Latest / Masters in Business / Inside the Gerber Statistic with Sander Gerber of Hudson Bay Capital

Transcript

Inside the Gerber Statistic with Sander Gerber of Hudson Bay Capital · Masters in Business

0:02
Speaker 1: Bloomberg Audio Studios, podcasts, radio news. This is Masters in
0:11
Speaker 1: Business with Barry Ritholts on Bloomberg Radio.
0:16
Speaker 2: Strap yourself in for another good one. Sander Gerber, CEO
0:20
Speaker 2: CIO of Hudson Bay Capital. What a fascinating background he has,
0:26
Speaker 2: starting in philosophy and ending up on the floor of
0:30
Speaker 2: the American Stock Exchange as an equity options trader. That experience,
0:36
Speaker 2: those two things combined to really create a kind of
0:41
Speaker 2: unique perspective on the world of markets, on the world
0:44
Speaker 2: of risk, and on the world of models. You know,
0:48
Speaker 2: I've used the George Box quote a million times. All
0:52
Speaker 2: models are wrong, but some are useful. And the way
0:56
Speaker 2: Gerber goes about using models is very much along the
1:01
Speaker 2: George Box lines, which is, not only are we going
1:05
Speaker 2: to assume that models are wrong, but we want to
1:08
Speaker 2: create our own models to be able to identify when
1:11
Speaker 2: they're going to be at a great variance to what's
1:14
Speaker 2: going on in reality, and then how to position ourselves
1:17
Speaker 2: to take advantage of it. They're less directional traders than
1:21
Speaker 2: they are arbitreasures. Hudson Big Capital runs a dozen different
1:28
Speaker 2: strategies and they're all quite fascinating. Everything from risk, arb
1:33
Speaker 2: to private credit and real estate in the first quarter
1:38
Speaker 2: of twenty twenty five, where volatility spikes and a lot
1:41
Speaker 2: of people's expectations are dashed. Their models do really well.
1:46
Speaker 2: I find his depth of knowledge and his technical expertise
1:50
Speaker 2: to be absolutely fascinating. I think you'll find him to
1:54
Speaker 2: be fascinating also, with no further ado, my conversation with
1:58
Speaker 2: Hudson Bay Capitals Xander Gerbert. So let's start a little
2:04
Speaker 2: bit with your background bachelors and humanistic philosophy and an
2:09
Speaker 2: MBA from Wharton Finance. What was the career plan?
2:13
Speaker 3: Well, actually, I was good at math, so I first
2:15
Speaker 3: entered the Wharton School undergrad. I don't have an MBA
2:18
Speaker 3: from Wharton. And then when I was at Wharton, I
2:21
Speaker 3: didn't think I was getting an education, so I decided
2:26
Speaker 3: to transfer into the College of Arts and Sciences, so
2:30
Speaker 3: I got two degrees. Concurrently, I picked up a degree
2:32
Speaker 3: in philosophy. Humanistic philosophy. I wanted to understand the development
2:37
Speaker 3: of thought, how we got to where we are in.
2:39
Speaker 2: Society, epistemology or something more specific.
2:43
Speaker 3: It was moral philosophy, generally, starting with the ancient Greeks
2:47
Speaker 3: through the existentialists. I think that I used my philosophy
2:51
Speaker 3: background much more than my finance background, because it really
2:54
Speaker 3: gives you a different view on the world. When I
2:57
Speaker 3: was at Wharton colleg Andrew Krieger came in nineteen eighty
3:03
Speaker 3: seven to speak. He had majored in Sanskrit Eastern philosophy
3:10
Speaker 3: and then he got his MBA at Wharton and he
3:13
Speaker 3: was the leading FX trader at Banker's Trust. And he
3:16
Speaker 3: spoke about how his philosophy Eastern philosophy helped him understand
3:20
Speaker 3: the markets. That you might feel very convicted the markets
3:24
Speaker 3: should go a certain way, but the markets have their
3:26
Speaker 3: own mindset and you have to accept what the markets have.
3:30
Speaker 3: And it helped him emotionally to trade better because he
3:33
Speaker 3: realized that mother market was going to be right, and
3:36
Speaker 3: so it was from his philosophy background that he was
3:39
Speaker 3: able to reconcile that with him with his beliefs in
3:44
Speaker 3: terms of where markets should go, and it helped him
3:46
Speaker 3: to be a better trader.
3:47
Speaker 2: That I definitely can see that. You know the concept.
3:50
Speaker 2: I don't know if I'm stealing this from Zen Buddhism,
3:53
Speaker 2: but it's the water flows, but the rigid tree breaks
3:57
Speaker 2: in the storm, and it's very similar to, hey, that's
4:01
Speaker 2: an Eastern way of saying, why are you finding the
4:03
Speaker 2: trend exactly?
4:04
Speaker 3: And so, you know, when I was in college, I
4:06
Speaker 3: really didn't know much about the markets. And as I
4:09
Speaker 3: told you, I still I had entered first the Wharton School,
4:12
Speaker 3: so I was still getting my degree there, but I
4:13
Speaker 3: was really focused on the philosophy. And you know, people
4:17
Speaker 3: think the philosophy is not so practical, what are you
4:19
Speaker 3: going to do with it? And here the top FX
4:21
Speaker 3: trader in the world came and said, this is what
4:23
Speaker 3: you should be doing. So it was it was sort of,
4:26
Speaker 3: you know, ratification of what I was studying.
4:30
Speaker 2: Huh. I think you're the first person who I've ever
4:33
Speaker 2: spoken to who said, yeah, the Wharton School of Finance
4:37
Speaker 2: at University of Pennsylvania not a great education. Isn't it
4:42
Speaker 2: really true that most of our education, or at least
4:45
Speaker 2: for a lot of people, you're just self taught. Schools
4:48
Speaker 2: will give you a curriculum and here's the reading list,
4:51
Speaker 2: but it's up to you to kind of learn whatever
4:54
Speaker 2: there is to learn.
4:56
Speaker 3: I think it's a good point. You know, the Wharton
4:58
Speaker 3: School is arguable the finest finance school, but finance is
5:03
Speaker 3: a technical discipline, and I wanted to understand the world.
5:07
Speaker 3: And I think that you can only go a certain
5:10
Speaker 3: degree using that background. And it's true. Then in order
5:14
Speaker 3: to I think, upgrade yourself, you've got to be able
5:19
Speaker 3: to develop the capacity to self learn, to take in
5:24
Speaker 3: from the environment around you, to enable yourself to grow
5:28
Speaker 3: your skill set to your experiences through working with others.
5:32
Speaker 3: And that's something we try to incorporate within Hudson Bay
5:36
Speaker 3: is the ability for people's careers to develop, and it
5:40
Speaker 3: is something that you have to rely on self learning
5:43
Speaker 3: and within college in certain disciplines. In college, like in philosophy,
5:48
Speaker 3: a lot of it is you know, discovery, self discovery,
5:52
Speaker 3: and other disciplines there is no self discovery. So I
5:55
Speaker 3: think it is important to the humanistic background.
5:58
Speaker 2: So you come out of of Wharton and University of Pennsylvania,
6:03
Speaker 2: you start your career on the floor of the American
6:05
Speaker 2: Stock Exchange as an equity options market maker. That had
6:10
Speaker 2: to be a fascinating experience, especially nineteen nineties and two thousands,
6:16
Speaker 2: that was a hot period and option trading. Tell us
6:19
Speaker 2: a little bit about that experience.
6:21
Speaker 3: Well, actually, when I graduated Penn I had been I'd
6:25
Speaker 3: clerked on the floor of the Philadelphia Options Exchange in
6:29
Speaker 3: nineteen eighty seven, and I liked it. But my parents
6:33
Speaker 3: had spent all this money to send me to a
6:35
Speaker 3: fancy school. They had taken out a home equity loan
6:38
Speaker 3: to pay for my college tuition. So I thought to
6:41
Speaker 3: be a muslely floor trader would be disrespectful. So I
6:43
Speaker 3: went to Banning Company for two years, and I was
6:46
Speaker 3: in management consulting for two years. It was boring, but
6:50
Speaker 3: I did learn something from it, and then I came
6:52
Speaker 3: to the floor of the AMEX.
6:54
Speaker 2: Wait before you jump to the AMEX. Aside from learning
6:59
Speaker 2: that being was boring, what else did you learn?
7:02
Speaker 3: I learned how people can work together in good conscious
7:08
Speaker 3: with dedication and still muck things up. Because what we
7:12
Speaker 3: would do is we would parachute into places like British Airways,
7:16
Speaker 3: Montreal Trusts, uh CIA Industries, and we were like the
7:21
Speaker 3: external strategic planning and we would They would put young
7:25
Speaker 3: people like me, and we'd sit next to people and
7:28
Speaker 3: interview them and figure out why projects went to muck.
7:31
Speaker 3: And I understood from that that well meaning people can
7:35
Speaker 3: still muck things up because they don't have an appropriate
7:38
Speaker 3: guide frame or appropriate leadership. Or they're not so like
7:43
Speaker 3: little things can take projects astray.
7:47
Speaker 2: So what was it that drew you to the floor
7:49
Speaker 2: of the well?
7:50
Speaker 3: I enjoyed the Philadelphia floor, and also I was I
7:54
Speaker 3: always liked games, and so I and I had a
8:00
Speaker 3: talent I thought for trading, and so I went to
8:05
Speaker 3: the the AMEX. Someone gave me it was like eleven
8:09
Speaker 3: hundred dollars a month as a stipend, and I kept
8:12
Speaker 3: roughly half the profits. And there was no training. They
8:15
Speaker 3: just threw me there very in the deep.
8:17
Speaker 2: End of the pool. Whoever doesn't drown. Hey, you can grab.
8:20
Speaker 3: Exactly right, exactly right. And it took me from July
8:26
Speaker 3: of ninety one till December of ninety one. I made
8:28
Speaker 3: five hundred dollars profit. Not for me, five hundred dollars.
8:33
Speaker 3: Trading had a split which I had to split. Yes, well,
8:37
Speaker 3: actually because I had a draw, I didn't get anything.
8:39
Speaker 3: But then the next year I took off and it
8:41
Speaker 3: turned out that I did have a knack fort I
8:44
Speaker 3: was able to understand the volatility of the market. Is
8:48
Speaker 3: usually we're vol traders, and I did something that was
8:53
Speaker 3: two things that were novel on the floor. The first
8:57
Speaker 3: is I understood that you have to break down your
8:59
Speaker 3: volatile exposure month by month, which back then was unusual.
9:04
Speaker 3: In other words, people had these models that would give
9:06
Speaker 3: you one volatility exposure across the entire portfolio. And I
9:10
Speaker 3: realized that julys and earnings month, and August is a
9:13
Speaker 3: beach month, so you can't use those two months to
9:16
Speaker 3: offset each other. And so I was able to jerry
9:18
Speaker 3: rig the models that were early then to be able
9:21
Speaker 3: to look at my VEGA exposure month by month. That was,
9:24
Speaker 3: believe it or not unusual. And the second thing that
9:27
Speaker 3: that's early nineties is yes, that was ninety one, ninety two.
9:31
Speaker 2: Okay, all these things we kind of take for granted.
9:34
Speaker 2: I know, right at one point in time, you wonder
9:37
Speaker 2: why it's become so increasingly difficult to beat the broad index.
9:41
Speaker 2: It was a ton of inefficiencies, that's right, that's right.
9:43
Speaker 3: And it was a great edge for me to come
9:45
Speaker 3: to that realization. And maybe it was because I had
9:48
Speaker 3: studied the models at the Wharton School. We had broken
9:51
Speaker 3: them down, and I understood that the models are only
9:53
Speaker 3: as good as the inputs. And a lot of people
9:56
Speaker 3: back then were doing spreads in their head and the
9:59
Speaker 3: other group were using these canned models that would give
10:02
Speaker 3: you one volatility exposure across you know, the entire model.
10:07
Speaker 3: And the second thing that I realized was that you
10:11
Speaker 3: need to combine fundamentals with the technicals of the models.
10:16
Speaker 3: In other words, the models assumer normal distribution of returns,
10:21
Speaker 3: but when you get into some kind of event, it's
10:26
Speaker 3: no longer a normal distribution returns. It's you know, the
10:29
Speaker 3: stock's either going to go up a lot or down
10:31
Speaker 3: a lot. That's a barbell distribution, right as opposed to
10:35
Speaker 3: normal distribution. And so by looking at events and when
10:39
Speaker 3: they're going to happen and breaking down the VEGA exposure
10:43
Speaker 3: month by month, that gave me an edge that I
10:45
Speaker 3: was able to exploit. Do you find vega for listeners
10:48
Speaker 3: who are Vega is the volatility So of the an
10:53
Speaker 3: option has premium, and that premium is the extra amount
10:57
Speaker 3: you pay for the right to have limited loss and
11:00
Speaker 3: unlimited gain. And so that premium, that value of that
11:07
Speaker 3: option to exercise or not exercise with limited loss, goes
11:12
Speaker 3: up and down in value based upon the degree of movements.
11:16
Speaker 3: So when something's moving around a lot, that has a
11:18
Speaker 3: lot more value. So premium value goes up when things
11:21
Speaker 3: are not moving a lot, premium value goes down, and
11:24
Speaker 3: so by trading this range of volatility up and down,
11:29
Speaker 3: which is in part dependent on what's happening with the
11:32
Speaker 3: fundamentals of the stock, you were able to grab edge.
11:37
Speaker 2: So these are really second or third level derivatives. It's
11:40
Speaker 2: not the underlying value. It's the increase in value of
11:44
Speaker 2: the option and then within that the range and the
11:49
Speaker 2: variability of that increase in option value. That's what you
11:52
Speaker 2: were trading.
11:53
Speaker 3: Yes, and you know it's really not complicated. I mean
11:58
Speaker 3: Wall Street tries to make things much more more complicated
12:00
Speaker 3: than they are, but the simple, elegant solution is always better.
12:06
Speaker 3: So it might sound complicated, but it's really not right.
12:09
Speaker 2: And that complexity is a feature, not a bug. You
12:13
Speaker 2: can sell stuff if it's complicated and hard to understand.
12:16
Speaker 2: If it's simple, well I think I could do that.
12:19
Speaker 3: That's right. Wall Street tries to make things more complicated
12:21
Speaker 3: because it has to justify the sales commission and if
12:26
Speaker 3: but things really are not so complicated.
12:29
Speaker 2: So what was your biggest takeaway from your experiences as
12:33
Speaker 2: a trader? How did it shape how you look at
12:36
Speaker 2: the world of investing, How did it affect what you're
12:40
Speaker 2: doing in Hudson Bay today.
12:41
Speaker 3: Well, I really was grounded by that three and a
12:44
Speaker 3: half years of watching every tick on the stock. You know,
12:47
Speaker 3: and you're you're geographically limited on the floor. You can
12:52
Speaker 3: only trade at the post that you're standing by.
12:54
Speaker 2: Like physically in space, your physically heether to that trading
12:59
Speaker 2: thing exactly.
13:00
Speaker 3: And there are even rules that you had to do
13:02
Speaker 3: most of your trading in that geography, so you couldn't
13:05
Speaker 3: move around a lot. And what it taught me is that,
13:09
Speaker 3: you know, like a trading post, a strategy goes in
13:13
Speaker 3: and out of favor, and if you want to be
13:15
Speaker 3: able to make money in all markets all the time,
13:18
Speaker 3: you have to develop a toolkit that can go beyond
13:22
Speaker 3: one particular strategy. So you need to have multiple strategies
13:26
Speaker 3: to develop persistent profitability. The other thing that I learned
13:31
Speaker 3: was that you can make the right decisions and still
13:33
Speaker 3: lose money. I had plenty of time where looking back,
13:37
Speaker 3: it was the right decision, but the markets thought differently,
13:40
Speaker 3: and so you always have to be worried about what
13:43
Speaker 3: can go wrong. And risk is not about not losing money.
13:48
Speaker 3: Risk management is not about not losing money. Risk management
13:51
Speaker 3: is about unexpectedly losing money. In other words, when you're
13:56
Speaker 3: evaluating a situation, you should know what is your reason
14:00
Speaker 3: worst case downside. Now there's always the black swan that
14:04
Speaker 3: maybe you can't figure on, but you should. But risk
14:08
Speaker 3: management is always about understanding what could go wrong and
14:12
Speaker 3: quantifying what could go wrong.
14:14
Speaker 2: So I want to unpack what you just said, because
14:16
Speaker 2: it's filled with goodness. First, you're referring to your approach
14:23
Speaker 2: is Hey, we're really more process focused than outcome focused,
14:27
Speaker 2: because if you have a good process, even if you
14:30
Speaker 2: get a bad outcome, it doesn't matter. Probabilities will eventually
14:34
Speaker 2: work in your fay.
14:35
Speaker 3: Exactly right.
14:36
Speaker 2: That's number one. But then the part two, which I
14:39
Speaker 2: think a lot of investors overlook, is and a risk
14:44
Speaker 2: management component that if the worst case happens, we still
14:48
Speaker 2: survive and lift to trade another.
14:50
Speaker 3: That's right, exactly right. And so at Hudson Bay, I
14:53
Speaker 3: created the deal code system, uh.
14:56
Speaker 2: Deal code system.
14:58
Speaker 3: Yes, so at the time, well, I left the floor
15:02
Speaker 3: beginning of ninety five and started deploying just the money
15:05
Speaker 3: I'd earned on the floor in an off floor trading account.
15:09
Speaker 3: And I would develop a strategy and hire someone else
15:13
Speaker 3: to run it and develop another strategy and hire someone
15:15
Speaker 3: else to run it. And as I was having other
15:18
Speaker 3: people manage basically my trading account, I realized I had
15:24
Speaker 3: to scale my risk profile that I developed on the
15:26
Speaker 3: floor over multiple risk takers, and I needed to do
15:31
Speaker 3: it in a manner that would produce persistent profitability. So
15:35
Speaker 3: at the time, we were trading a lot of risk
15:36
Speaker 3: garbitrage deals, so we called it a deal code, and
15:40
Speaker 3: a deal code is just a numerical moniker that we
15:43
Speaker 3: put on each trading idea within the book, and that
15:47
Speaker 3: enables us to focus in on how is that trade hedged,
15:51
Speaker 3: what's the risk riskiness? How much could that trade lose
15:55
Speaker 3: in a reasonable worst case scenario, and it gives us
15:58
Speaker 3: a batting average, so we can under stand is a
16:03
Speaker 3: portfolio manager winning more ideas than they lose so to
16:06
Speaker 3: be persistently profitable. I think it's not just about winning
16:08
Speaker 3: more dollars than you lose. It's about winning more ideas
16:11
Speaker 3: than you lose.
16:13
Speaker 2: So let's talk a little bit about Hudson Bay's strategy.
16:17
Speaker 2: You've been managing outside capital across a variety of asset
16:22
Speaker 2: classes and strategies. Tell us talk about some of the
16:27
Speaker 2: key strategies and what has been the drivers of making
16:31
Speaker 2: those strategies successful. Well.
16:33
Speaker 3: As I mentioned, I wanted to be able to make
16:35
Speaker 3: money in all market environments, so you need a tool
16:37
Speaker 3: set to do that. So our strategies are equity long
16:41
Speaker 3: short converts, credit, event merger, volatility trading.
16:48
Speaker 2: This isn't just I'm going to buy the S and
16:50
Speaker 2: P five hundred and put it away for a decade.
16:53
Speaker 2: You're active traders, and you're really looking to take advantage
16:57
Speaker 2: of situations where you have a fairly good idea of
17:01
Speaker 2: what the outcome is going to look like. It's not hey,
17:04
Speaker 2: this is open ended. Usually you're pretty confident in here's
17:08
Speaker 2: what our range of potential outcomes are.
17:10
Speaker 3: I think that, especially in today's world, you have to
17:13
Speaker 3: understand what your edge is versus the machines. And a
17:17
Speaker 3: machine can calculate risk based on historical precedent, but a
17:23
Speaker 3: machine cannot calculate risk based upon some kind of uncertainty
17:26
Speaker 3: due to some kind of event, callous or change that's
17:29
Speaker 3: coming up because it's new. So the machine doesn't have
17:32
Speaker 3: the ability to calibrate for something that's new. And so
17:35
Speaker 3: generally across all our strategies, that's what we're focused on,
17:38
Speaker 3: is we're focused on event callous change. How can we
17:42
Speaker 3: profit off of that in a way that machines cannot.
17:45
Speaker 2: So that's the fundamental criticism of models. All models assume
17:50
Speaker 2: that the world in the future is going to look
17:52
Speaker 2: like the world in the past. Risk management is what
17:55
Speaker 2: happens if the world doesn't look like out at.
17:57
Speaker 3: Us precisely, And that's why we don't use the standard
18:00
Speaker 3: risk management models. I actually created a statistic, the Gerber statistic,
18:05
Speaker 3: that helps to understand diversification between our deal codes, between
18:10
Speaker 3: our investment positions. A lot of our competitors are tied
18:13
Speaker 3: to factor based modeling, which ultimately, underneath it is reliant
18:19
Speaker 3: on regression analysis. Regressions. Our straight line fits through normalized
18:24
Speaker 3: sets of data, and human relationships don't file straight lines,
18:28
Speaker 3: and certainly market relationships don't file straight lines. So using
18:32
Speaker 3: that as the underpinning of a risk management system is
18:38
Speaker 3: just incorrect. And so we've created a whole different structure.
18:42
Speaker 3: As I said, we've used since nineteen ninety eight, and
18:44
Speaker 3: I think that's given us the ability to weather storms
18:49
Speaker 3: and profit from it in ways that our competitors can.
18:52
Speaker 2: So let's talk a little bit about the Gerber statistic.
18:56
Speaker 2: You had this validated by Harry Markowitz, the creator of
19:02
Speaker 2: modern portfolio folio theory. Tell us about that collaboration and
19:09
Speaker 2: break down the Garberg statistic a little bit. How do
19:11
Speaker 2: you guys actually use it?
19:14
Speaker 3: So, because of my distrust of models, based upon my
19:16
Speaker 3: experience on the floor, in particularly the guts of the models,
19:20
Speaker 3: I never believed in the correlation statistic, that correlation is predictive,
19:27
Speaker 3: and this was I thought one of the underpinnings of
19:30
Speaker 3: modern portfolio theory that you look at the expected return
19:33
Speaker 3: of the stock, the expected variants of the stock, and
19:37
Speaker 3: the covariance of correlation between the different components of a portfolio.
19:43
Speaker 3: And at the time, you know, we used the deal
19:46
Speaker 3: code system and on Wall Street the banks were telling
19:50
Speaker 3: me this is nonsense, but don't even talk about it
19:52
Speaker 3: with investors. And then in eight when everyone lost money
19:55
Speaker 3: and we made money, I realized we were doing something different.
19:59
Speaker 3: And then I had the idea. Because of course I'd
20:00
Speaker 3: studied about Harry in modern portfolio theory. Everyone in finance has.
20:05
Speaker 3: He won the Nobel Prize. I decided, you know what,
20:08
Speaker 3: I'm going to go out to see him to see
20:10
Speaker 3: what he thinks about the Gerber statistic, and at the
20:13
Speaker 3: time it wasn't called the Gerber statistic. But a friend
20:15
Speaker 3: of mine said, gee, you really should file a patent
20:17
Speaker 3: on this before you see Harry, and so I did,
20:20
Speaker 3: and I had to name it something, so I called
20:22
Speaker 3: it the Gerber statistic. And we now have I think
20:24
Speaker 3: we just got our sixth patent on our process for diversification.
20:28
Speaker 3: So I got to see Harry in San Diego. Lovely guy.
20:32
Speaker 3: He welcomed me, and we're walking. He liked to walk
20:36
Speaker 3: along the beach and I said, Harry, you know, I
20:39
Speaker 3: don't think that correlation's predictive, and Harry said, you're right.
20:43
Speaker 3: I said, no, no, Harry, you don't understand it. I don't
20:46
Speaker 3: think that because it's one of the base foundational bases
20:49
Speaker 3: for what She won the Nobel Prize in Modern portfolio theory.
20:52
Speaker 3: Said Harry, I don't think that historical correlation has relevance
20:56
Speaker 3: to the future. And he said, you're right. And it
20:59
Speaker 3: turns out that in his nineteen fifty two paper that
21:02
Speaker 3: sets forth modern portfolio theory, he said that correlation should
21:06
Speaker 3: be determined by the judgment of practical men. In other words,
21:10
Speaker 3: the stock analysts should think what will be the relationship
21:13
Speaker 3: going forward, not to mind the past, but be forward looking.
21:17
Speaker 3: But in the nineteen sixties, as computing power increase, people said, oh,
21:22
Speaker 3: we can mind the statistic, this row statistic correlation, and
21:26
Speaker 3: then we can plug it into the model as correlation.
21:29
Speaker 3: He meant correlation in a semantic sense, not in a
21:33
Speaker 3: mathematical sense in terms of using in his model. So
21:37
Speaker 3: he actually said that the deal code system uses his system,
21:41
Speaker 3: the modern portfolio theory system. He said that there's three
21:45
Speaker 3: legs to his system. And so because we use limited loss,
21:49
Speaker 3: because we seek to diversification through hesing on the own,
21:53
Speaker 3: because we seek to win more than we lose in
21:56
Speaker 3: each investment idea, he said that is accordance with his system.
22:01
Speaker 3: But in any way, we we've written several papers together
22:05
Speaker 3: on the Gerber statistic within modern portfolio theory and have
22:08
Speaker 3: demonstrated that you get better performance with less risk by
22:12
Speaker 3: replacing historical covariance with the Gerber statistic. And Harry and
22:17
Speaker 3: I actually we only had really one disagreement, and the
22:20
Speaker 3: one disagreement was on factors. There's all these you know,
22:23
Speaker 3: factor methodologies, and Harry believed that only one factor matters
22:27
Speaker 3: for portfolios, and I think two factors matter, so and
22:31
Speaker 3: so that, but the other twenty three factors we both
22:34
Speaker 3: agree are complete nonsets.
22:36
Speaker 2: So if you look at the fomb of French model,
22:41
Speaker 2: which started out as two or three factors and then
22:43
Speaker 2: became five fact.
22:44
Speaker 3: And then grow and grow. If you speak to the
22:47
Speaker 3: research departments of bar Axioma, they'll tell you that thirty
22:52
Speaker 3: four to forty percent of a stock price movement can
22:55
Speaker 3: be explained by factors.
22:57
Speaker 2: Okay, so that's third, let's.
22:59
Speaker 3: Roll it a and of that third, eighty five percent
23:03
Speaker 3: of that third can be explained by the first five factors, okay,
23:07
Speaker 3: which means giving credit to five, which that's bar Naxioma
23:11
Speaker 3: tells you eighty five percent of the forty percent can
23:16
Speaker 3: be explained by five factors, which means the other twenty
23:19
Speaker 3: factors explain the fifteen percent of forty percent of the words.
23:22
Speaker 3: Six percent of a stock price movement can be explained
23:25
Speaker 3: by twenty one factors, right, meaning which is complete. You know, nonsense,
23:30
Speaker 3: but no, if you lever a portfolio up, you know
23:34
Speaker 3: ten times, all of a sudden, that six percent looks
23:37
Speaker 3: like it's sixty percent, but it's all complete nonsense. It's
23:40
Speaker 3: numerical mumbo jumbo. It's part of the whole Wall Street
23:44
Speaker 3: pizazz that is not based on reality. But you know
23:48
Speaker 3: it sells.
23:49
Speaker 2: So so I want to guess the two factors. If
23:52
Speaker 2: I had a guess, I'm going to rely on a
23:54
Speaker 2: paper by Wes Gray of Alpha Architect and guess it's
23:58
Speaker 2: value and momentum. But I'm curious what you feel.
24:01
Speaker 3: Well, actually, Harry thought it was market. I think his
24:03
Speaker 3: market and section.
24:04
Speaker 2: So is market and sector. But are those really factors?
24:07
Speaker 2: Do we really The.
24:07
Speaker 3: Whole idea of factors is kind of like, you know,
24:12
Speaker 3: a little nonsense. It's like beta, you know, like market
24:15
Speaker 3: we think of as beta. It's now been called a factor.
24:19
Speaker 2: So oh, I never really thought of beta as a factor.
24:22
Speaker 2: It's just it's, hey, if you do nothing, you get
24:26
Speaker 2: But that's market, you know, So huh, that's really it.
24:29
Speaker 2: So you're looking at the sector it's in and the
24:32
Speaker 2: overall market as the two driving facts.
24:34
Speaker 3: I think those are Now it's true that momentum, value,
24:38
Speaker 3: these other things are relevant today because everyone else has
24:42
Speaker 3: glommed onto it, because we have so many statistical, process
24:45
Speaker 3: driven strategies that try to trade momentum. You know, buy cheap,
24:50
Speaker 3: sell expensive. It pushes everything in line. And this is
24:53
Speaker 3: what I found on the floor using models to trade options,
24:57
Speaker 3: that the models would push the the values of the
25:01
Speaker 3: options into alignment in accordance with the model because everyone's
25:05
Speaker 3: using the same model, and so the same thing is
25:08
Speaker 3: true in the broader market because everyone's using basically the
25:10
Speaker 3: same factor models. It pushes things in alignment, which works
25:14
Speaker 3: in normal market environments. But when things you know, have
25:19
Speaker 3: a dislocation, it no longer works, which is why people say, oh,
25:22
Speaker 3: our risk model broke down or whatever, because these aren't
25:25
Speaker 3: really risk models.
25:26
Speaker 1: Now.
25:26
Speaker 3: It's one thing to use a model to trade because
25:31
Speaker 3: the model's telling you something is some expensive or cheap.
25:34
Speaker 2: And relative to history.
25:36
Speaker 3: Right, And if something's always cheap, you just adjust the model.
25:39
Speaker 3: So there's a validity to that. But that's different than
25:42
Speaker 3: using the same model for risk management. Risk management, again,
25:46
Speaker 3: is about avoiding unexpected loss.
25:49
Speaker 2: Huh. That's really interesting. So when I started on a
25:53
Speaker 2: trading desk, one of the things that I was always taught,
25:56
Speaker 2: which I never contextualized as a factor, is, hey, what's
26:02
Speaker 2: driving the stock? Well, the stock is only a tiny
26:06
Speaker 2: part of it. The stock is twenty percent, the sector
26:10
Speaker 2: is thirty percent, and half is the market. So you
26:13
Speaker 2: could be the greatest stock in the world. If the
26:15
Speaker 2: market's going down, it doesn't matter, and it could be
26:18
Speaker 2: a really good stock. But if it's in a terrible sector.
26:21
Speaker 2: You know, the metaphor was always great house in a
26:25
Speaker 2: crappy neighborhood is a crappy house. You're really putting that
26:29
Speaker 2: into the context of these are the broader factors that
26:32
Speaker 2: are affecting that single holding.
26:34
Speaker 3: That's right, that's right. And you know, in our at
26:37
Speaker 3: Hudson Bay, we seek to produce the alpha. So it's
26:41
Speaker 3: true that the market is moving the stock, but we
26:44
Speaker 3: try to pick stocks that outperform the market or pick
26:48
Speaker 3: shorts that will go down more than the market. So
26:51
Speaker 3: we seek to focus on the alpha provision.
26:54
Speaker 2: So let's talk about something related to this. A paper
26:58
Speaker 2: you published, environment eats culture for lunch. It sounds like
27:02
Speaker 2: the environment is what the market's doing with the sector is,
27:06
Speaker 2: but give us a little detail about.
27:08
Speaker 3: Well, actually, I mean that that paper was related to
27:12
Speaker 3: the human aspect, not the market. So Peter Drucker came
27:17
Speaker 3: up with this idea that culture eats strategy for breakfast
27:20
Speaker 3: that corporate culture is actually more important than corporate strategy
27:26
Speaker 3: for the success of a firm. I think there's a
27:29
Speaker 3: lot to that that, you know, the way people work
27:33
Speaker 3: together in an organization. But I've always thought that this
27:36
Speaker 3: corporate culture thing is nonsense. If you have people try
27:39
Speaker 3: to describe their corporate culture, they cannot articulate it, right,
27:42
Speaker 3: you know, like what's the corporate culture here at Bloomberg,
27:45
Speaker 3: you know, like.
27:47
Speaker 2: Fun data driven? It's all about data, So you come up.
27:50
Speaker 3: On data driven. It's not a culture. Data driven is
27:52
Speaker 3: a process. But I'm talking about what's the human aspect
27:55
Speaker 3: of it? What's what's the human culture.
27:57
Speaker 2: I'm the wrong person to ask that, right.
27:59
Speaker 3: Because because no one can really describe corporate culture, what
28:03
Speaker 3: you can describe as an environment. What is the environment
28:06
Speaker 3: that people work within? And I kind of learned this
28:09
Speaker 3: at Band and Company because Baine was described as this
28:13
Speaker 3: like fun loving place, everyone has fun. And then when
28:16
Speaker 3: I was there, two guys died on the locker bee
28:19
Speaker 3: crash and Bill Bayn had milked the esop and so
28:21
Speaker 3: the company almost collapsed. When I was there, they fired
28:24
Speaker 3: half of my class, not me, They fired all the
28:27
Speaker 3: incoming MBAs and it was the avarice of Bill Bain
28:32
Speaker 3: that nearly collapsed the firm. We're talking back in nineteen
28:37
Speaker 3: eighty nine to ninety.
28:38
Speaker 2: So the corporate culture was with pacious greed, well did
28:42
Speaker 2: you know, and almost destroy.
28:44
Speaker 3: It was inauthentic. And when people try to describe culture,
28:48
Speaker 3: they can't. And so what I wanted to do was
28:50
Speaker 3: to describe an environment. What is the environment that you
28:53
Speaker 3: want to work within? And you know when you speak
28:56
Speaker 3: to when you speak to people on other firms, what's
29:01
Speaker 3: your corporate culture? What's your value statements? Usually these things
29:04
Speaker 3: go on and on and on. No one can really
29:06
Speaker 3: remember all the value statement. If you can't remember your
29:09
Speaker 3: value statement, it has no value.
29:12
Speaker 2: I'm going to imagine that twenty two twenty three when
29:16
Speaker 2: all the big firms were saying, we want our employees
29:19
Speaker 2: back in the office. We don't want any more remote work.
29:23
Speaker 2: It's a matter of corporate culture. How did you think
29:27
Speaker 2: about that? Was this a legitimate demand and is it
29:33
Speaker 2: not so much corporate culture? But we want an environment
29:35
Speaker 2: where people are in the office working together. Is that legit?
29:39
Speaker 3: I hate going to the office and seeing people not there.
29:43
Speaker 3: I think that people should work together. On the other hand,
29:46
Speaker 3: You can't force these things. You can't force independent thinking,
29:51
Speaker 3: you can't force collaboration. You can have an environment that
29:54
Speaker 3: engenders it, and so we try to have an environment
29:57
Speaker 3: that engenders it. So it's my opinion that people who
30:01
Speaker 3: come to the office are going to succeed more than
30:03
Speaker 3: people who don't. Now, I understand that, you know, the
30:06
Speaker 3: commute is a hassle and sometimes people, you know, want
30:11
Speaker 3: to take the day off, and so you know, our
30:13
Speaker 3: standard is two days in the office. Many teams have
30:17
Speaker 3: a third day, but a lot of people. Usually people
30:20
Speaker 3: are in our office three to five days a week.
30:22
Speaker 3: But we don't force it. If once you force people
30:24
Speaker 3: to be in the office, I think you're losing this
30:26
Speaker 3: spree de corps. We want people to want to work
30:29
Speaker 3: at Hudson Bay. If they don't want to work at
30:31
Speaker 3: Hudson Bay, they should go elsewhere. But to force people,
30:34
Speaker 3: I think, you know, for high performers, I don't think
30:39
Speaker 3: that's the way to engender the right environment.
30:42
Speaker 2: And environment beats culture for work because the work environment
30:47
Speaker 2: is more important than some statement that nobody remembers. Correct.
30:51
Speaker 2: So you guys have let's talk a little bit about
30:53
Speaker 2: independent thought. You guys have done pretty well. When the
30:57
Speaker 2: expert's wrong. You throw five, seven, eight, and nine. You
31:01
Speaker 2: were notably up in years where most people were down.
31:06
Speaker 2: Again in Q one of twenty twenty, you guys did
31:09
Speaker 2: really well all periods of big market turmoil. I don't
31:13
Speaker 2: know what you were doing in two thousand and one two,
31:16
Speaker 2: but I'm imagining the same approach held true. How do
31:22
Speaker 2: you think about these periods? Are they truly black swans
31:28
Speaker 2: or are they things that, with the right approach to
31:31
Speaker 2: risk management, are create opportunities.
31:34
Speaker 3: Again, people are trying to assess risk based upon some
31:39
Speaker 3: kind of parametric distribution with you know, standard deviation movements,
31:45
Speaker 3: and I think that's just nonsense. The markets don't work
31:48
Speaker 3: like that. So our system enables us to weather all
31:52
Speaker 3: market environments through the deal code system by ignoring those
31:58
Speaker 3: parametric The Gerber statistic, which is the basis for the
32:02
Speaker 3: work with Harry, is a rank order statistic because it
32:08
Speaker 3: recognizes the failures of parametric normal distributions. And what we
32:14
Speaker 3: do is we set a threshold because a lot of
32:17
Speaker 3: data is noise in the markets. If the S and
32:19
Speaker 3: P moves by ten basis points, it doesn't communicate to
32:22
Speaker 3: you how the S and P affects other things. Yet,
32:25
Speaker 3: and all these statistical models, they're including every single data
32:28
Speaker 3: point because if you don't include every single data point,
32:32
Speaker 3: then in the matrix math you have a divide by
32:34
Speaker 3: zero issue. So they're forced in all these correlation statistics,
32:40
Speaker 3: these regression analyses to include every single data point. With
32:44
Speaker 3: the Gerba statistic, we are able to create thresholds where
32:47
Speaker 3: we ignore data below a certain degree of movement, and
32:51
Speaker 3: so that enables us to focus on Everyone wants meaningful relationships, right,
32:55
Speaker 3: So this is how we're able to focus on meaningful
32:58
Speaker 3: relationships within the market.
33:01
Speaker 2: You know, we talked a little bit about sub prime
33:03
Speaker 2: real estate and how the models it wasn't even that
33:07
Speaker 2: they broke. They were so poorly constructed they were destined
33:10
Speaker 2: to fail. You know, if you build a house really poorly,
33:13
Speaker 2: you don't need an earthquake. Eventually, it's just going to
33:15
Speaker 2: collapse under its own weight. But I have to ask
33:18
Speaker 2: you some questions about real estate because Hudson Bay has
33:22
Speaker 2: been increasingly invested in private credit and real estate. You've
33:27
Speaker 2: done a number of major refinancings in and around New
33:30
Speaker 2: York City. Six twenty Avenue of the America's is tell
33:35
Speaker 2: Us a little bit about the work you're doing at
33:37
Speaker 2: Hudson Bay with private credit and real estate.
33:40
Speaker 3: Well, we saw beginning with the the transitory higher rates,
33:47
Speaker 3: which we thought was nonsense, right. We saw that rates
33:50
Speaker 3: were going to be higher for longer, and we had
33:55
Speaker 3: believed that the market had been anchored in this idea
33:57
Speaker 3: of ultra low rates, which was really a manipulation of
34:02
Speaker 3: the monetary system. So we started thinking about what's the
34:05
Speaker 3: implications of that, and came to the notion that the
34:09
Speaker 3: banking system would be under stress. And what's the implication
34:13
Speaker 3: of the banking system under stress. Well, that means that
34:16
Speaker 3: they can't extend loans in the same way, you know,
34:20
Speaker 3: corporate as well as real estate. So we started staffing
34:24
Speaker 3: up in those areas to take advantage. And now I'm
34:27
Speaker 3: convinced that the there's now going to be a structural
34:31
Speaker 3: shift in credit provision in the US economy, that the
34:34
Speaker 3: banks are no longer going to be the mainstay for credit.
34:38
Speaker 3: And that's because the government has effectively guaranteed our banking system,
34:44
Speaker 3: which creates moral hazard. We have on the order of,
34:48
Speaker 3: you know, forty three hundred banks in the United States.
34:51
Speaker 3: It's a lot, especially when you compare it to Canada
34:54
Speaker 3: that's got the big you know, handful, and you know
34:58
Speaker 3: when you deposit money in the bank, that bank is
35:02
Speaker 3: lending it out long.
35:04
Speaker 2: And fractionally reserving it. So it's ten to one, whatever
35:08
Speaker 2: the precise the leverage there using.
35:10
Speaker 3: So I think that the whole fractional banking system notion
35:15
Speaker 3: is challenged, particularly in the idea of the ease of
35:20
Speaker 3: information transparency among depositors coupled with the necessity for government
35:26
Speaker 3: guarantee and moral hazards. So private credit firms like ours
35:31
Speaker 3: people invest in Hudson Bay and they know it's not
35:34
Speaker 3: a bank account, and that gives us license to deploy
35:38
Speaker 3: the money in ways that are appropriate, and so we
35:42
Speaker 3: began staffing up in those areas. And now in real estate,
35:45
Speaker 3: for instance, we have teams that work in real estate equity,
35:49
Speaker 3: in CMBs, distress CMBs, and direct provision of real estate credit.
35:57
Speaker 3: And as part of the core vet you've Hudson Bay.
36:00
Speaker 3: These teams work together, which give us a better understanding.
36:04
Speaker 3: It's a great advantage to have equity teams working with
36:07
Speaker 3: credit teams, particularly all real estate's local It gives us
36:11
Speaker 3: a much better understanding of the asset that we're looking at.
36:16
Speaker 2: Huh, that's really kind of interesting. You know, Ever since
36:19
Speaker 2: the financial crisis, some of the new regulations and bank
36:23
Speaker 2: regulations directly led to the rise of private equity, private credit.
36:31
Speaker 2: You know, some of the forecasts are over the next decade,
36:33
Speaker 2: this blows up to a thirteen trillion dollar asset class.
36:37
Speaker 3: I think we're in the third inning, early early days here, Yeah,
36:40
Speaker 3: I think so.
36:41
Speaker 2: And it it feels like it's been so big because
36:45
Speaker 2: we started with practically nothing in that space, and the
36:48
Speaker 2: first couple of trillion dollars felt like, oh, my goodness,
36:51
Speaker 2: is just so much capital washing over this. But this
36:55
Speaker 2: seems to have happened in the past where woll banks
37:00
Speaker 2: and brokers kind of move up market, they create a
37:03
Speaker 2: void in the space they left, and private money rushes
37:08
Speaker 2: in to fill that void. Is that what's going on
37:11
Speaker 2: with private credit and real estate?
37:14
Speaker 3: Well, it's still early in that. I think it's a
37:16
Speaker 3: golden age for real estate credit. The banks are not
37:20
Speaker 3: able to they don't have the capital now to lend,
37:24
Speaker 3: and so there's it's open season.
37:27
Speaker 2: Huh. Really really interesting. So how do you identify opportunities
37:31
Speaker 2: in the real estate space. It seems like there are
37:34
Speaker 2: so many buildings that are half empty, and yet it's
37:39
Speaker 2: a slow motion train wreck because most of their tenants
37:43
Speaker 2: have ten or longer year leases and they're just slowly
37:49
Speaker 2: starting to recognize unless you're a super A class building,
37:54
Speaker 2: even A buildings are having a hard time attracting renewals
37:59
Speaker 2: and tenants. How you identify these and how far along
38:02
Speaker 2: the repricing of commercial real estate or at least offices
38:08
Speaker 2: do you think we are?
38:09
Speaker 3: Well, those are big questions. And I'm from Annaburg, Michigan,
38:13
Speaker 3: and I saw how in Detroit, Detroit was going to
38:15
Speaker 3: be called the museum to the desolate city because downtown
38:22
Speaker 3: Detroit went empty when they built the Renaissance Center. Everyone
38:25
Speaker 3: moved to the Renaissance Center and left these empty, huge
38:29
Speaker 3: buildings in Detroit. And you see aspects of that now
38:32
Speaker 3: where the A buildings, the new buildings are attracting very
38:36
Speaker 3: high rents and buildings in other areas are you going empty?
38:43
Speaker 3: So to understand what's going on, you really have to
38:46
Speaker 3: understand the asset, and so that's why it's important to
38:48
Speaker 3: have teams from different disciplines being able to understand the asset. Obviously,
38:54
Speaker 3: looking through the rent rolls and understanding you know, the
38:58
Speaker 3: weight to average lease, but also understanding the macro environment.
39:04
Speaker 3: You know, are things growing And we have so much
39:05
Speaker 3: uncertainty now going on, not just because of work from
39:09
Speaker 3: home with Zoom, but also the longer term implications of
39:14
Speaker 3: AI and what's that going to mean for the workforce
39:16
Speaker 3: and even cities like New York City. It's possible that
39:20
Speaker 3: we're not going to need the same number of junior lawyers,
39:23
Speaker 3: junior accountants, junior bankers.
39:26
Speaker 2: So I've heard some people discuss AI as a tool,
39:30
Speaker 2: and it's not that you're going to lose your job
39:32
Speaker 2: to AI, but you're more likely to lose your job
39:36
Speaker 2: to someone working with AI. Is that a fair assessment
39:41
Speaker 2: or is it just still way too early to take.
39:42
Speaker 3: I think we still don't know. I think AI is
39:45
Speaker 3: the greatest change in my lifetime, bigger than the Internet.
39:48
Speaker 3: I think so, yeah, really yeah, because the ability for
39:53
Speaker 3: natural language processing goes far beyond what I thought was possible.
39:57
Speaker 3: You know, I studied linguistics a bit in college. The
40:00
Speaker 3: whole idea of how we form language is a fascinating subject.
40:05
Speaker 3: And now the computer is able to be coachent in
40:07
Speaker 3: their responses, We've you know, kind of approaching hard AI
40:14
Speaker 3: in a way that I did not think was possible,
40:17
Speaker 3: and it's only going to get better.
40:19
Speaker 2: Let me push back a little bit. And I'm not
40:21
Speaker 2: necessarily saying I believe this, but so I've had this
40:28
Speaker 2: conversation over and over again with a number of different people.
40:31
Speaker 2: How are you using AI in your daily work? What
40:34
Speaker 2: are you finding? And someone who hosts a different podcast said,
40:40
Speaker 2: they created this really interesting set of prompts with AI
40:47
Speaker 2: to get an answer to how to do certain things,
40:50
Speaker 2: and the first time they got the answer, they were
40:54
Speaker 2: really impressed. Oh my god, this is a genius insight,
40:57
Speaker 2: and look how smart this is and how it it
41:00
Speaker 2: figured out exactly what I needed. And then they asked
41:03
Speaker 2: a different question with a different subject kind of got
41:07
Speaker 2: the same answer, and it was like, oh, this is
41:10
Speaker 2: a party trick. This isn't really intelligence. It just looks
41:16
Speaker 2: like intelligence, and even though it's getting better, it's still
41:21
Speaker 2: kind of dumb relative to it impresses us. But once
41:26
Speaker 2: you peer behind the curtain and see the wizard is
41:29
Speaker 2: just a man, you figure out this is less what
41:33
Speaker 2: it purports to be in more like a very useful,
41:37
Speaker 2: clever trick.
41:38
Speaker 3: I was thinking of a Wizard of Oz also while
41:40
Speaker 3: you were while you were saying that, But I don't
41:42
Speaker 3: think there's a guy behind the curtain that's giving the answers.
41:44
Speaker 3: That's why I think that it helps with the junior
41:48
Speaker 3: analysts that you have to check anyway, and it certainly
41:52
Speaker 3: speeds up the research process in ways that were not
41:56
Speaker 3: possible before, for sure, and it's only going to get better.
41:59
Speaker 3: And it may makes mistakes, but the junior analyst makes
42:02
Speaker 3: mistakes also. I mean, I've used it for things my
42:06
Speaker 3: lawyers probably will hate me, but sometimes when I've had
42:08
Speaker 3: a discussion with the lawyers on how to express something
42:12
Speaker 3: in a document, to all ask AI the question. It
42:15
Speaker 3: will give me a range of possibilities and enables me
42:17
Speaker 3: then to be more on a level playing field with
42:20
Speaker 3: my lawyers who have had a lot more experience than
42:22
Speaker 3: I have. But it has enabled me to bring to
42:25
Speaker 3: the discussion insights that we might not have thought of.
42:28
Speaker 2: I'm glad you brought up the attorneys, because a judge
42:32
Speaker 2: just sanctions a lawyer for using AI and in certain
42:37
Speaker 2: of his answers, and this unfortunate tendency to hallucinate. I
42:45
Speaker 2: don't think the problem was that he used AI to
42:47
Speaker 2: help him in research. He didn't double check it, and
42:49
Speaker 2: he failed to disclose that AI was plathiness.
42:52
Speaker 3: You know, it's just plain laziness. The the AI is
42:57
Speaker 3: good for the junior person, and I think as implications
43:00
Speaker 3: for the workforce, you know, what is the workforce going
43:02
Speaker 3: to look like? Given that, maybe we don't need the
43:07
Speaker 3: same failans of junior accountants, junior lawyers, junior bankers.
43:12
Speaker 2: How do you become a senior account lawyer, banker if
43:15
Speaker 2: you're never a junior It's a tough question. So let
43:18
Speaker 2: me give you an opportunity to update your twenty twenty
43:23
Speaker 2: one piece in investing. Don't short human judgment? Do you?
43:28
Speaker 2: Are you still holding that for you?
43:29
Speaker 3: Absolutely? I mean we are in the human judgment business. Really,
43:34
Speaker 3: we are trying to beat the machines. We do that,
43:40
Speaker 3: as I said, through understanding uncertainty, events, catalysts, and change,
43:45
Speaker 3: and I think ultimately human judgment is superior in the machines.
43:49
Speaker 3: I hope we won't go into a Hell two thousand
43:52
Speaker 3: type situation that human judgment will always be superior. You
43:56
Speaker 3: wouldn't want to have a machine be the a in
44:00
Speaker 3: the United States. How could a machine possibly make those decisions,
44:05
Speaker 3: you know. So obviously human judgment will always be there,
44:08
Speaker 3: and I don't think that we're at a terminator type,
44:11
Speaker 3: you know, situation, but there are certain experts that say
44:14
Speaker 3: that ultimately that's where we'll go. I mean, I do
44:17
Speaker 3: know that in the military, you know, the idea of
44:20
Speaker 3: robots creating robots is a real idea, and it very
44:24
Speaker 3: might well change battlefield dynamics. But I believe that certainly,
44:32
Speaker 3: at this point in time, the human capacity to ingest
44:36
Speaker 3: a mosaic of information and to make the right decision
44:42
Speaker 3: is superior. If you take a chessboard, the machine can
44:46
Speaker 3: beat the master, but if you put an extra bishop
44:49
Speaker 3: on the board, the machine can't deal with it, right,
44:53
Speaker 3: And I think that's the paradigm. And life does not
44:56
Speaker 3: mimic a chessboard, you know. Life mimics the chessboard with
45:00
Speaker 3: extra pieces being put on randomly, and it's that randomness
45:04
Speaker 3: that I don't think the machines will be superior than
45:06
Speaker 3: human judgment. Now, it might appear at times that the
45:09
Speaker 3: machine can beat the human, but I think ultimately the
45:12
Speaker 3: human judgment is superior, and so our business is based
45:16
Speaker 3: on human judgment.
45:18
Speaker 2: You mentioned the wartime usage of AI. There was a
45:22
Speaker 2: pretty big article I don't remember. I want to say
45:24
Speaker 2: the Times, not the journal, that figured out that in
45:28
Speaker 2: the Ukraine Russian War, which started out as a conventional
45:33
Speaker 2: bombardment between tanks and mortars and anti tank weapons, over
45:38
Speaker 2: the past six twelve months, seventy percent of the casualties
45:43
Speaker 2: have been drone AI warfare driven, and it's very much
45:49
Speaker 2: a brave new world. It's not like the old world
45:52
Speaker 2: of warfare. What it sounds like you're suggesting with AI
45:58
Speaker 2: is that they're both code developed, that you'll still have
46:01
Speaker 2: humans driving the process, but AIS become an increasingly large
46:07
Speaker 2: part of it, regardless of whether we're talking about warfare,
46:10
Speaker 2: business or investing. I don't want to put words into
46:12
Speaker 2: your mouth, but is that a fair way to assess that.
46:15
Speaker 3: I think so. I mean, I think that the humans
46:18
Speaker 3: always have to be on top of the machines. Machines
46:20
Speaker 3: have a lot of latitude, both to produce themselves as
46:24
Speaker 3: as well as to target. You know, the markets are
46:28
Speaker 3: different because the markets follow a behavioral dynamic. The evaluation
46:32
Speaker 3: of risk versus reward is something that I think a
46:36
Speaker 3: machine cannot do in the same way the human can.
46:39
Speaker 2: So given some of the volatility we've been seeing in
46:43
Speaker 2: the first quarter of twenty twenty five. Has that changed
46:48
Speaker 2: how you're looking at your models, how you're viewing your
46:52
Speaker 2: approach or is it, Hey, this is just another one
46:55
Speaker 2: of those things that comes along and we have to
46:57
Speaker 2: be able to trade through.
47:00
Speaker 3: We actually like the dislocation because the dislocation proves the
47:04
Speaker 3: models are wrong.
47:05
Speaker 2: Well, I know you guys don't release public performance numbers,
47:09
Speaker 2: but I know you're doing much better than your benchmark
47:13
Speaker 2: this quarter. Volatility is your friend. Is that what you're saying?
47:17
Speaker 2: Because volatility disrupts traditional models and you're a non traditional model. Correct.
47:22
Speaker 2: So I know you've worked with Harry Markowitz. What other
47:26
Speaker 2: academics and what other institutions have you worked with?
47:29
Speaker 3: Well, at Imperial College London, there's further work being done
47:33
Speaker 3: on the Gerber statistic and incorporating it. The idea of
47:37
Speaker 3: thresholding data and ways to do it to For instance,
47:43
Speaker 3: if you want to understand the significance of a stock
47:46
Speaker 3: price movement, maybe you should exclude days where there's very
47:50
Speaker 3: low volume and only include days when there's high volume.
47:53
Speaker 3: There's a variety of ways to incorporate it.
47:57
Speaker 2: I know, I only have you for a limited amount
47:59
Speaker 2: of time. Let me jump some of my favorite questions.
48:02
Speaker 2: I ask all of our guests, what are you watching
48:04
Speaker 2: or listening to? With? What's keeping you entertained?
48:07
Speaker 3: Recently I streamed Eastern Gate?
48:10
Speaker 2: Oh really?
48:11
Speaker 3: Which is I saw in the New York Times. It
48:13
Speaker 3: was this spy thriller series on the conflict between Poland
48:18
Speaker 3: and Belarus, and I wanted to understand the dynamic between it.
48:22
Speaker 3: So I thought I'd get a little entertainment and understand
48:25
Speaker 3: something I couldn't pick up here. And it's a little slapstick,
48:29
Speaker 3: but I think it's worth it.
48:30
Speaker 2: Eastern Gate. Yes, did you happen to watch any of
48:33
Speaker 2: Fouda when that was just the most heart wrenching stuff
48:38
Speaker 2: to watch? It's so stressful.
48:40
Speaker 3: Yeah, and pretty realistic.
48:42
Speaker 2: Very realistic. Let's talk about mentors who helped shape your career.
48:47
Speaker 3: I gotta give a lot of credit to Dave Patrice.
48:50
Speaker 2: Who I know that name, who really.
48:53
Speaker 3: Helped me get into shape. And he was on my
48:57
Speaker 3: case every day, the diet, the working out. We're workout partners,
49:04
Speaker 3: and I was thirty five forty pounds heavier, uh huh,
49:08
Speaker 3: and he got me to recognize they needed to get
49:12
Speaker 3: in shape. I thought I was in shape, but I
49:13
Speaker 3: wasn't in shape. I think I think a lot of
49:15
Speaker 3: people think they're doing okay when they could do a
49:17
Speaker 3: lot better. And he taught me I could do a
49:20
Speaker 3: lot better. And I think it's affected me overall, my
49:22
Speaker 3: mental acuity, my mood, my stamina. I really give him
49:29
Speaker 3: a lot of credit.
49:30
Speaker 2: You mentioned books earlier. What are some of your favorites?
49:32
Speaker 2: What are you reading right now?
49:34
Speaker 3: One book that I really enjoyed, which was long, was
49:37
Speaker 3: Walter Isaacson's book on Elon Musk, which I read before
49:40
Speaker 3: the election, and it made a big impact on me
49:43
Speaker 3: because I believe in questioning the experts, but must takes
49:46
Speaker 3: it to a different level. He's questioning metallurgical properties that
49:50
Speaker 3: were well grounded in science and engineering, and he's saying,
49:54
Speaker 3: why does that have to be? And oftentimes he was
49:57
Speaker 3: right that the established can census regarding properties of medals
50:03
Speaker 3: was wrong.
50:04
Speaker 2: M really really interesting. Any of the books you want
50:07
Speaker 2: to mention?
50:09
Speaker 3: I read The Melting Point by Frank Mackenzie recently. He
50:13
Speaker 3: was the head of Scentcom and he talked about what
50:16
Speaker 3: it was like to lead Sentcom and he also had
50:22
Speaker 3: a MA He measured in English, and he thought that
50:25
Speaker 3: his English background to be a commanding general was very
50:28
Speaker 3: helpful because I helped him to articulate better and to
50:33
Speaker 3: form consensus, you know, among his colleagues.
50:36
Speaker 2: Really really interesting. Our final two questions what sort of
50:41
Speaker 2: advice would you give to a recent grad interested in
50:44
Speaker 2: a career in either filling the blank, investing options trading,
50:50
Speaker 2: multi strategy management. What advice would you give to them?
50:54
Speaker 3: I think it's, you know, across all certainly service occupations,
50:59
Speaker 3: is you got to be will beat the machines, and
51:02
Speaker 3: to do that, you need to be independent thinker. You
51:07
Speaker 3: need to go against the grain, question the experts. You
51:10
Speaker 3: need to be able to do that. You need ab
51:12
Speaker 3: to work with other people, to learn from them, to
51:16
Speaker 3: expand your horizons, to expand the mosaic that you can
51:19
Speaker 3: bring to your independent thinking. And you got to be
51:22
Speaker 3: able to respect your colleague. So I think that those
51:26
Speaker 3: three things are a real guideposts for people.
51:28
Speaker 2: This goes back to your corporate culture, which your environment,
51:32
Speaker 2: corporate environment, my bad, your corporate environment, think independently, collaborate
51:38
Speaker 2: and respect the individual. Correct huh? And our final question,
51:42
Speaker 2: what do you know about the world of investing in finance?
51:45
Speaker 2: Today would have been useful when you were first getting
51:48
Speaker 2: started in the early nineties.
51:51
Speaker 3: I think that you know everything you learn in business
51:54
Speaker 3: school or economics, you can just throw out the window
51:59
Speaker 3: economics and of science. People try to portray economics as
52:03
Speaker 3: a science, and it simply is not. And so all
52:06
Speaker 3: the notions that we brought up regarding money supply, you know,
52:11
Speaker 3: Milton Freem would be turning over in his grave. Even
52:13
Speaker 3: though these principles might have some grounding, It's not scientific,
52:17
Speaker 3: you know. This is this is not a natural science.
52:21
Speaker 3: It's a behavioral science, and it's based upon how people
52:25
Speaker 3: interact with each other. And I think that that appreciation
52:29
Speaker 3: leads to the notion that oftentimes the academy or the
52:32
Speaker 3: experts try to profer things that everyone everyone seems to
52:40
Speaker 3: believe one way, and you think, how could I be right?
52:42
Speaker 3: Because everyone believes one way because this is what they
52:44
Speaker 3: studied in school, and if the authorities say it's that
52:47
Speaker 3: one way. And I think that as you go through
52:51
Speaker 3: life and you age, you realize that the Ivory Tower
52:55
Speaker 3: isn't always correct. In fact, a lot of times the
52:57
Speaker 3: Ivory Tower doesn't have the real life experience, and so
53:01
Speaker 3: they're flat out wrong.
53:03
Speaker 2: I'm trying to remember where I'm stealing this quote from
53:07
Speaker 2: Science Advance's One Funeral at a Time. The same is
53:11
Speaker 2: true with other things that Dick Thaylor said. Rather than
53:16
Speaker 2: wait for the rest of economics to catch up with
53:20
Speaker 2: behavioral finance, I'm just going to teach it to the
53:23
Speaker 2: younger generation and it'll infiltrate much more quickly than waiting
53:29
Speaker 2: for all of my peers to accept it. Really really fascinating, Sander.
53:34
Speaker 2: Thank you for being so generous with your time. We
53:37
Speaker 2: have been speaking with Sandra Gerber. He is CEO and
53:41
Speaker 2: CIO of Hudson Bay Capital. If you enjoy this conversation, well,
53:46
Speaker 2: be sure and check out any of the previous five
53:48
Speaker 2: hundred and fifty we've done over the past eleven years.
53:52
Speaker 2: You can find those at iTunes, Spotify, YouTube, Bloomberg, wherever
53:57
Speaker 2: you find your favorite podcast. And be sure and check
54:01
Speaker 2: out my new book How Not to Invest The Ideas,
54:05
Speaker 2: numbers and behavior that Destroys Wealth out today wherever you
54:10
Speaker 2: find your favorite books. I would be remiss if I
54:13
Speaker 2: do not thank the correct team that helps put these
54:15
Speaker 2: conversations together each week. John Washerman is my audio engineer.
54:20
Speaker 2: Ana Luke is my producer Sean Russo is my researcher.
54:25
Speaker 2: I'm Barry Ritholtz. You've been listening to Masters in Business
54:29
Speaker 2: on Bloomberg Radio