Whitepaper · 2025

AI-Native Fundamental Investing

A first-principles framework for agent–human collaboration across research, decision, and monitoring.

Discipline
Fundamental · >3 months
Thesis
Remove the constraints
Feasibility
Emerging · ~2027
Read
Abstract

This paper sets out a first-principles view of how AI reshapes fundamental investing. Genuinely AI-native systems — redesigned around agent–human collaboration rather than bolted onto legacy workflows — create durable advantage specifically in fundamental investing beyond three months, where the game is probabilistic prediction over vast bodies of information.

It examines what agents and humans are each good at, how work should be divided, the prerequisites for AI-native teams to match and then surpass human teams, where the technology stands today, and how a legacy fund makes the transition.

Once the constraints that bound human investing fall away, the optimal move is to run the most rigorous analysis on as many ideas, as often, as possible.

01 Why use AI for fundamental investing?

AI doesn't crown new winners everywhere — only where the game is synthesis.

Move along the holding-period spectrum. The durable edge sits at the far right.

AI narrows or decays the edge AI alters the game Durable AI advantage

Why is the long-horizon game different? No single data feed, proprietary call, or trend line deterministically predicts an outcome past the quarterly cycle. The work is large-scale information synthesis and pattern recognition — exactly where AI beats humans. And the market won't shrink as AI arrives; it grows more vibrant, because three properties hold.

i

The sport always exists

No one — human or AI — predicts perfectly. Many winners coexist, and any system can keep improving.

ii

Tolerance is higher

Returns trend positive over time as society advances — which is why new GPs launch every year.

iii

A winner for every strategy

Different horizons, markets, and sectors each demand their own framework to predict well.

02 What is each side good at?

Complementary strengths, not substitutes.

The agents

Built for scale & objectivity

No capacity constraints
Faster processing of information
Consistent execution — they do exactly what's asked
No emotion — more objective decisions
+
The humans

Built for the world off-screen

Friendship, reputation & social capital
Access to offline conversations & observations
Intuition honed from lived experience
Accountability for the outcome
03 What drives the outcome?

Five factors decide a fundamental fund's result.

01
🔍

Breadth of coverage

How many companies are considered.

02
📥

Information quality

The quantity and quality of info synthesized.

03
⚖️

Rigor of analysis

The correctness of analysis applied to it.

04
🧭

Pattern recognition

The framework used to reach judgment.

05
🤝

Brand & connections

The personal capital that drives fundraising.

04 From first principles — who does what?

Assign each task to the side built for it.

Agents

Scale · synthesis · objectivity
  • Screen as many companies as possible, as often as possible
  • Collect and synthesize all available digital information
  • Follow analytical guardrails faithful to the strategy
  • Make recommendations objectively
  • Root-cause and learn from mistakes objectively

Humans

Strategy · tacit knowledge · accountability
  • Use social identity to market and explain the system
  • Define the strategy and the analytical guardrails
  • Gather high-quality non-digital information to combine with the rest
  • Encode intuition into agents — and review every decision
  • Remain accountable for outcomes
05 Why can't agents set the framework themselves?

Judgment has to be encoded by humans — for three reasons.

Reason 01

Internet-average worldview

An agent leaning on raw model knowledge inherits the average of the internet — and cannot, in principle, out-judge the best human investors.

Reason 02

Completeness & locality

Not everything in a great investor's head is written down, and what is gets polished. What worked for one investor won't transfer wholesale to another.

Reason 03

Coherence

Analyses, frameworks, and strategy must cohere. Horizon, sector, market, and the investor's own strengths all shape the right guardrails.

Agents can refine on top of a framework — but they need it guided and encoded by humans. And whoever stays accountable for the system must understand its guardrails deeply.

06 Can it work on digital data alone?

The internet isn't the territory.

What's online

Eventually knowable by agents

As digitization improves, everything on the internet should be considered accessible. But the internet doesn't reflect reality — emotions are performed, propaganda and censorship abound, some voices dominate, and most people stay silent.

What stays offline

Where conviction is built

Skilled investigators surface what people won't volunteer online. Economic temperature is felt on a busy street, not in cold reporting. Seeing and interacting carry higher resolution.

The exception: longer-horizon strategies on stable businesses lean on established online facts — and a basket approach dilutes any single offline edge.

07 Prerequisites to match human teams

All of these must be true. Check them off.

00 / 15

Tap each requirement

These are the conditions an AI-native team must meet just to match a strong human team — before it can surpass one.

08 Where are we on the journey?

Early emergence — with a path to feasibility around 2027.

Tech-stack maturity extrapolates toward feasibility; human behavior and system iteration may delay the timing of real success.

Hallucination & data-source integration

Predictable progress

On a clear path toward inflection points, depending on model and harness.

Analysis & tacit-knowledge encoding

Largest gap

Design patterns are maturing, but seasoned investors lack the time, mindset, and setup to transfer what they know.

Non-digital data infrastructure

Behavior + engineering

Clear pathway — meeting bots, expert transcripts, RAG-enabled stores — not yet perfected.

Human–agent collaboration system

No efficient system yet

DIY builds underestimate the burden; legacy tools only assist existing workflows rather than enabling AI-native practice.

Mindset, org & process

Innovator's dilemma

Trust in agents fluctuates; the shift is hardest for older, successful organizations.

Linear extrapolation points to 2027 — though human change sets the real pace.
09 How AI-native investing outperforms

Human investing is optimized around constraints nobody questioned — until agents joined the team.

20companies
monitored closely at once
3models / day
valuation models updated
2in parallel
strategies pursued
Fading
memory horizon for past lessons

Illustrative figures. The principle: do one more analysis whenever its marginal value beats its (falling) token cost.

Human ceiling

Death to 80/20. Burn the tokens, and do all of it.

10 The ideal AI-native team

Smaller, sharper, and differently shaped.

The Architect (once called the PM)

×1
StrategizingDefine strategy, framework, analyses
TeachingEncode & refine agent guardrails
SupervisingScrutinize pieces, then the whole system
CommunicatingMarket & explain to LPs and media
AdministratingFinance, audit, legal, licensing
LearningFollow where the system leads

The Investigators (once called analysts)

×1–2
ConversingManagement, investors, experts
CollectingOffline, proprietary information
ShadowingTrace decisions full-cycle — how an analyst grows into a PM
11 The ideal enabling platform

Five layers, end to end.

Wide, timely data integration

High-quality sources, as broad as possible.

A reliable harness

Surfaces the right information and applies the right skills without errors.

Full-cycle automation

Ideation, macro, qual & quant analysis, private-info integration, human review, portfolio construction, risk, monitoring, buy/sell.

Effortless framework encoding

Skills, multi-step orchestration, and scheduling — all customizable with little effort.

A memory & decision-reference system

Human rules of thumb, agentic reflection on past trades, reusable wisdom, and a conflict-resolution agent to keep it coherent.

12 Build in-house or buy a platform?

A customizable third-party platform is the likely end state.

Proprietary DIY

  • Encoding, monitoring and refining is already a full-time job
  • Little bandwidth to build & upgrade the stack itself
  • Models imperfect, best practices still immature, ecosystem moving fast
  • UX and flexibility become second-tier when built only for in-house use
  • Result: slower, narrower evolution
Likely answer

Shared, customizable platform

  • Build once, serve many — a viable market forms
  • Data, infra, harness & reusable encodings get productized
  • Costs spread; smaller teams start with a lower barrier
  • Customizability and fast iteration are first-class priorities
  • Talent stays focused on investing, not infrastructure
13 How do you prove it's better?

Judge the system's logic, not any single quarter.

📈

Outperformance vs market

Relative return over a benchmark.

📊

Risk-adjusted return

Sharpe and similar measures.

🔁

Absolute across cycles

Performance through regimes.

Long feedback cycles and many interacting variables make attribution hard — but that's no different from how human investors have always been judged. Frame the fund as an evolving system: every component improves over time, and performance should follow.

14 The transition roadmap

Top-down, and deliberately sequenced.

PMs own the relationships, the strategy, and the team — so the decision is theirs. Once committed, the hard part is steering the chaotic in-between without breaking risk controls or culture.

1

Map & identify +

Map the current strategy → framework → analyses, then work backward to find the analyses agents can perform.

2

Encode into skills +

Assign those analyses to analysts; have them encode the guidance and steps into skills and multi-step flows, and own the results.

3

Stitch into automations +

Combine the tested blocks into partial automations — screening & ideation, initial analysis, monitoring, buy/sell, portfolio construction.

4

Run alongside humans +

Operate each automation beside existing routines as a check and supplement; refine frameworks toward optimal output.

5

Shift analyst time +

Move analysts from hands-on analysis to monitoring the automation — and to sourcing proprietary information offline.

6

Pilot full automation +

Run the full cycle without scaling yet, with analysts checking and refining the system.

7

Improve & expand +

Keep improving the system and developing new routines.

8

Scale beyond human +

Scale ideation breadth, analytical depth, and frequency until performance surpasses human teams.