Olivia
Online
Marco Jean Aboav
@macro_fintech
public capital markets and AI nerd
London, England
Joined January 2016
1.6K Following
251 Followers
116 Posts
@LucaProsperi It boils down to who's the next buyer:
degen imploded
new tradfi tourists are exhausted
Btc maximialists are maxed out
big tradfi asset owners and central banks are the missing ones
Like in EM, the traders who survive are those who could read the flow.
Lots of people are concerned about the short energy supply in the AI capex cycle.
Open source and small (custom) language models will eat alive the closed-source stack.
It's just a (massive) ROIC problem.
You need to be plugged into the SV ecosystem to truly appreciate the disruption that is coming in your vertical.
Best ideas hardly come from vertical proximity.
This a typical problem I've faced in academic research, if you build on top of someone else research you're hardly gonna have disruptive innovation.
Sometimes you need to unplug, go places, and rely on serendipity to truly accelerate your human intuition. That's the only things AI will never replace: creativity and intuition.
You need to fish around from different fields, different crowds, different experiences.
I've truly enjoyed sharing our experience with other builders and judges in the agentic space at SVAI in Snowflake Palo Alto office.
Customization + managed accounts + AI
You don't need anything else in 2026
https://t.co/pRQ7NyaVCn
The biggest breakthrough in AI will be in science by far.
Glad to hear from Kevin Weil from openAI.
Somehow it's not that different from embedding the "physics" laws of financial markets into AI, like we do in @etnaresearcher
100x difference compared to the current prompt + sql circus
Total Portfolio Approach and AI in asset management are the biggest mega trend, and they are feeding each other, the rest is a second derivatives effect
More here
https://t.co/zhsm9b1t0G
Hierarchical reasoning is a smart move for superintelligence
https://t.co/97AfzjInzV
Griffin from Citadel on GenAI: No Impact on Alpha
Everyone’s talking about GenAI. It writes, summarizes, and automates.
But let’s be clear: it hasn’t moved the needle on alpha.
Here’s why:
1) It’s universal. Everyone’s using the same models — no edge. This is not that different when people claim they do unique things (coming from the same hedge funds and same schools)
2) It’s synthetic. GenAI predicts words, not markets. No context, no cognitive capabilities.
3) It’s derivative. It reflects consensus, not discovery. PMs care about tiny mispricing, not what the average analyst has to say.
4) It’s efficient, not insightful. Faster info retrieval doesn’t mean better ideas. The risk is to face info overload which in pre-genAI cycle implied terrible decision making
5) It homogenizes thinking — the opposite of alpha.
GenAI changes how we work, not what we know.
Alpha still comes from a cocktail of unique data, proprietary models, and human conviction — not from the latest prompt.
Human heuristics will always win in financial services, we need to empower it instead of eliminating it (as we currently do it with the current genAI trend).
AI pricing in financial services is upside down. If you’re charging a flat fee “for value,” either: 1) You’re underpricing, or 2) There’s no real value. Value in finance isn’t flat. It’s outcome-weighted. If your AI can’t justify variable pricing tied to results — you’re not building impact, you’re selling software.
Prove me wrong. ⚔️
@LucaProsperi If someone sends me an actionable plan on X to reach Y with aligned incentive, they might have my attention.
Unfortunately they are all impressed by our achievements...
BEAT THE PODS: A 7-POINT RECIPE FOR SINGLE MANAGERS
There has been some fun discussion here recently about the trend from single manager hedge funds (SMs) to multi-manager hedge funds (MMs or "pods"). Which is very sensical after a year like 2022 where there was a big performance differential between most pods and most SMs. With many SMs well below high-water marks, a year like '22 can also accelerate talent flow to MMs.
In my career, I had the really educational experience of both working at a very well run single-manager fund as well as three well run & well known multi-PM funds. In my role now at Fundamental Edge, I have the pleasure to work both with many top multi-PM funds and single-PM funds. Nothing here will be specific to any of these funds, but simply my high level observations & personal opinions.
I had a friend at a SM fund a couple weeks ago ask me "Brett, you've worked at both, what advice would you give to SM's to compete against MM's?". I also listened to Vinny & Porter's podcast talking about using their learnings from Citadel to accelerate their family office investing returns. Both got me thinking, and I'm in the process of developing a 100+ page deck with some structured thoughts to do some consulting with funds working through this change in the investment ecosystem.
The way I usually use twitter (X) is as a "first draft" of my thoughts. These is my napkin scratch, and I will refine this over time - in that vein I highly welcome your feedback since it helps me improve the thinking.
So, here we go, a recipe for single manager hedge funds to compete when "pods are eating the world".
1) KEEP YOUR TALENT.
Be honest, you expected point 1 to be "leverage time arbitrage" right? I'll get to that point. But in the SM vs. MM market share game, the foundational job as a SM GP/CIO is to keep your team.
The hedge fund world is unique in that even large, well-known single managers might only have investment teams of 6-15 people. And if you've lived in the world, you know that there are often 3-5 truly special investment professionals in that group who have that "nose for money" and are consistent money-makers.
I've seen plenty of SM funds gutted by losing talent to MMs. If the firm has a good culture & internal training & development, a "next person up" approach can work. But turnover of your A+ talent in the ultimate human capital endeavor - investing - is costly & risky.
My advice? Practice preventative maintenance with your talent. The old paradigm of "you should be lucky to work here and I'll pay you whatever I decide to pay you" for a talented 4-7+ year proven money maker just doesn't work anymore. The pods will give that person a $500m-$1bn+ portfolio, 15%+ payout, usually some nice up-front money and the autonomy and alignment that money maker desires. The pitch is compelling, and lots of the top SM talent is biting. (what shines is not always gold, but we will get to that part).
How do you compete?
Money is an obvious vector here. But very few SMs will compete with the bull case for a PM at a MM (i.e. if you build a portfolio to $2bn and make 5% on that you are consistently taking home $10m+ as the PM). And by the time that person has accepted the MM seat, more money isn't likely to sway them (and can set a bad precedent).
What I think matters more is alignment, visibility & a sense of partnership & shared mission. If you have an investment team of 12, who are the 3-4 on that team you absolutely don't want to lose (CIOs, you know who those are). Maybe make them a partner in the fund. Give them a contractual share in the management fee. Give them a 3-year comp plan with P&L contingencies. Your job is to try to create more visibility & confidence in compensation outlook so that the analyst view the SM seat as a higher multiple, more durable comp stream that that person can build a nice life around, even if that comp level falls short of the MM bull case (which it will).
Also, the soft stuff is the big stuff. Celebrate your team's life events, encourage them to bring their kids to the office, do off-sites & outings. Make it a "special" place to work and bring humanity into the day to day.
Understand burnout is a common thing in the 4-7 year range and sometimes burnout leads to exploring new pastures. Consider a sabbatical offering. I honestly loved the part of moving jobs where I got to take 1-2 months off between seats. Build that into your talent development plan.
Work to really articulate the pros & cons of the SM vs. MM seat. Tactfully walk through the risk of the MM seat - drawdowns, risk model constraints, top-line cost flow throughs, high turnover, etc (this is where I think our content will hopefully be helpful to our clients as we break down the MM approach in excruciating detail).
And if all else fails, and you really lose a star, keep an open door. If it were me, I'd let them know they have a seat when they want it. The grass looks greener for them right now, but when they get caught in a 200bps drawdown for positioning reasons and get a capital cut and 2 analysts quit...the grass doesn't look so green. Make it easy for them to come back. And, by the way, the pressure cooker of understanding that way to invest is likely to make them a monster upon return.
2) TIME ARBITRAGE.
This one isn't surprising, but it's true. In my MM teams, I always felt like we had a really good sense of the 9-18 month winners, but it is very hard to express those trades and stick with them. I can't just be long UNH, short WBA and wait it out - particularly if UNH is going to have a 10% pullback and WBA is going to beat a quarter & squeeze.
I run through this math in my deck, but a 20% drawdown on a 15% of LMV (long market value) position in a $50 vol, $1.5bn GMV book ALONE will hit my 150bps drawdown trigger and give me a capital cut. I simply cannot bear that risk as a MM PM - it is existential, as I then need a full 3% / 1-sharpe performance on my 50% book size just to get back to even.
What does that mean in simplistic terms? If I think there is a chance a stock goes down 15-25% before it goes up, I either can't own it or I can't own it in size. With more capital flowing to MMs, and that mathematical constraint true across the board, this behavior starts to create distortions to the price discovery mechanism in the markets.
How does that manifest? Via big over-reactions to near-term squishiness. Maybe a good company with a good 3-year story is going to miss a quarter, and due to that overhang the company has underperformed by 15-20%. Have a bias to lean into those trades.
The hero MM PM makes money 9-10 out of 12 months and limits losing months to under 100bps. That is done via a very sharp process around identifying inflections, revisions, and catalyst driven narrative shifts.
Find the "cheap but no catalyst" stories. Or sometimes the play might just to figure out when pods might want to buy the story and be there 3-6 months before. I came to LOVE situations where there was an obvious catalyst, but the catalyst wasn't coming for 4-6 months. Believe it or not, many traders won't wait that long, but the IRR of waiting can be really superb.
3) CONCENTRATE.
I am bearish on single-manager portfolios of yesteryear with 150+ positions. That is too many, in my mind. Markets are growing inescapably more efficient with alpha windows tighter and alpha pools more shallow. How wide is the alpha load on position 150 in that portfolio? I would submit not wide at all.
If 75% of stocks were fairly valued a decade ago, my guess is that is 90% today. There will always be anomalies & inefficiencies in markets, but the ecology of players has shifted.
The quotient of "dumb money" has decreased due to the secular trend of indexing.
Between quants arbitraging systematic, observable anomalies and pods arbitraging inflections & revisions applying their data & corporate access driven approach, anomalies are smaller.
So the bar should be higher for ideas in your portfolio. Concentration, to me, is the only way for SMs to survive. Find your great ideas and act decisively.
I strongly believe the days of the 40 person, 7-sector single manager 2 & 20 hedge fund are over. Markets won't allow that model to work anymore. The future for SM are smaller, nimbler teams with concentrated portfolios & low latency decisive decision making. It's the only way.
4) WHERE POSSIBLE, ALPHA ISOLATE.
Limited Partners (investors in hedge funds) have become much more sophisticated over the last decade. Particularly around factor attribution.
The MM's are true "alpha factories" in that the return stream is primarily idiosyncratic vs. systematic. That might seem academic, but in a year like 2022 when beta, investor overlay & LT momentum smashed returns at Tiger cubs, it becomes not so academic. MM's performed well last year by using market volatility to exploit idiosyncratic mispricings. If I'm an LP, that's EXACTLY why I'm paying 2 & 20.
I don't want to pay 2 & 20 for a HF portfolio to give me beta & exposure being long LT Mo and short residual volatility. I can do that now with factor baskets. I can buy the Goldman VIP ETF.
You have to give me something I can't get somewhere else for cheaper.
My advice - get a risk model. There are now some vendors who offer off-the-shelf risk models that are as good or better than exist at MMs (Equity Data Science is one of our speakers in Academy). Learn to understand the risk decomposition in your portfolio, learn to do an attribution analysis for your LPs. The good ones will ask for this.
And my advice would be, if possible, to use a portfolio construction approach with a 60-65%+ beta-neutralized idiosyncratic bar with as low a possible correlation to the S&P 500 and GS VIP. If I'm an LP building a portfolio of hedge funds, that's what I want in my stable.
5) RECONTEXTUALIZE EARNINGS.
Given the desire to be a 9+/12 month positive P&L generator and given the drawdown risk characteristics inherent in the 400-600% gross leverage wrapper, earnings season becomes a critical catalyst event for MM portfolios. And not just print day, but the run-up and post-print mis-pricings.
As a MM PMs live for earnings season. The print to print activity was all with an eye towards the next earnings catalyst. The modeling, channel checks, calls w/ sell-side, 2-4 interactions with corporates each quarter, and the end of quarter structured and exhaustive earnings preview process was all done with an eye towards monetizing the volatility inherent in earnings season (print days are 2% of a year's trading days but generate ~20% of idio volatility).
What does that detailed process mean for SMs? Well, earning prints are the ultimate moving bar, the ultimate expectations gap game. There is simply a low chance you are going to compete & win this game consistently in today's market - even if you wanted to, the massive sell-side wallet at MM firms gets them top priority in corporate access & sell-side access. Throw in multi-million dollar alt data budgets and it is an arms race that very few SMs can compete in.
My advice would be to try to re-contextualize earnings as a primarily defensive minded period where the understanding is the SM team might not have the full context of expectations. That seems disempowering, but the alternative of spending ~50% of your research time on the earnings cycle seems like a bad decision too. Listen, will a huge NVDA-like Q1 beat & raise still move stocks? Yes, obviously. Certainly find the inflection in your thesis & understand the catalyst path.
But know that if, increasingly, you don't understand price movement in your space on print day - that's ok. Use earnings as thesis check in. Perhaps do some positioning work (via MM PM friends or spec sales) and look for counter-positioning moves to fade. Company misses MM whisper and get's hate-sold? Take advantage of those counter-positioning moves to leg into positions you like. More react than predict, would be my advice.
6) GO WHERE THEY AIN'T.
Try to understand where pods play to better understand the ecology & alpha pools. Pods generally will really like three things.
1) highly liquid stocks
2) catalyst-rich situations, particularly latent revision potential
3) tangible/reliable downside
The life-blood of the MM model is liquidity. I need liquidity for a couple reasons.
1) I need to risk manage my book in a kerfuffle so I don't get my capital cut. I can't do that if I am 5 days of volume.
2) I am turning over my book 5-10x per year, so with shorter trades I need to get in and out quickly.
Generally my experience trading large portfolios is that anything north of 2-3% of daily volume (outside of a natural cross) starts to create slippage, i.e. I'm pushing a stock up or down. For a 10% LMV position on a $1.5bn book ($75m position), even if that stock trades $100m ADV, that is taking me 33 days to enter, 33 days to exit with minimal slippage.
What does that mean? It means that I might cover 250 stocks on my team, but I REALLY want to cover the most liquid 50. That is really where my P&L is coming from. I will still cover $25-$75m names, but anything under $50m and certainly anything under $25m starts to become very difficult to enter & exit, particularly if I am wrong on a trade and get stuck.
My belief is that the trend toward MM HF's will hollow out the $5-50m ADV competitive set and paradoxically make this a reliable alpha opportunity for the remaining single manager funds.
On revisions, understand that in most sectors the forward trajectory of EPS revisions is deterministic to stock price returns, with some exceptions. A more creative, value driven mindset can surface ideas where the revisions and stock price might diverge.
6) FADE VOLATILITY.
My sense is that there is now over $1tn of gross capital deployed in "short part of the alpha curve" volatility budgeted market neutral HF strategies.
What are the implications?
Well, when a volatility event happens and I have a 1.5-3% drawdown, i.e. my longs go down and my shorts go up, I more likely than now will have to exit that trade after it has gone against me.
In a MM wrapper gross capital effectively becomes pro-cyclical with returns. From the Tiger perspective on things (i.e. "if I like a stock and it goes down for a stupid reason I like i more"), this is non-sensical. If my longs go down 3% and my shorts go up 3% I should take gross exposure higher.
But...LEVERAGE. It is generally believed that MM funds will run with 400-600% gross exposure. Basically a MM hedge fund is like your typical 20% down buyer on a house - if the house value goes down 10%, I'm not down 10%, I'm down 50%. Leverage accelerates outcomes, and in a MM context turns 3% returns on gross into 15 returns on equity via the acceleration dynamics of leverage.
What does this mean in practice? Well, if leveraged players have a drawdown, their reaction to that drawdown is likely to actually ACCELERATE that drawdown via de-grossing of portfolios. If you've spent any time trading the last decade, you know the intensity & frequency of these de-grossing kerfuffles is only increasing.
This is just a math problem. If I run a big book at a pod and I have a 3% drawdown and the pod doubled my book from $2bn to $4bn to try to take advantage of that drawdown and I'm down ANOTHER 3%, it's a huge problem. A risk these funds can't and generally won't take. So there is a systemic risk approach to limit left tail outcomes by truncating your PMs who are underperforming. And it's worked at the GP level beautifully, so I wouldn't expect that to change (if anything, more replicators will come).
What does that mean for you, dear single manager?
These kerfuffles are your friend. Learn to identify them and monetize them. You are the only player at the table who has the capital to step into this kerfuffle. Do the work, but don't be afraid to fade a kerfuffle. Preferably on a beta-limited way, looking for blown out spreads. There are different ways I've monitored for kerfuffle and spread-blow outs and I am building that into my content.
7) UNDERSTAND THE ECOLOGY.
Don't have your head in the sand with regards to the other players in the market.
Great poker players study their opponents more than they study the cards - their approach to the game changes based on who is at the table. Annie Duke was a guest speaker at at analyst training and told us to "run the nuts" against retail investors, play the obvious hand when presented (that's worked beautifully on AMC & GME).
But the poker game changes when you are at a 12-top full of pros. Their tendencies & tells are harder to glean.
The same is true in markets in 2023.
When Buffett started investing professionally in 1956, there were virtually no investors doing fundamental analysis. Reading a 10-K was an alpha generator.
When Julian Robertson started investing professionally in 1980 the fundamental investing hedge fund industry was a cottage industry. You could have a massive informational edge simply with some effort.
The ecology is different in 2023. The hedge fund industry is $4.5tn+ with over 12,000 funds in the US alone. Information is distributed more uniformly. Quants have eaten the alpha on Buffett's favorite factors & turned those factors into beta. And thousands and thousands of sharks are out their looking for any sign of an edge.
My advice, study the players at the table. Understand the incremental buyer of your position - who are you selling it to. Work to understand when the MM, when the SM, when the LO might want to get involved.
And, importantly, spend time trying to understand how the shifting capital players create new anomalies in market. For example, the indexing wave has created a large alpha opportunity for index-rebalance.
It is my belief that 1) the shift to MM biz models in the HF industry is secular not cyclical. Shared overheads, risk diffusion (P&L is additive & risk is sum of squares), the challenges of a SM launch all align to support the view that the MM model is here to stay.
But out of that trend, I also believe that new alpha opportunities will emerge for the enterprising, adaptive single manager.
If you've read to this point, damn that ended up being a LOT longer than I expected it to be. I'd love your thoughts & feedback.
If you are a SM or MM who thought any of this is helpful, please DM me or e-mail me at [email protected]. As I said, I am working on a lengthy deck and some deeper sessions from clients on this front both delivered via zoom & in-person. I will have some capacity for consulting on this front.
Have a great Friday & weekend all!
@quant_arb I've been playing around with all the major LLMs on this stuff, same conclusion.
If you master the subject, you get a junior to mid dev support for 20 buck/month.
LLMs don't replace creativity, making money in markets requires a ton of it.
@investingidiocy what should be the way to run marketing in your opinion?
My company, Newfound Research, turned 15 today.
Coming up on this anniversary, I reflected quite a bit on my career. I’m not sure why, but this milestone feels larger than I would've expected.
So I decided to write something.
15 Ideas, Frameworks, and Lessons from 15 Years
@HarryStebbings Leverage an AI intern to extract audio, videos and text features to maximize audience dynamics. It's a bit art & a bit science, but the science could go well beyond those stats and could give you ton of insights, txt me if you need more context
@quant_arb The exchange stops you before you implode I suspect, depends on t&c, never allow cross margining among strategies/accounts.
@quant_arb Da Prado called the cafade of factors investing out: before fund launches is essentially just a simple overfit, it's impossible to consider capacity the reason for underperformance just 3 months.
Not surprised a big theme is causal factor investing.
@AlexanderGerko A macro trader betting on the event without getting exposure on the pre-event noise days/hours before
Last Seen Users on Sotwe
Trends for you
Most Popular Users
Elon Musk 
@elonmusk
241.7M followers
Barack Obama
@barackobama
119M followers
Cristiano Ronaldo
@cristiano
114.2M followers
Donald J. Trump
@realdonaldtrump
111.8M followers
Narendra Modi
@narendramodi
107.2M followers
Rihanna
@rihanna
98.7M followers
NASA
@nasa
92.4M followers
Justin Bieber
@justinbieber
91.8M followers
KATY PERRY
@katyperry
89.9M followers
Taylor Swift
@taylorswift13
83.8M followers
Lady Gaga
@ladygaga
75.3M followers
Virat Kohli
@imvkohli
73.2M followers
Kim Kardashian
@kimkardashian
70.8M followers
YouTube
@youtube
68.8M followers
Neymar Jr
@neymarjr
66.2M followers
Bill Gates
@billgates
65.1M followers
Selena Gomez
@selenagomez
63M followers
The Ellen Show
@theellenshow
62.3M followers
CNN
@cnn
61.8M followers
X
@x
60.7M followers