Ex-Citadel PM: All Hedge Fund Failures Are Because of One Reason - Rich Falk Wallace
from Odds on Open
1h 27m 8s
The job of a hedge fund portfolio manager (PM) is fundamentally to generate risk-adjusted alpha by delivering a financial product that outperforms the market benchmark. While many attribute success to unique fundamental insights, the majority of failures stem from poor portfolio construction and risk management—particularly misalignment with market correlations. Any correlation in a portfolio drags the sharp ratio down toward the market’s return-volatility ratio, making correlation control essential. Top PMs succeed by combining deep fundamental analysis with systematic risk modeling, identifying undervalued opportunities where market pricing is misaligned. They treat factor exposures as if they were individual stock positions, enforcing strict limits and actively managing anti-correlations. The most effective PMs integrate macro trends, supply chain dynamics, and thematic shifts into a cohesive analysis, using data and models to continuously test and refine their strategies. This structured, analytical approach—grounded in both intuition and modeling—enables them to generate consistent, sustainable returns. Crucially, the market’s ability to price in trends means that true alpha comes from identifying and acting on rare, non-correlated, or mispriced opportunities. Modern tools now allow real-time decomposition of returns, making it easier to assess exposures, risks, and performance across the portfolio. Ultimately, success hinges not on grand, untested bets, but on disciplined, data-informed, and rigorously managed portfolio construction that consistently beats the risk-free market alternative.
what is the job of a hedge fund PM? 90% of hedge fund failures or PM failures, or even probably
at some level analyst failures, are that portfolio construction, deciding all those things. Alpha
is competitive, and the only way to generate it today is to have a reasonably correct perspective
and advantage at several layers of the stack. Effectively, any amount of correlation of portfolio
drags your sharp ratio towards that market sharp ratio. When they blow up, whether it's equities
or other asset class or whatever, the story is never shucks. I actually am not as brilliant
as I was before. They got wrong, was risk, portfolio construction, being analytical about the
other half of the portfolio. Hey, when I look at the back end of our YouTube statistics, I can see
that only around 25% of you guys are subscribed. And I'd like to ask a favor. If you like the content
we create and you want to help us grow this show, if you could hit that subscribe button,
I would really appreciate it. I promise that in return, I will do my best to give you guys the
best content we can with the best guests and best interviews we can do. Thanks and back to the
episode. Rich, thank you so much for coming on the pod. For sure, thanks so much for having me.
Fundamentally, what is the job of a hedge fund PM? I think ultimately what a hedge fund PM
is doing is contributing to a financial product for the ultimate end user, which is an allocator,
somebody who's allocating capital, hopefully on behalf of a broader group. But so I think that
that framing actually translates into what the immediate, proximate thing that a hedge fund
manager has to do, which is generate outperformance relative to what the free alternative is.
Again, I'm framing everything in terms of a product that ultimately is being delivered by an
investor. But that product is designed to, like, the sort of product that someone is willing to
pay for at the level of fees that a hedge fund, for example, captures, is a return in excess of
the risk. And that's kind of, you know, obvious to say, and the definition of risk has many different
components, right? I think the classic Warren Buffett kind of school of things is that, you know,
that risk is purely fundamental. Let's just look at things in terms of the absolute long-term risk.
But there's a whole evolving school. I'm sure we'll talk about it. But obviously, that looks at,
let's model this in a more sort of systematic, thoughtful way. But so, you know, fundamentally the job
of a portfolio manager is to pick stocks, ultimately, if we're talking about the world of stock picking,
that are sort of outperform the fair kind of risk of that set of names. We can talk about how that's
achieved, what are the ways and drivers of alpha that people kind of generate. But fundamentally,
it's ultimately to be a component, at least, of delivering an attractive product to the end investor,
to the end owner of assets to the allocator. And so how do top hedge fund PMs outperform? How do they
find, how do they have a differentiated view? How do they, I guess, analyze the world, see everything
and go, okay, this isn't priced in. Let's make the markets a little bit more efficient.
For sure. So I think there's two angles to delivering that attractive, ultimate financial
product. And you highlighted, I'd say, the second, the more classically considered one. But I'll
just mention briefly the first, we can come back to it. But is that you don't leak excess
or unintentional sources of risk that could be not doing your research sufficiently about some
kind of, if you're short, a stock that could be bought out, if your meaning could be fundamental
types of risks, it could be, you know, a fundamental business question that you're not thinking
about, right? It could be the modellable types of factor risks and exposures, you know,
drivers of investment return or at least a volatility that don't necessarily have any demonstrated
amount of skill that could be a level deeper, probably, or a half level deeper is sizing and
selection. And you're not sort of way over sizing or undersizing ideas in a way that, again, leaks
kind of volatility. Because again, if you think about the product that's ultimately being sold,
just as any business has revenue and cost, fundamentally, to some extent, again, this volatility
definition is one of the vectors of risk. But fundamentally, that financial product that's
being sold has the return is the revenue and the cost line item is the volatility of that return
stream and is the risk. But anyway, so just to mention, half of the story are probably even more
than half of the story of generating, you know, outside's performance, what drives like the best
people in the world. I think it's very rare that firms say, you have a story of, you know, of a
great hedge fund manager who blows up that says, you know, oh, you know, I did, I cared too much
about risk and portfolio construction. You know, that was my vector of failure. I think it's much
more common that people say, oh, I did these incredible feces. I had this incredible differentiation,
but I messed up on portfolio construction and risk. I screwed up the book from that standpoint
and blew up that way as opposed to the other way around, right? So it's much more common like
that the vector of failure in achieving something excellent, because there are, and you know,
this is probably itself a debatable point, but there are a lot of people, not that it's that many
in the global scale, but there are a lot of people in, you know, if you walk in a midtown through a
dark, you find somebody who can generate a positive sort of idiosyncratic or non in the broadest
definition, some, you know, non non kind of beta based return. There's a good number of people who
can generate a positive hit rate on that. And downstream of that, you can generate positive
attractive products on the back of that positive alpha hitting rate. But the, but again,
the vector of, you know, 90% of hedge fund failures or PM failures, or even probably at some
mobile analyst failures are that portfolio construction, miss sizing, all those things. So I guess
the short way answer to that question is the biggest vector of failure is actually everything
below the, I think what has classically been the kind of, you know, level of, how do I find it
inside? But anyway, so that's half of the story at least if not more. But I'd say dealing with the
the first part, the way you mentioned, which is, okay, how do you actually do the research process?
What do you do? How do you like generate something insightful? I think ultimately for everyone,
it's a little bit different. Again, what you're trying to do is find a thesis that is misunderstood
by everybody else, or at least by the pricing in the market. It doesn't have to be everybody else,
right? I think that's one of the things that often people don't fully understand is, you know,
there's a, you know, many tens of trillions active equity market that is, you know, trading stocks,
and if three people have the same idea in your, in your management meeting, it doesn't like
meaningfully diminish the, the realized, you know, sort of potential alpha in that name. Of course,
people can bid up, you know, stock prices to the point that they diminish that. But the real thing is,
do you have an idea that has both, you know, a level of, kind of, that is both different and correct
about a stock? That's the threshold question in order to generate, you know, positive alpha
hitting. And then underneath that, how do you do that? I guess there's the whole, the whole
question of like, what research do you do? Do you look at for process differences? Do you look for
data advantages? And my perspective, and again, everybody will have a slightly different answer
to this question. Like every, if you have, you know, kind of, and I've seen you've had some
different styles of fundamental equity investor on, some people will have, hey, I really,
really the only alpha is heuristic is, okay, I've got a framework for looking at things. Yeah,
this is the recent argument of like, many people got the look on investing at SpaceX at various
fallowations, and only a handful of people made the size of the bad at the right kind of thing.
So that's one argument that there's sort of a heuristic elements. There's some that are like,
okay, you really have either differentiated data or less accessible data or data that is sort of
put in an amount of sort of refining of that data to the point that it can be usable to
synthesize with that heuristic. And then there's, so I think the, I mean, in the end,
those are kind of the two buckets heuristic encompassing a huge amount of different frameworks of
value and macro and management and ideas about what's changing in the world. And I think
my perspective on it is that I have a, I have a good friend who's a great investor as a partner
in a big hedge fund. And we both sort of say this, we're like, I mean, we sort of started our
careers and went up the path together. And the question was like, okay, do you care about valuation?
Or do you care about not beating numbers? And the answer is, is both, right? Do you care about
the macro environment? How it affects the business? Or do you care about just purely the fundamentals?
The answer is both. And so I think, you know, the truth is everything matters. And
everything matters for lots of reasons. But one key thing is that a lot of the alpha, my basic
perspective is made in interaction terms where some people are misunderstanding both of the underlying
data and the, you know, don't have sort of haven't put together a piece of the underlying data.
And they haven't understood heuristically how to think about if that data were to arise in
that way, what it would do to the business. Because again, a lot of people, even if they have
reasonable heuristics for valuing stocks or values or understanding sector dynamics or what they
would do, you know, in certain environments. If a certain set of data points or a certain
environment of the world, if it sort of feels remote to you, it's like, oh, well, if I had that
level of granular data, I would do this or that. A lot of people don't even think through every
sort of eventuality and don't have the sort of conviction in a path of the world to kind of,
to kind of think through all the heuristics that would even make that relevant. So I guess what I,
my basic view is alpha is competitive. And the only way to generate it today is to sort of be
have a reasonably correct perspective and advantage at like several layers of the stack today.
That's really what I kind of view of this. If you're a student who wants to work at a great
trading firm,
Listen up. Our brand partner, Onyx, the largest oil derivatives trading firm in the world,
is hiring junior rust developers. They're opening this up to people who haven't written
the line of rust in their lives because they'll train you from the ground up. From day one,
you're on a small team working on real projects, learning from senior engineers who built the systems.
If you're self-taught with serious projects to show for it, apply the link below.
I have a friend and mentor who's come on the podcast, Hedge Fund Manager, and he says to me,
your job as a Hedge Fund Manager, Hedge Fund PM, is number one, make money, and everything else
follows. That's it. The other thing, so the first part of what you were saying there about
you seeing more people get into trouble from a mismanagement of risk perspective rather than
actually making the wrong bet. I'd like to ask, how do the top hedge funds do it? You've worked
at many of the top hedge funds, you worked at Citadel, and I think everyone today talks about
how well these machines run. They, as you said, generate alpha for the allocators in a way
where everyone wins. They run an incredible quality, but a human strategy of getting the best people
plugged into their system and have them generate alpha for allocators. They can pay them a lot because
they're basically never down. They manage risk exceptionally well. I guess what's your take?
You worked at some of the top hedge funds. What do they do that makes them special in that regard
from a risk management perspective? What any fund does well that's succeeding is it manages in
the long term. You either, there's really only two ways out, and it's either to have an incredibly
high hit rate or incredibly high slugging or managing correlations. Those are really the only
two kind of things. If you're going to have an even reasonably high degree of concentration
in a portfolio, you have to have such high hitting and/or slugging to make that math work. And
and Warren Buffett would say that too. I forget what number he cites in his latest thing,
but it'll be like seven stocks made most of the P&L in my life. And of course, there's nuance to
it. There's those numbers happening on top of compounding and all kinds of subtleties. But big
picture, if you're going to have a reason, even not even that high, but a reasonably reasonable
degree of concentration, you have to have the burden of proof on your, on this skill demonstrated,
on, on hit rate and, and slugging has to be so enormous versus the much greater share of people
who have, you know, demonstrated long term success is managing correlations. And that's true of
every sort of multi-strat that you're thinking of. If it's the big names, if it's the less well-known
ones. And how do they manage correlation is the subtlety? Because if you simply take a real
quick picture, just for our audience who doesn't know exactly what that means. What does it mean to
manage correlations? Why is that important for a hedge fund business for a PM, et cetera?
Yeah. I'll say it maybe in one way. There's like probably 15 framings you could come up with.
One simple way to say it is you're kind of, you're fundamentally, you're bounded on your sort of
theoretical, at least. And again, nuance is a bound here. But I'll just sort of, you're bounded
on your ability to generate long term sharp ratio by your, what's the way to say this? You're
fundamentally, you're bounded by your correlation. And the reason you're bounded by your correlation
is because of your, of your, of the names in your portfolio. And the reason is basically because
any set of very correlated bets ends up replicating the sharp ratio of a, let's say, factor. And
that factor, you know, you could think about this as abstractly as we're stating here or just
think about it as the market factor or any factor in a portfolio. Factors, there's almost no,
like, think of the market factor. The market factor has a.6 sharp roughly. I mean, of course,
it varies between, depending on what his, his purity look at. And it can, and even if you look
even at whatever five and 10 year, maybe you find, if you, if you pick the right points on the,
on the time series, you could be between point four and point eight or something like that,
or even stretch a bit wider than that. But broad, big picture, something like a half, half sharp
ratio is what the market gives you for free. By the way, and you can go now get that from Vanguard,
obviously. And that's a free product you can, you can generate. And so, effectively, any amount
of correlation to portfolio, drags your sharp ratio towards that market sharp ratio. That's
like the simplest way to say it or any market factor, correlational subset. And so, meaning,
that isn't to say that the market doesn't generate great returns with long term, of course, it does,
right? There's a, there's a priced in ROIC to owning stocks in the long term, right? You can
pick that, that equity risk premium, is it, is it eight, is it six, is it 10, whatever you think?
But fundamentally, in conjunction with that expected return is a volatility cost that you're
taking on in order to get that long term compounding. And again, crucially, for anybody trying to
generate alpha, you shouldn't, in principle, over the long term, get paid for that kind of risk
return ratio. Like, you could just own the market for free. And so, everybody's job is to compete
with that very attractive alternative of, of the free market, right? And so again, anytime you generate
correlation in a book, you're sort of dragging your sharp ratio almost in rough proportion to
the extent to which you're correlated to that, you know, to that market portfolio, which is free,
right? And so, again, that's not to say that you can't ever accept that, accept that bet. And
again, there are environments or cases where if you're hit, if you're hit rate and you're slugging
or so great, that you can overcome that. Hey, I took a, I took a dead weight in my book that
drove me towards a sort of market type sharp in a portion of that total. But, but it was great enough
bet. I was, I was correctly predicting the AI supply chain at each level of the game. And we're
done. There's, there's a bunch of people who today could, could kind of point to that track record.
And again, and that's sort of the, the environment of old, right? If you look at, if it's the classic,
you know, kind of kind of Soros lineage or the Tiger lineage, there was a degree of abundant
enough alpha where some of the sort of utility curve of sharp ratio kind of came out in the wash
where you could generate almost market like returns or better in the top, top cases at a volatility
amount that was lower than the market, right? And that's, that was the equation. Today, it's,
it's kind of a different picture where often what these top hedge funds in the world are doing is they
are generating, you know, maybe on average, not that far from what the market compounds at. Although
there are years will be very different, but it's at a much lower volatility profile. So the sharp
ratio is just, you know, much, much higher than what the market delivers you. But again, effectively,
what you're managing. And again, the way people manage it is more sophisticated than just looking
at a correlational chart. We can talk about the details of that because correlations in realized
terms slip around constantly. It's why people, their, the whole job is to manage X anti correlations,
not backward looking correlations because, you know, long-term capital management is the classic
backward looking correlational management problem, which is correlation change in a moment, right? And
so managing backward looking correlation, but really what you're trying to do is manage forward
looking correlation. And so, but again, because any amount of correlation is what drags your sharp
ratio towards the market chart ratio is kind of the cleanest way to say. Oh, yes. And I remember
first learning about this by reading over Gappies work, you know, reading his tweets. And I think
people today fundamentally misunderstand the job of hedge funds. You have all the retail guys
talking about how well said it all only did 15%. I don't know what it is. They don't understand
that that 15% is at a much lower volatility. And they're selling a product to allocators, right?
That is, that is what they do. I guess on the way these businesses work and on the
specifics of how they manage risk or maybe, maybe even in general terms, if you can't speak to
the specifics, what are at a base level? What is, what are say the three things with what they do
with regards to risk management that are most important? Is it hedging out market or turn all
the different factors? If you're in the center book, what are the most important things? What is the
most important job for you to do? Yeah. Fundamentally, the most important job is to manage, yeah,
whether it's factor risk or whether it's, you know, and factors, factors, again, kind of framing every,
I'm sort of tying everything I guess in this conversation back to this concept of ex ante correlation,
because that's all ultimately a factor is. There's a few different ways to frame it or talk about
it. But fundamentally, a factor is any source of kind of correlation in a portfolio. It's a common
in the classic normal way of phrasing and a factor, a component of a, you know, a driver of outcomes,
right? And so in that same light that the core goal is to generate, as I said, manage correlations,
that's assuming you have, on average, across however many, however much breadth you have,
positive alpha hitting. But yeah, because those are the two fundamental goals to manage,
you know, forward looking correlation and to drive as high as you can, sort of,
Breath basically and positive hit rate at that breath hit rate and perhaps slugging perhaps sizing but those typically at
You know in broad scope if you're looking at hundreds of
tens or hundreds of
Portfolio managers. It's rare to see massively positive sizing skill or slugging
It's much more common to see the skill show up in in some kind of an aggregate hit rate on outfits rare to see those other pieces
But but so again your goals are basically to drive that kind of average hit rate
To drive breath in the book at that breath of that positive skill of course breath without that is not helpful necessarily and to reduce
X anti correlation and so of those three kind of drop like it's sort of simple to model what these businesses do in that
Framework like if you said okay, how do I drive higher average hit rate?
It's a talent to your point and it's and then there are scale advantages that help you get that talent
And how does the math of all that work fine? There's the breath which is again the recruiting effort and having more people and
Generating talent and then there's the X anti correlation piece and the way that people do that is basically look at
Any source of factor exposures at the at the individual portfolio level and then crucially how they add up across the book because
A piece that sometimes people don't appreciate is like let's say you have two books that have or ten books that have
You know momentum exposure and they're long tech and they're long the latest exciting kind of tech
But they're market neutral even if they have 80 80% idiot at scale that 80 comes down if every if every one of your pms has a similar ish
Exposure that book average even though you've gotten more breadth is actually much lower your idiot might be 60 or something like that and the reason for that is
That factor bets compound on each other because they've been designed to be those things which are sources of correlation
So those things compound whereas the remainder term whatever's not a factor bet that 15% in each of the case that they're all 85
percent or whatever
It becomes like a loaded dice that has a 50 or a loaded coin that has like 52% hit rate and if you take 10,000 coin flips or a thousand coin flips
It's going to roughly the the average is going to reduce to the the hit rate right on that those things diversify
Whereas the factor bets even if you keep adding more and more factor bets that compounds on one another right and so even you could add 100
You could have 10,000 it wouldn't reduce the volatility driven by that by that similarly pointed factor exposure
Whereas the independent idiosyncratic kinds of bets
You know do diversify way at scale and of course
What a factor is could be the classic momentum and value and those are all super those are all time tested true
That's a fact there's no doubt that sometimes there's a momentum rally and that there's that high beta stocks move together
There's no dispute about that. That's all that semi stocks move together or that software names have have to have a reasonably high degree of
anti-correlation there's no question, but then there's also additional ones people look at they look at crowding they look at
revision they look at
And then there's kind of this third level the people begin to think about which is thematics
Hey, I want to look at the and people do this via bank baskets
They do this via just constructing sort of groups and buckets of thematics
But it's basically just saying like is there some transient factor like okay, there's AI trade and you know
And do you want to model at that level of depth open question and open practice and sort of debate constantly about how much to go down the rabbit hole of
characterizing factors
One of the thing all that is
You could build a factor model constantly right a model of x-andie correlation saying it that way, but you can build a factor model that goes
that is you know more accurate that is ever more
Had ever higher r-squared in terms of explaining volatility in a portfolio
But you don't always want to go to the final limit like one way to increase the r-squared of a risk model would be to have a one-day
correlation like there are actually commercial models that have one day and some quants use them if they trade in today and things like that
But for mo you have to match the kind of model horizon to the investment horizon because otherwise even if you have like a statistical model that again
You could throw in PCA factors meaning like you know factors that are grasping at observed correlations even if you don't have a name to that thematic that's happening
effectively
In numbers, but um, but you could you could even grasp at those at that kind of statistical exposure
But from a practical day-to-day standpoint if somebody even somebody has relatively high turnover like let's say somebody has
15 turnover or 12 or even 20 turnover
You're not trading at a frequency to to be in and out of names where you know
You're gonna necessarily want to have the risk model constantly change up on you every day and then in the one-day
Extremies that which does exist by their commercial models with one-day factors which can be useful
Again if your horizon matches that if you're trading in today
You actually might really want to look at one day your one-day reversal
Kind of factor, but if you're
If you're trading over the course of a month on average it was 12 turn over in a year or if it's even half a month
Some of those transient thematic some of the PCA driven factors that can just flip on flip off
You know aren't going to let you sit there and do the hard work of
Okay, well what data what what do I think about what's new about the world? What's because again, I think we're a lot of
um fundamental investors are demonstrating their ability to have positive alpha-hitting where they have skill
It's actually not I think there's always this hope in a funny way. I've seen it myself. I've seen it with others
You know that you can build and you can there are places where this happens right where a fundamental type of
Investor will build sort of pseudo statistical arbitrage like oh when this happens this happen when this simple idea happens
That simple idea happens then you always got to be long this or that and I think it's rare for those to
Consistently succeed on a back test in a way that is that is really monetizable. I think what what the
Key area where people generate positive P&L are are predicting something that is changing about the world
Something that hasn't exactly now can rhyme
It can have a correlation to events and again where I think the best people that I've seen do it is they have those kinds of rules
Those kind of connection points and they marry that set of heuristics with what happens when
You know um the refinery the when capacity and refineries is doing this and chemicals industries doing that and rates are doing this other thing
Um, and the factor model says this the and then there's a new thing that's happening
Either technologically or in the model or in the management commentary and you're marrying that set of like
Heuristics with what are the general rules about the world with
But here's what's changing in the world and here's how my sort of mental model of how that change will overlay these kind of rules and insights
And when you're pulling all those together
It sounds awfully complicated
I guess but but that's that's where people are able to generate positive
You know alpha hitting and those are the people I've seen do it the best is people who kind of overlay all those pieces with
Something that change because again no quant is going to arb that
Totally new environment. Here's what how I think about the new environment
I mean the example the recent examples all of the
AI supply chain problems that people have been working out and thinking about and hearing and thinking about the
consensus and the crowding and the market
Interpretation of that because at all times you're also competing with
What does everyone already think you know, and so that layers adds a whole layer of difficulty to the whole game
So that's kind of how I think about that. Oh 100% and I think that the factor risk models
Like you said are another variable for a fundamental to fundamental PM to better understand the world
they
can build
Not just intuition around what's going to happen in the supply chains, but also how the factors move as well as that
And I think what you've laid out there and very very well is that
the top PMs at the top hedge funds
This is how they think they're given all the best tools all the best research all the best talent
um
The best risk models to
Understand the world at a more granular granular level and then make bets on top of that
And I think in some sense what you're doing right now is kind of
Giving smaller PMs or given PM giving PMs that aren't at the top hedge funds or maybe you have clients there
um
These tools
To see the world in in a similar way. So I guess could you talk a little bit? What is our corner? What do you guys do?
and
Yeah, what problem are you trying to solve? Yeah, for sure um, so yeah, no we actually work with um
most of the largest hedge funds in the world today um uh like
You know um something like
Seven of the ten largest multi-managers in the world were in like six of the trillion dollar plus asset managers
Our large deployments like a hundred and eighty-seat. We have we have a few hundred
Even larger deployment kind of in process of rolling out all the way down to
single-seat new launches spinouts things like that um
And we're in you know something like a reasonably large fraction of our businesses in that kind of
hedge fund very intense very focused on the kind of issues you mentioned
That's a large fraction of the of the business, but it's but you know
It's it's less than half the other you know
Maybe 30 40% of the business that are tiger cubs and deep value funds and long-only is in firms that are
Increasingly adopting um this kind of way of thinking and I think there's a and then another 15% are allocators both asset owners
lots of the sovereign will funds and other firms in that kind of category and um and then another
This whole world of kind of platform SMA platforms that have kind of come into being we can talk about what's kind of a sub trend?
What happens is happening in the world from a
capital raising standpoint?
But I think there are two or three overlying
Trends that are happening. I mean one is from an allocation standpoint the entire world is headed to
Kind of beta one or beta zero products
I think we could talk about why that is and come back to that the other thing that's happening is just this broad
hope that
the financial industry as an investor that you can have more excellent tools about to analyze
everything from. I think there's been the muscle of that sort of fundamental, deep value,
deep research stockpicking has gotten very strong. But the muscle around risk management portfolio
construction is just beginning, it's sort of in the beginning stages of it being really adopted,
thoughtfully by the entire industry. I think what's happened in the last almost 20 years, 15 years,
has been a vanguard of a handful of people. It's not just one firm, but it's probably, you know,
20 firms that have really embraced this kind of way of thinking about X-anti correlation and managing
that on fundamental investors. The whole quant industry is fully adopted on this way of thinking
and has been for since the dawn of, you know, since Rentech and the whole thing was always about
portfolio construction, the alpha associated with portfolio construction. But translating that
into the fundamental realm where, you know, I think there's this, you know, Warren Buffett,
and I'm a Buffett stan, like I've read every book, every letter Warren Buffett has ever written,
I started out as a total deep value person and fully aligned with lots of the ways of thinking
about the world that kind of implies. But I think this new model of thinking about performance,
decomposition is just to say we should be thinking about everything at the level of, like the most
fundamental person, if you said, okay, why is a stock up in the past year? You would say, like,
if I pulled up Microsoft and you'd say, okay, why is Microsoft up or down in a five-year span or
whatever it is? If you just said, okay, you could narrate that chart or something like that and you'd
say, well, why is Microsoft up or down? It would be totally remiss not to look at what happened to
the Q, the NASDAQ in that period, right? Of course, if you put those two things on top of each other,
huge fractions, 60%, 70%, you could debate the factor model on which one you want to pick,
but is explained by the basics of what happened in the technology market over any span of time.
And the more fundamental question would be to say, well, why did it deviate from that sort of
benchmark at any point in time? Why did it deviate up? Why did it deviate down at any point in time?
And so if I were to ask an analyst, like, why did Microsoft go up and down, like, do you understand
this stock? Sure, you could say, why did the stock price move? But much more fundamentally,
you'd say, why did it deviate from its correct benchmark? And fundamentally, by the way,
what a factor model does, stripping away the x-anti-correlational, the sort of quant framing of it,
more, you could also frame it in terms of benchmarks. All it does, so Microsoft first accused,
maybe that's a fair benchmark. But if I said a small cap, because I said data, maybe data dogs
large enough, but like some really smaller cap technology stock, and I said, okay, we'll compare
that to the to the to the cues. You might say, well, it's a small cap. The cues are a large cap,
you know, and here's some other, you know, it's it's more of a value stock within technology versus
this is a more high growth company. And so you said, well, the cues aren't a perfect benchmark.
And again, all the factor, factor models do is effectively create a perfect bench, a sort of academic
or theoretically perfect benchmark for every stock at every moment in time, and model everything
relative to that sort of theoretically perfect benchmark. Because of course, we can say,
Microsoft first cues, reasonably fair, whatever. But of course, and certainly if we said
this small cap tech stop, what if we said like a, you know, a an Indonesian steel mill? Of course,
that shouldn't be come to the to the cues, right? So like there's a different, more perfect benchmark
than any broad based benchmark, of course. And that's effectively what factor models do is they
create a benchmark for every and then they do all this downstream modeling, having built the perfect
benchmark for every stock at every moment in time. They do all this downstream math on what that
implies about correlations across names that have those same attributes that have non-overlapping
aspects of the benchmark, right? But so again, there's this kind of framing that okay, if I was trying
to say I really understand a stock, why did the stock price go up and down? But really what you'd
want to say is, why did the stock deviate from its perfect benchmark, you know, sort of attribute
perfect benchmark from being a large cap tech, you know, name with this kind of value value structure,
whatever it is, why did it deviate? And those deviations are what explain its idiosyncratic,
more fundamental, deep stock research, the management did this, its strategy was that it's product
line did the other thing. That's kind of, so I think what that whole side of the world has,
has sort of come to the conclusion of is actually there is all this downstream portfolio construction,
aggregate PM modeling benefit from this kind of way of thinking, but actually the more fundamental
way of looking at stocks is actually to look at why did it idiosyncratically deviate from its sort
of, you know, sort of attribute driven perfect benchmark. And it actually reframes how you think
about, well, what KPIs matter to some Microsoft? What narratives drove the outperformance or
underperformance at different points in time? And so as you're trying to say, okay, well, why did
it stock? Do I understand, as I'm trying to initiate on a stock? The best firms in the world will
say annotate the idiosyncratic return of the stock, not the price of the stock to understand what
KPIs drove, what movements over time, for example, let's just one example, but there's tons of them
that say, so again, I think what's been and what's kind of secondally changing is that reframing of
actually this kind of modeling has both the downstream quant kind of aggregation of correlations
benefit to a book, but it also actually lets you elucidate what's fundamental, what's benchmark
driven and keeps us intellectually honest about analyzing stocks, portfolios and stocks is kind of
this, to some people is mind blowing actually still to this day. And it will be probably in five
years as a group of people for whom that's like a totally, and there's a group of people for
whom that's just, I don't agree with that at all. I just, what do you mean a stock went up? What
are you talking about, right? And that, you know, there's the, there's that kind of die hard,
fundamental only person. And then there's this camp of, hey, look, everything matters at some level
in terms of understanding how to how to think about stocks. And so what's happening is that way of
thinking about decomposing returns of stocks is moving from kind of the back office where that was
sure there's these guys modeling this stuff and sort of creative interesting way. All the
quant funds seem to find some utility for it in risk management, but, but I'm not sure that
matters to somebody who's looking at, you know, what's NRR on some software business that's
growing, you know, 27% a year. But I think what's, and what has changed in over the last 15 years
and is sort of rapidly being adopted is, hey, this is actually the way to, that we need the front
office to be aware of that interplay between my stocks and what their sort of, you know, factor
relative performances, what their exposures are, are we meaning to take these huge factor swings
that are implicit sometimes in large portfolios that we take. And so as that is shifting from this
kind of back office risk CIO type office thing, it's now becoming core to, you know, whether even at
non, you know, kind of pod type firms, but at every single long only at every single, you know,
kind of at many of the tiger cubs already and that's growing. If you think of that school of
thinking, certainly at the long only at the at the each type at that beta one and beta zero,
if it's what's going to be, you know, just driving if you're hired fired, promoted, demoted,
those tools need to be excellent. And in the front office's hands and serving them as opposed to
like, hey, somebody built the tool internally that it works some of the time or it's great. It's
some ways, but you want some degree of, hey, we work for, you know, we build an excellence around
that that kind of toolkit. So I think it's the combination of kind of arming everybody with that
awareness of all those pieces in a way that feels native immediate intuitive. And then also,
just this great desire for, you know, of the new era of investor to be very analytical about stocks,
to not just sort of take quotes at face value, not to take price movements at face values for people
to say, okay, why is the stock moving? You know, how much if a stock, if we said Microsoft is,
and but take any stock and video anything is 60% factor driven, 70% factor driven at a stock level
in certain your portfolio is probably even more. You know, if you just sort of simplified that way,
like, okay, my book is down. Why is like the number one motivating question of Wall Street? Some of
its news flow that's idiosyncratic to the stock, but a huge fraction is factors and flows.
And being able to model that, see that precisely intraday live over a longer sweep to
over 25 years, whatever it is, that, there's just sort of this demand and desire for people to be
analytical, thoughtful, intellectually honest across the stack of like fundamental type thinking
of single stock research all the way to, you know, portfolio construction, performance analysis,
all those pieces. When there's an old school PM and he's never thought in terms of factors,
you know, he, he sees Nvidia up and he's, and he's like, oh, you know, he, he has all the reasons
in his head, right, and maybe the AI factor, whatever is, is up big. Same day. And so most of that
variation is explained by that, but he goes to you and he says, Rich, I think that's bullshit,
you know, this is, there's no way, you know, the world is simple. What do you say to that?
I think the reality is, you know, it doesn't, you know, the, like old school, new school, it
doesn't cut across quite that cleanly in that we see, we see people who are, in some ways I find
actually that the person, if you said old school, and maybe you met it more stylistically,
but if people have seen many market cycles, often the survivors of that are those who keep
adapting themselves to the environment that they find themselves in, right, to the market as it
changes, right, the person who says, hey, look, I was like,
you know I mean obviously markets change right there was an era we're just getting the
filing before everybody else was a source of alpha that obviously eroded pick your time for
him uh go right and so if part of what you meant is in ten or you know ten year at at investing
and seeing cycles often I think that people who survive best for those who kind of adapt
themselves to the environment and what the best evidence shows right like you couldn't you
wouldn't have been you wouldn't even have the tools to really excel at this way of thinking
20 years ago 10 years ago right probably five in some ways um and now it's at your fingertips
instant right if I want to see this decomposition in this way that I'm saying you can do that
instantly and then you can close the window and do your other stuff too by the way right like
meaning it's not all consuming it like in an earlier era you might have had to like build all
spreadsheets or write it in code and there's all kinds of things you would have had to do to make
this even practicable worth today um a lot of this stuff can be immediately at your fingertips
excellent insightful you can tweak it you can debate some of the assumptions to lots of the
different pieces well what if I tweak this and change that and what would that imply about my
about my risk and exposures and so um all those pieces I think are you know conspired to make
people who want to be adapted which I think actually most not everyone of course you're
there are definitely there's always diehards writing every in every category of everything right
um I'm investing um for sure um but um but it isn't like that I think it's a matter of degree
and it's about you know I think a lot of people are will acknowledge because it's a fact I mean
the truth is I don't think there's anybody who would doubt the concept of benchmarking right like
of course like someone can get a return stream for free um you have to at some level beat that
free alternative right um and there's all kinds of mental gymnastics you can do and and should do
um and it's funny because um you know that kind of the correct benchmark is so important meaning
at the level of granularity that we were just talking about but at the level you know you mentioned
earlier like oh there's retail people who look at whether it's any of the the major pods like okay
this one any of these big pot you know multi managers generated 10% last year that's terrible compared
to my I was long Tesla what are you talking about these guys 500% I'm better than and looking
and by the way some of them are like no no pride of of authorship at all like I'm not saying
you know your local head fund person is some you know gods gift to to to insight like they get
around all the time people get um stories well ahead of the market it's not it's not impossible
to believe that it well we're talking about is is not like can that happen right or can it happen
like at one insight or one kind of thing you know but but it's like can you productionize it into a
business where it's a product you can sell at scale distribute you know what I mean like we're
talking about like when because again um it's not like there there are many cases where the retail
insight is more correct and has been I mean look at probably Tesla for its longest period of time
right and then you could debate the nuances and why and all that stuff but there are tons of
examples where um you know the market you know the the sort of when I say the the hedge fund crowd
isn't perceiving seeing the ball correctly all the time right um but it is to say that what you're
solving for is is there something you can build at scale that has a sort of productionize ability
that has a sale ability like hey this is an attractive financial product that you should put
you know lock up money with for a long period of time that's going to do right by typically the
allocator is somebody allocating pension fund public pension funds endowments sovereign wealth funds
that's ultimately who all of this is serving is like is there a risk-adjusted product and to your
point earlier you know a multi-manager kind of market neutral product isn't really competing with
an equity index as a product I mean it's sometimes every dollar is competing with every other
dollar but if you it's closer comp is going to be maybe it's fixed income right with a different
maybe tax profile or a different kind of market profile probably to some extent but um but in
terms of the return stream you have to you have to comp it correctly right for and then of course
there's this well different asset class comparability question and all those things but um
but yeah so that's getting the right benchmark in every sense is it been so back to your question
like I don't think there's that many people who would at least um intellectually disagree
like oh we should be benchmarking everything appropriately and we should be thinking about our
performance relative benchmarks and I suppose sure we should aggregate the those things and see how
those things add up to big unintended exposures in my portfolio right that makes sense um and it's
just a matter of um it's almost like a matter of almost energy like how much energy do I have to
apply this to my process today like I'm looking at five things I'm doing research in this way like
do I have the resources do I have the tools yeah on that actually um how do the best pms
incorporate it into their process so of course it's a risk management tool but you built some
intuition around it um you start to feel how a different factors move you start to select bets
that you know are idiosyncratic in nature and aren't just betting on AI factor or levered
you know x amount um and so from the people you've seen the pms you've observed and spoken to
how does it fit into their process when they also have to juggle you know macro view um supply chain
all these different things at the same time yeah for a lot of uh people it's a part of a it's a um
kind of requirement in a process either ensued by a firm or by themselves to say there's certain
sort of non-negotiables about how large I'll get in any one of these kind of factors because the truth
is um there are some factors like people will say this about crowding the kind of by word is um this
is a it's crowding and you can debate this too but um it's a positive return low sharp factor
and that's to say that again it might be it's correlational and almost um almost
necessarily I say almost because like even a stat arc strategy has um is is effectively building
factors and those might have positive sharp but those are such subtle factors such nuanced factors
right like the broad brush factors the widely knowable factors the momentum the um are not are in
the best case in some of them are are zero or negative return over the long term um and have been
or in my chain of course but um they're like the market in that sense and that they um you know
might be positive return but they're not positive sharp ratio structurally um they will create a
drag to your uh to your sharp ratio if you're producing a level of sharp ratio that is you know
contributory to a firm beating inaccurate um and so um so practically what people do is they
often have kind of constraints and what I've seen really um people who I think kind of nail the
model do is they think of each factor bet as if it's a stock bet and they just sort of say hey
whatever limits I put on single stock exposures I'll put that same kind of limit on a
fact about and I'll sort of almost diligence it doesn't mean you spend your whole day thinking about
you know how's momentum doing but it is it just mathematically and just true and in truth
it is going to contribute to your PNL to your career you know you're piring firing promotion everything
as much as any individual stock bet you have and so they'll treat it both in terms of applying
limits on their portfolio construction and in terms of the diligence they'll do um they'll sort of
treat it like it's a stock position they won't ignore it but they won't over index to it
they won't spend all their time researching what do I think is going to happen to this or that
factor but they will think carefully about okay I've got this exposure in my book here's what it's
done historically here's what I think it might do in this environment here's what I think about
those things and sort of treat it like uh treat it like a view um and I think that's the most
like prudent rational perspective um because again keeping strictures on it and it varies by
firm and it varies by structure what percentage of your vol or your forward-looking volatility is
driven by factor bets versus idiosyncratic stock specific bets like the most intense kind of
limit limiting firms will kind of have people live anywhere from 80s to 90s percent idiosyncratic
like of the volatility they take is 80 to 90 percent stock specific and any silver major of 10 to
even 20 percent that is factor kind of driven some firms will let you go down to 60 percent idiosyncratic
and 40 percent factor bet um sort of thinking um in some ways that they'll manage that at the top
co you know in some ways and and they can either manage via factor baskets they can manage via
all kinds of strategies they can just find pms that maybe have inclinations to those factors that
are opposite of their typical pm like if we said most people often will want to be long a similar
set of factors of today that's you know today or tomorrow that's semis or this part of the
technology value chain whatever it is and um and momentum and whatever it is you might have some
books that are kind of structurally against that right that's in the hedging at the top top kind
of level and so structurally what people are doing is kind of managing to either internally
imposed or externally imposed kind of limits and thinking about the exposures just sort of
analytically like what are the pieces what am I exposed to what are the components what nuances
could I be missing what aspect of my momentum is going to be somewhat you know subtly correlated
to other risks I'm taking in the book um and then treating it like a like a position it
fundamentally a lot of what these tools do us included is they they make
it all more tractable more controlled more logical it makes the whole thing instead of a guessing
game a speculation uh oh uh you know I think that you know momentum is going to contribute this
and that it's like all right there's a number here I mean you know and there's debates on
but we're the debates on precision are significant digits different questions that at a different
level from um you know from like that sort of speculative it's like it's like it's like this or zero
right so definitely I've heard I heard somebody say this that as they uh a friend of mine who
who had worked at a kind of tiger cub
and stepped into a more pod to multi-manager type C.
They felt that they, as they looked at the toolkit
that they were sort of playing a video game, actually,
that it felt like investing in stocks had gone
from this sort of, yeah, those goods and bats, this right,
but I think, but I think this person was framing it as good,
that it sort of felt like they were almost on a quest,
an intellectual quest of right and wrong
in the previous job where it was like,
this is right, and sort of, you know, this person described,
they were almost sort of blocking out the results,
the, you know, the outcome, the accountability, whatever,
and that as you stepped into that multi-manager seat,
you sort of had this toolkit and the system
that made the whole thing feel like a video game
to some extent, like, okay, I'm, and what that meant,
what they meant by that is like, I mean, control of the levers,
I can understand the drivers of my risk and exposure,
that doesn't mean that it's easy,
it doesn't make it easier at all.
And in some ways, imposing that kind of stricture
of accountability, clarities, truth, you know,
raises the bar, right, on, you know,
are we actually generating value on average
in a way that's provable?
'Cause you can also think about a sharp ratio fundamentally
as a teased statistic against the mel hypothesis
that you don't have skill, right?
And that's not a new framing to a quant, you know,
or a sample teased statistic, right?
But, but to a fundamental person, that's kind of a,
'cause again, the sharp ratio is mostly an attribute
of how high quality the product is to the investor,
to the end investor, or the intermediary to the end investor.
But it's also like a statement of how certain we are
that we're generating returns materially different
from zero, right?
And so all those things, again, like, what these tools do,
is they sort of put the world in a tractable standpoint,
let you tweak it, play with it, poke that risk,
and say, okay, what if I change my positioning
from this to this, what if I run me an optimizer
that says, okay, give me a more optimal book construction,
like on the fly, let me see, okay, well,
if I didn't want to run a full optimization
that rebounds my book, just give me new ideas
that would kind of hedge my book,
or actually look at the entire S&P,
look at my entire shorts candidate list,
or I've got, you know, two universes that look like
they could be attractive, but I'm not sure,
like, run it top to bottom.
What would most hedge the rest of my exposure in my book,
like, give me that answer in an instant?
That's what these tools can do,
is kind of cut the whole guessing process into instant, right?
And so that's, and then translate that into performance,
like, okay, how did that show up?
What did that turn into?
Like, did I succeed at sizing differently,
or it stares at you on paper?
And so there's, you know, in some ways,
it kind of reduces the artisanal quality
of investing, in a way, right, the, like, sort of,
oh, I put the hand worked sourdough, or whatever, right?
It's sort of, okay, there's a game here.
There's a, there are points, there are level ups,
there are level downs, there's risk and exposures.
And then, and then there's the really hard part,
which is generating positive alpha-hitting,
and that is so hard in practice, right?
That, and it involves all those nuances
that we talked about, that it's crazy to me
after that, if you have that,
or if you put so much effort to achieving that,
to be handicapped at the, to leak P&L or volatility,
that is not associated with your work,
all of this work that you do,
generates essentially no P&L from a factor stamp,
and all the factors do is drive drawdown, volatility,
I mean, all of the drawdown, you know,
volatility is drawdown modeling,
and drawdown modeling is correlational modeling,
unless you have high concentration again,
that's a slightly separate point, but any all drawdowns
are fundamentally correlational at, at breadth, right?
Or at scale.
And so, that's really it is how,
it's, it's amazing how hard it is to generate alpha
in general, and to sort of do that,
and then, and then just let, like as we said,
this at the top, the failure vector of like pick your,
pick your famous, we don't have to name anybody,
but pick your famous, deep value, fundamental,
or any type of investor who has had a lot of success
and had some really obvious demonstrations of brilliance
and insight, and then when they blow up,
whether it's equities or other asset class or whatever,
the story is never, you know, shucks,
I actually am not as brilliant as I was before,
and maybe that's just what they tell themselves,
but I think mostly it's not that they had a,
that they're fundamental quality of thinking
to deteriorate so greatly at all.
Like I think it's the same person,
there was flashes of brilliance in that person,
but what they got wrong was risk, portfolio construction,
you know, kind of being analytical
about the other half of the portfolio.
And so I think that's, you know,
that kind of perspective that I'm reflecting
is what the market is saying is saying, okay, yeah,
we, the thing we can do is isolate people
to find the thing that's really hard,
that's really special is finding ways to analyze
what's different above and beyond the sort of classic
factor exposures or classic,
and let's let people do that hard work, you know,
and not, and not leak P and L that isn't associated
with any of that.
- Oh, 100% you're allowing them to isolate their pets
and really, and I just, at the firm level
and all the top hedge funds do it,
they allow the PMs and the people who sit in the center
book seat to see how skilled the PMs actually are, right?
And you have a PM and he, it's a tech investor,
software or semis or, I mean, if he's in semis
and he's, and he's up big and there's no risk model
around that.
I mean, there isn't much, there isn't a good way
to measure how differentiated their thinking is.
I guess the pod shops, a lot of them, you know,
big hedge funds have you as client, you know,
use your product or rather, what do the smaller
or like long-only shops, what use do they have?
I mean, I guess they're long-only, they're, in my mind,
they're not trying to hedge anything out, what's the use?
- Totally, so really at some level,
so two different classes, there's kind of super concentrated,
whether it's long-only or long biased,
like somebody who is really in a very concentrated portfolio
and we'll talk about the uses there.
And then there's the kind of benchmark relative,
long-only person who is trying to make, trying to comp
themselves to a benchmark and trying to make active,
you know, everything is in active terms.
And in the second case, the active benchmark relative person,
kind of fundamentally, they are a long,
they're a, they're kind of a market neutral short,
it's a benchmark, I mean, every position, right?
You are taking constant, and somebody's a greater breadth,
depending on your relative sizing, across the book,
but you're kind of a market neutral against a benchmark,
like truly, like you actually, and what's funny
about this stuff is you asked about like,
different frameworks and styles,
maybe surprisingly to some, like the market neutral person
is like, fully now, conceptually and in many,
in most cases, practically adopted on this way
of thinking about things, and it's sort of expected
that you would be, and surprising if you're not,
in lots of cases, but the second most kind of adopted
on this framework, thinking are, you know,
benchmark relative, long-only people who want to at least,
'cause they, 'cause everything, again,
it's the same exact question.
Why, what is driving my performance roles as benchmark?
How much of this is factors and flows
versus idiosyncratic ideas?
Every moment of time I am implicitly short ideas
by being underweight the benchmark,
just mathematical, right?
That is what I am.
I'm running a kind of market neutral book
against this benchmark effectively
by my, you know, overweights and underweights.
And so, you know, they,
each person emphasizes different pieces of this,
like, right, whether it's the sort of risk
and exposure, understanding and choosing
what I want to be long and how big I want to be,
bets that are not kind of stock-specific
in idiosyncratic, the second category being iterating
on portfolio construction around that, right?
Like, what, what if I do this?
How do I optimize?
How might I find hedges?
And the third category being seeing performance decomposed
through the lens of this stuff
and then translating that back into that kind of feedback loop,
okay, well, what could we do better, et cetera?
So people focus on different amounts, right?
But, no, and then, and then,
so that's the long, only person, you know,
increasingly is just living in exactly that same mindset, right?
It's done with, I suppose, a little bit less leverage,
but when it's always a bigger scale, a dollar scale often,
depending on the place,
but, and then there's kind of a different bucket,
which is, you know, that kind of very concentrated person.
And for that person, again, almost paradoxically,
they're taking the biggest factor bets,
like even though their kind of discussion point, right,
is hey, we're just, we're pure fundamental investors,
like just like your company fundamentals.
Again, in the reframing of it,
they actually have the biggest factor bets.
Their books are squared to factors is actually highest.
It might be 80% because they're not
comping to a benchmark and they're not, you know,
building a kind of, uh, hedged portfolio.
And so they have the biggest leverage to each of those factors.
And if you're going to take these huge factor bets,
you ought to be intentional, thoughtful, aware of those pieces.
And then, so that's kind of on the sort of,
what you mentioned, things like risk, portfolio,
construction factor exposures, and then maybe this kind of relates to the rest of what we do,
which is the combination of performance decomposition and process integration.
We do all of this enormous RMS business, research management process business,
but meaning you can take in conviction scores and price target and see, did our highest,
again, tie into performance, that our highest conviction names produce our best returns.
Did we have sizing skill on our highest conviction names, actually?
Or did we actually done much better, had we just sort of bucketed all of our highest conviction names into a single size,
instead of, hey, our one idea is what we should be kind of in three times the size of the next guy.
Reality, we actually have only skill at sort of differentiate, high conviction, low conviction,
you know, as opposed to number one, number two, number three, right?
We're actually really not that granular at our relative judgments.
So pulling in all those, and so the, and maybe even a third leg of the stool,
what's happening in the world is, it's part of this AI theme.
It's part of everything that's happening, which is the sort of just build a model,
do the kind of painstaking, you know, labor aspect of the analyst job is declining pretty rapidly, right?
And what the expectations are is that you're hiring an analyst to become an allocator of capital relatively soon.
Whether that's called a PM, or that's called a senior analyst, or whatever it is.
But the best firms in the world, actually both market neutral, long only and everywhere across the board,
are saying, I've been doing this by the way for like 10 years, right?
Again, and it's not, they're all kind of related, these kind of trends,
because it's really, are you being, are you insisting upon being intellectually honest, clear,
having objective criteria for success and failure?
They all reduced to some, you know, probably McKinsey deck of what it means to be excellent at operational excellence
and transiting that from taking the investing world, which sort of had this, as I mentioned,
this, my friend said, this sort of deep value questing aspect, which is true.
And I love that stuff, right?
I've read, you know, every, every, you know, I've read the ironhorn book,
and I've read every word, Warren Buffett, and I love that you have to.
I think if you like investing, you have to be, you know, like this kind of value nerd, right?
And love this kind of thing, but it's also important if you're selling that as a product to, you know,
the kind of long-term investor base that is, that is trying to produce a consistent return stream
for a set of people's retirements, right?
That packaging of that questing is great, but you have to sort of turn it into a systematized process.
And again, but so back to the, the previous point, what people have been doing is sort of that,
just sort of analyst as number, enterer, and compiler of information is, is reducing.
And what everybody is seeing is that, that analyst hire is meant to be allocating capital,
relatively maybe earlier in their career, or is, or is it least on that track to the purpose
of an analyst is to be making active investment decisions, or at least contributing,
that doesn't mean they're pushing the button necessarily, but that they're having sort of P&L-analysable ideas, right?
And again, that a lot of the best pods in the world will have, you know, analysts,
trackers, and mock portfolios, and we enable that.
That's like a big part of what we do as well as analysts have mock portfolios as soon as firms want them
to idea trackers, things that characterize, am I long or short, different ideas?
And is that generating skill?
And even if that's not at the early stages of a career dictating your, you know, what you take home,
it's at least giving you the feedback loop of, I'm keeping me honest of, okay, got it one day.
Hopefully in the not-too-distant future, I'm going to be accountable to these outcomes, right?
And I can't sort of lie to myself, lie to my PM, whatever.
Like it's the most classic thing in an investment firm for, and I've done this myself.
So this is said with love, right?
But like for an analyst to point, I said it by here, and it was, you know, but then when it went down,
I said to celebrate before it went down, like that sort of, you know, that whole thing,
which is actually endemic, I mean, it's a big part of a lot of investment.
That is, that is also dying in a way, because people are, you know, giving the tools across the board
to say, okay, well, let's keep ourselves intellectually honest across the stack.
And those are, those are kind of trends, which is, let's be more analytical.
Let's give, let's empower people to make decisions that are accountable earlier.
Let's give them the feedback loop earlier.
You know, it's just like, you know, it's, it's giving you, you know, I think it's my elementary school,
they'd give us test grades, but not final grades until some year of school, six grades,
seventh grade.
Oh, that was a good system in some ways, right?
It's like, you don't need to have this final mark on what feels like at the time, you know,
a kid feels like a permanent record, but, but it can be a good system to begin to get you.
Okay, at least on the test thing, you know, loosely where this would head, right?
And so, but that gives me the opportunity to, you know, begin a feedback loop.
Okay, well, I'm bad at this.
I'm good at that.
How do I get better at this thing?
Or this was random chance.
And I don't need to worry about that in the long term that'll work itself out.
But that, you know, beginning of that sort of just intellectual honesty, that processization,
I think the tools to kind of do that full stack of stuff are just getting so much better
than they used to be.
Whereas previously, I might have taken just tons of resources to build that out.
Now, you can kind of get that instantly and have that sort of best in class quality immediately.
You touched on it there with regards to junior analysts, it's a new world, and then
also some other tools you provide so that these junior analysts can get that, I guess that
feeling of a test, that managing a paper book.
But I want to ask about your take on the future of junior analysts, given that you work
with so many top funds, so many different funds, different strategies, and being reasonably
integrated, that quite integrated into their process by selling your product.
Are junior analysts screwed, you know?
I don't think so, I mean, I definitely, I definitely think it will be different as all
things will at every change of technology, you know, well before any of the current waves
of AI integration into process and AI integration to everything we do, right.
All of, but this was true at every revolution, it's the job will change, right?
Constantly, as, you know, like I think Lloyd Blankfine talks about, you know, writing
in the bond, the prices into the desk, I mean, you know, like in the early days of bond
trading, like paper slips, like his job changed, but, and I'm sure what they were looking
for then and every firm today is looking for is somebody who can in the end generate
that kind of positive alpha hitting rate, that's exceptionally hard.
There is not enough, there is way less supply than there is demand for that person.
And so what people are looking for are people who can ultimately generate or contribute
to generating or have an input to generating or, you know, any part of that stack that
are adding to the job of adding positive alpha hit rate, that is a scarce resource, not
that it's not that it's zero, there are, you know, there's a, like an entire industry
of people who on average actually seem to do it, not just seem to just mathematically do
it today at least.
And that's a very scarce resource.
So that pipeline of looking for talent that has the capacity to scale into becoming that
is super valuable and exactly the path, exactly the kinds of programs, exactly what
their job will be day one, exactly how quickly they get to mock books, exactly how quickly,
because I think one of the things is, you know, that trend is to more quickly say, is this
person going to ultimately be able to get one vector is to give a mock portfolio, and
you might, there's no harm to that for sure.
But there's also so many different ways people can, and like evaluate that person's ability
to get to that point and every investment firm has a different philosophy that some people
might, yeah, quickly kind of think, hey, let's begin to look at track records of, was your
idea good two months into your investment job.
Some firms might say, hey, look, I just want to see it, can you do great research?
Can you do, and I fundamentally believe that if you do great research in the limit that
will translate to, and by the way, that philosophy is true, like, there are, there are firms
that are the best firms in the world that say, hey, I want to pretty quickly see if you're
generating, if your idea is roughly have positive directional signal, and there are equally
excellent firms that take that second approach that say, hey, I don't, I don't, yeah, I
certainly don't believe two months in how could anybody, right?
I really want to see, can you generate excellent insights and research and find what we talked
about, that kind of interaction term, that, that, that fourth level of insight, that that's
pretty rare to be able to be all that, you know, compilation of different insights, and
then right.
I mean, that's so hard.
If you can do that kind of a thing and show that you have that ability, regardless of
which direction you kind of came out to, if I, if I see that, I can train that into somebody
who can generate positive, positive kind of P and L, however you define it in the long
term, both exist.
But yeah, I think I think the, that's a scarce resource that people are looking, looking
for it again.
But again, I think this world of, of all this stuff, of AI and, but it's not just right,
AI is a huge enabler, and it lets people have more leverage day one.
You can build all kinds of internal trackers and tools and, and all kinds of things leveraging
all of the tools that exist, there's also just better data everywhere, all of the different
datasets, the accessibility of that data, our tooling, others, you've had some on the podcast.
And then also just the, your podcast itself, right, and just sharing the knowledge that
all these things have you and, and the rest of the industry.
I think there's just a greater ability to,
to more rapidly level up into the types of knowledge
that kind of brings you, makes you comparable
to the person who's been around for 10 years.
I mean, I think the slope of somebody's really talented
and intelligent and hardworking,
to get to the sort of knowledge base
that a senior person has is just faster, actually.
So in some ways, I think, as with any technology,
it scales the best, it probably creates dispersion.
It's a put it one way between top and bottom.
And I think that's the case.
More than it is, like, new analysts or anything like that.
- I agree with that.
And I'm saying new analysts,
not that there won't be analysts at all,
but it's that, you know, that 1th percent,
like, you know, like certain 99th percentile guy,
or even 99.99th percentile guy.
He's just a machine, or she is just a machine.
You know, it's just all the tools feed them.
They're able to build up this corpus of knowledge,
this differentiation.
And then they're put into positions
where that can expand rapidly, right?
Whereas the median talent, what is he or she do?
You know, for me, and I see this just with friends
or, you know, junior, very junior in their career.
I see the dispersion.
I see the, the modes that my friends who are,
who are most intelligent, I see what they're doing.
And I'm very much long, but for, for, for the median,
for me, it looks extremely bleak.
And I'm just curious, how do you think about that, you know?
- Yeah, I don't, I don't necessarily agree
with extremely bleak, but I do agree with your overall point,
which is, for sure.
I think, I think every technology without strictures
which can come from all kinds of sources, but it's,
you know, expands dispersion.
I think it, it will make the, the top person.
And by the way, the technology stack that exists
to do that, one of the vectors is AI,
but other vectors are the technology of,
I mean, capital markets themselves,
is the expansion of the ability for young people
to get large pools of, of hedge fund capital.
I mean, this whole SMA world is enabling people to start
market neutral type firms more easily than ever before.
And so, and again, not to say that anything is easy,
but just that, that there's a large pool of capital
that is looking to allocate that.
And that itself, that is a capital markets technology
that, you know, gives the rewards to the top person
as compared to, and that's just one vertical.
But, you know, there are other,
people who've raised capital in different sources
that are, that are young and, and, and doing well.
And it's not a, as you, as you're kind of pointing out,
it's not necessarily an age thing.
It's a, every technological revolution
had increases, is tends to increase the ability
to have some dispersion at the top
and empower the person, but, you know,
but I do think there are also, like, limits to,
there are also frictions, right?
Meaning, somebody's coverage universe
while it can expand,
while you can go deeper into the, into the coverage universe
in terms of breadth and a portfolio by a single,
really excellent investor, you still can't get that level
of positive hit rate at, you know, some number of stocks.
You know, like, there's going, there is a limit.
There's a friction point as in so many areas
where there's 80, 20, or maybe if to certain technologies
enable things to be, you know, higher than 80, 20.
There is this kind of, you know,
and even if they slightly increase it,
there are just these frictions that limit,
it kind of be investing as a, as a profession
from being winner-take-all.
The way that some other areas are winner-take-all,
or at least niches are winner-take-all, right?
And so much of business, but investing
is a much more fragmented market.
If you just analyzed it like any other business,
they'll just do it like a Porter's five forces
of the investment business.
It's way more fragmented than so many other areas, right?
Even within the niches, like, you know,
if you took any technology area, sure,
I like, if you define technology,
it really brought this a ton of different.
But if you define any, like, CRMs or something,
like the market shares stack up in a,
in a pretty clean, organized industry structure kind of way.
And of course, new revolutions can upset that
and all that good stuff.
But the investing business doesn't stack up that way.
Like, even the biggest hedge funds in the world
have a tiny relative pool of capital, right?
Now, there are big, in, like, sort of ETFs
and long only's, there are, there are greater shares,
but even those, when you kind of compare those shares to,
you know, the total pool, it's not that,
it's not that enormous.
It's not nearly as the sort of herfiendall index
of the hedge fund business, somebody should check me
on the, do an analysis, you know, that's more rigorous here.
But, you know, it's not, it's not what it is
in many other industries, right?
And so I think there are a few,
could be cause in fundamental investing,
which is the vast majority of AUM.
That doesn't mean it's the vast, or of active AUM.
But that doesn't mean it's the vast majority of trading,
obviously, quonference trade a lot,
relative to there.
But in terms of active dollars of AUM
and you could debate price setting and stuff like that.
But I think, I think in terms of price setting,
probably as well.
There is the thing that is driving that out,
that justification for being in active markets,
which should be alpha in the long term,
is the capacity of a human to think.
And of course, every one of these technology revolutions
enables the human, one mind,
which has differentiation to expand its aperture
and do more, but there are still,
it can't cover every single stock under the sun, right?
With, with advantage.
And so, yeah, I don't think it's gonna be
some crazy hollowing out of the middle.
I definitely think it will continue to increase dispersion
as every technology has and will for the next hundred years.
- You sit in a very unique seat,
in that your next PM worked at, yeah, great hedge funds.
And you work now with best hedge funds in the world.
And you're deep in the weeds.
And I'm sure your network in the space is ridiculous,
talking to a bunch of different people.
Do you have any differentiated views or contrarian views
on where the future of the industry is headed
that you think most people would disagree with you on?
- That's an interesting way to phrase it.
What do I think that's contrarian
about the structure of the industry itself?
- I certainly think that, I think the trend to,
what we said, I think this is beginning to play out already,
but the sort of trend to beta one or beta zero products
as an allocate from the allocation standpoint
into the world.
I think people are seeing that,
but I think that will just straight continue
because it's so hard to benchmark anything but that.
And that's why there's all this interest
and extension products to, like, meaning long only,
to our long 130 and short 30 to add incremental alpha
to a long only product because, again,
that beta one effectively or market neutral
allows that kind of clean auditing of performance.
Like, there's no nonsense on,
is this truly skill or is this,
like, of course, there's still no one's out of the hood
and ways to subtly take factor bets, right?
When people talk about that,
like sort of hidden factor bets and everything else.
But in the big picture, I think that will continue.
And I think the world of being sort of hyper analytical
about everything involved in the outcomes,
investing, space will continue.
But I also think, you know, the big question is always,
how does a new thing that doesn't have a historical analogy?
How does that play out?
I think, you know, we see everything on the experiments level
from people building AI agents
that are managing mock portfolios
and allocating capital there.
Again, I think whether that's augmentation or that's kind of,
there are instances of a sort of fundamental,
replicating quant.
Whether people will nail that in the long term.
I think there will be.
But I don't think that will, you know,
totally displace anything.
I think that will be one component,
one set of strategies that deploy a lot of capital
and succeed in that respect.
No, so I think, I think if I, you know, I think,
I think if I had a contrarian view,
it would be that that there's,
I know, maybe it depends what area you're comparing it to,
but if you're comparing it to like, you know,
I think one thing was, I think five years ago,
there was a consensus view that everything would,
all capital would accrue to like the big four,
let's say multi-managers.
I think that's, I think today,
what's happening is almost the opposite.
There's this great like fragmentation,
lots of those, the very successful portfolio managers
from those firms are launching their own
and the ability to launch your own hedge fund
is easier than ever.
If you're kind of focusing on the vertical
that you're an expert in,
I think there's a difference in terms of true multi-strategy
from day one and multi-kind of engine,
that's maybe those businesses
have sort of huge economies of scale as well.
But if you're a equity-longshore PM,
the ability to launch a new hedge fund
with, you know, separately managed account SMA type capital
where an allocator is looking for you
to have some degree of risk limits
and allocate capital that looks like your,
there's some degree, fragmentation's the wrong word,
but just sort of an opening up of that market
where it's not all the capital turning
to those biggest firms.
I think that's happening, those are a couple.
- And with those small firms that you mentioned
that I'm sure you work with some of them as well.
- For sure.
- What are some mistakes they make?
What are some things you see come up over and over and over?
- You know what's funny is,
I think a lot of those are doing really well, actually.
I mean, of course, everyone has imperfections limitations, but every hour of every day.
But in the big picture, a lot of those people are people who came from firms where they had
established track records and processes and are raising capital and want their name on the door.
And there's I think plenty of room across the market for different economic negotiations
and whatever else happens on that side of the table. But I think if you come into this industry,
a lot of what people want to do is have, you know, build a firm, build a hedge fund, right? If you,
if you're a, if you love stocks and you care about this job, you know, money's a piece of it,
of course, obviously, but, but also, you know, building a firm is a really motivating and exciting
thing for a lot of people. And I think a lot of the best people partly want that. And the industry
is organizing itself to enable this talent pool that has kind of, because of the multi-manders,
not even just the big four, but all of them, has been able to demonstrate skill in a way that is not
as hype cycle. Like, oh, this guy had this one big trade that really killed her. But it's like,
oh, this person stayed this many years at this firm and really, and everybody, and they broadly made,
you know, kind of consistent P&L and they did that by grinding out a lot of, you know, consistent
success over over a multi-year track record. Like, that's a pretty high odds bet. Now, of course,
you can debate magnitudes versus odds, but broadly, that's a pretty high odds bet that I want to,
I want to kind of back. And so I, our broad impression, you know, is that a lot of those are doing,
doing very well and having a lot of success as long as they get the tools they need, they get the
resources they need. And again, the industry, I think we're one, you know, part of this, but I think
there's a whole group of pieces of it, whether it's the allocators, whether it's the, you know,
even banks and enabling those firms to have corporate access and have research access and all
the different pieces that kind of fit into a puzzle of running a great hedge fund. If you're kind of,
you know, disintermediating or allowing that to happen at the, at the, at the new launch level,
that's what's happening. That, and that, and therefore, the, the sort of ability to do that is
greater and greater. And so that's, that's interesting, but, but, but at the flip side, even the
large firms can adopt great technologies and see some, but there are still economies of scale,
both in the mentioned corporate access, like, meaning, I want to say corporate access, just
tangibly, like, if you're the CEO of some big company, right? You're just more inclined to
meet with a large, as any business would be, with a larger investor in your business than a
smaller investor in your business, they're going to have more meaningful impact on your company
in the long term, right? Of course. So, there are kind of economies of scale, two size, of course.
That's one like minor example, but there are plenty of those. And so, yeah, I think, I think that's
kind of, um, the trade up, but I think a lot of those guys are doing the right things. They're not
trying to, I think there's a, there's a different trap. I'll put it this way where there's,
what's right way to say it? Like, very large launches that come in with a perspective that they
need to build. And what's funny is, um, yeah, and maybe some of those existed a slight era before
the latest era of tooling. Again, I think we're part of it, but I think there are other elements of
that stack where firms try to build everything, want to build the toy, you know, to an extent.
Build it fast. Yeah, but they also want to build. And I think over trying to do too many things
at once is, is the biggest mistake in the world. And so, when you talk about, um, either really
enormous firms have the resources to attack those different verticals, right? Whether that's,
and the, by different things at once, I mean, everything from different, uh, asset classes, you know,
different diversification within, within managers, within how you build your research process,
what things you build versus buy in terms of tooling, you know, all those things. Um, you know,
I'd say really big firms and really small firms often get that really right in the sense that
big firms know how to procure intelligently. Know how to quickly make the correct evaluation
about build versus buy. Oh, okay. Some cases we have to build because there isn't a great thing.
Some cases, there's a really best in class when we have to go with that. Um, you know,
how to quickly differentiate and, and then build in the spots that don't have buy in this because
you're always going to get more efficiency using a tool, right? Like if you, um, wanted to store
data in a database, you're going to do much better using AWS than building your own database up.
That's sort of obvious to say, right? But some people still don't always have that, right? But,
but large firms get that and get that at scale. That's how they, and so always that's how they become
large by having process excellence around making those decisions thoughtfully because it's not like
you should always build or always buy. It's just, it's case by case, you have to look at the set of
things and evaluate really carefully. Um, and small firms kind of can benefit from, hey, look,
I don't have the bandwidth to try to overdo everything. Um, I need to, I have to go, I have to almost,
I have to stack up. I have to just evaluate the best tools in every category and I need to do what
I've done and focus and be really excellent at that. There's kind of this middle sometimes new launches,
that, that, that, you know, try to do too many things at once. And I think that's the biggest,
and there's probably a metaphor for everything in life in some ways, right? But it's, um,
there's a lot of returns to being excellent at what at the unique thing. And for a stock picker,
again, kind of back to the earlier point, um, picking stocks is really hard. It's not, it's not a joke,
right? Um, it's, meaning in a consistent, constant way, where you're constantly held to, to,
to expectations and limits around losing money, um, picking generating alpha is really hard and, um,
distract taking your eye off that ball. Um, you know, it's like you're a hitter, right? Or,
you know, baseball or whatever the analogy is. If you're trying to do that and, um, you know, um,
do anything else besides stare at the ball and try to swing the bat correctly to get your
250 or 300 batting average or are in the investing case, 53, 55, 50, whatever it is, um, you know,
doing anything besides, like it's, it's one of those games where the odds are, are low enough
where if you try to, if you distract and try to do too many other things, you're almost definitely
in trouble. Um, let alone the aspects of scale that kind of crude. So, um, I think focus is the
most important thing. And those, and again, I think I, I feel like I see in really large firms
and in really new launches, I think I see that kind of ability to focus slightly different
ways they do it. Um, but that kind of ability to implement focus, um, is, uh, is, uh, is definitely
the kind of. And I think the business you've built is helping in some sense, helping PM's focus,
knowing what it is they're trying to predict, um, and really squeezing the juice out of that.
And I think what you said there is a lesson that is obviously the case of investing. But I think
all things in life, all things in business, um, you don't focus, you're dead. Thank you so
much for coming on outside open, which this was wonderful. For sure. Thanks so much for having me.
Podcast Summary
Key Points:
The primary role of a hedge fund portfolio manager is to generate excess returns relative to market risk through disciplined, analytical portfolio construction.
Over 90% of hedge fund failures stem from poor risk management and portfolio construction, not from missing fundamental insights.
Alpha is competitive and requires a correct, multi-layered perspective—combining fundamental analysis with quantitative risk modeling.
Portfolio correlation drags the sharp ratio toward the market’s return-volatility profile, making correlation management critical to outperformance.
Top performers succeed by managing forward-looking factor exposures and achieving idiosyncratic returns through deep, cross-verified analysis.
A key skill is identifying undervalued or mispriced opportunities by integrating macro trends, supply chain dynamics, and thematic shifts.
Leading hedge funds manage risk by treating factor exposures as if they were individual stock positions, enforcing strict limits and diversifying against correlations.
The best portfolio managers combine intuition with systematic modeling, using tools to constantly test, refine, and validate their investment thesis.
Summary:
The job of a hedge fund portfolio manager (PM) is fundamentally to generate risk-adjusted alpha by delivering a financial product that outperforms the market benchmark. While many attribute success to unique fundamental insights, the majority of failures stem from poor portfolio construction and risk management—particularly misalignment with market correlations. Any correlation in a portfolio drags the sharp ratio down toward the market’s return-volatility ratio, making correlation control essential.
Top PMs succeed by combining deep fundamental analysis with systematic risk modeling, identifying undervalued opportunities where market pricing is misaligned. They treat factor exposures as if they were individual stock positions, enforcing strict limits and actively managing anti-correlations. The most effective PMs integrate macro trends, supply chain dynamics, and thematic shifts into a cohesive analysis, using data and models to continuously test and refine their strategies.
This structured, analytical approach—grounded in both intuition and modeling—enables them to generate consistent, sustainable returns. Crucially, the market’s ability to price in trends means that true alpha comes from identifying and acting on rare, non-correlated, or mispriced opportunities. Modern tools now allow real-time decomposition of returns, making it easier to assess exposures, risks, and performance across the portfolio.
Ultimately, success hinges not on grand, untested bets, but on disciplined, data-informed, and rigorously managed portfolio construction that consistently beats the risk-free market alternative.
FAQs
The primary job of a hedge fund PM is to generate outperformance relative to a risk-free alternative by delivering a risk-adjusted return that exceeds market benchmarks, ultimately creating an attractive financial product for allocators.
Portfolio construction is a key driver of performance, as 90% of hedge fund and PM failures stem from mismanagement of risk, poor sizing, or inadequate diversification—especially in managing correlations that drag down the risk-adjusted return.
Top PMs generate alpha by identifying undervalued opportunities—stocks or assets that are mispriced—through a combination of deep fundamental research, data advantages, and heuristics that reveal market inefficiencies.
Correlation between portfolio positions drags the overall risk-adjusted return (sharp ratio) toward the market's level, as any correlated bet reduces the potential for excess returns beyond the market average.
Top funds manage risk by minimizing factor exposure, enforcing strict limits on correlations, and maintaining strong X-anti-correlation across portfolios to preserve a high risk-adjusted return over time.
Factor risk arises from broad market trends or systematic exposures (like momentum or value), while idiosyncratic risk comes from unique, stock-specific events—both are critical to managing portfolio volatility and returns.
Chat with AI
Loading...
Pro features
Go deeper with this episode
Unlock creator-grade tools that turn any transcript into show notes and subtitle files.