI going to suss out about– Perhaps, there’s some correlations there that still haven’t been found between economics and business results. It’s possible, but I’m a little skeptical about that. Then behaviorally, unless AI, I guess, is making less mistakes than the humans. But to me, it seems like AI would make different mistakes, repeatable mistakes. So, I don’t know, it’d be interesting. [crosstalk] Humans can harness it to do a lot more, which will be, I think, amazing, but I don’t know if it’s like– I think augmented intelligence is probably more the apt AI, so than artificial intelligence.
Jack: Yeah. In a lot of ways, things stay the same. People are analyzing the crap out of the data. Like you said right now, if you have information that other people don’t have right now, you have an advantage. If an AI has information other people doesn’t have, it has an advantage. So, in a lot of ways, it’s a lot of the same type stuff. Maybe the changes are going to be more on the actual job side of Wall Street. You need less analysts, that kind of stuff. We just did a podcast about this. So, we’ve been thinking about this a lot. But I would say that’s what it is. The good analysts become a lot better, and you need less analysts, maybe stuff like that versus on the actual stock picking side.
Justin: The one thing with us though is, we’ve been talking about, we have proprietary data. We have almost 20 years of rating individual equities through anywhere from 12 strategies when we started to 20. Well, actually, in total, there’s 45 different models that we run, but not all of them are on the Validea site. But could we utilize AI to improve our investment process? I’m sure there’s something we could do there, but then it becomes a big back testing exercise, and you got to be careful with that too, because you’re just looking at the past historical data and saying, “Okay, what has worked the best?” But in terms of having a– [crosstalk]
Tobias: Hoping that 20 years–
Tobias: Hoping that 20 years is representative of what comes before and after?
Jake: And you are not overfitting the model in a big way?
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:
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: