· Graham Duncan

Graham Duncan: Talent Is the Best Asset Class

Talent is the only real asset class: the job of an allocator is to find people who are better at a thing than you are, understand what makes them tick, and put them into positive feedback loops.

talent-assessmenthiringreferencesinvestingallocationself-awareness0% confidence

Why this is in the corpus

Graham Duncan (East Rock Capital) is a legendary talent-evaluator and allocator. This episode is a dense framework goldmine on assessing talent, references-as-the-whole-thing, the aggression/integrity tension, and moving hidden assumptions from subject to object.

Summary for skimmers

Graham Duncan on how to find, assess, and unleash exceptional talent — the "what do you do here" origin story, the trace-of-fear hiring test, references as the real interview, commercial vs transactional operators, and infinite-game time horizons.

Briefing

What survives the editorial filter

This page should feel like a smart colleague already listened for you and left only the operating logic worth keeping. Not everything said in the episode makes it through.

Trust signal

Direct episode extraction

Best used for

Decision-grade retrieval metadata not yet added for this episode.

Hold lightly

No explicit downgrade reason stored yet for this episode.

Principles

Durable claims that survive beyond the speaker's biography — each with explicit limits, transferability judgment, and evidence.

Principle

Obsession is the highest talent signal

The strongest talent signal is obsessive intensity; a hiring manager wanting a trace of fear that the analyst will replace them keys on it.

Duncan compares it to Wayne Gretzky finding hockey at five, and to Buffett and John Arnold being obsessed from a young age.

Look for the person you'd be a little afraid is coming for your job.

“what I'm looking for is a trace of fear and myself that this guy is coming for me, that he will replace me.”Graham Duncan

Principle

A tight grip on a belief signals ego, not truth

When you grip a belief so tightly you cannot state its opposite, that is a sign of ego attachment, not of the belief's correctness.

Duncan uses this both to coach managers and on himself; his partner flagged his "futurism" grip around disruption risk.

If you can't argue the opposite, you're holding ego, not truth.

“that might be a sign that you have identity or ego caught up in that thing.”Graham Duncan

Principle

Everyone's genius sits next to their dysfunction

A person's greatest strength is structurally adjacent to their characteristic weakness; you buy them as a package.

Duncan says this tolerance for ambiguity can make him look indecisive because he holds the contradiction of someone being genuinely great and genuinely flawed.

Don't try to hire genius without its attached dysfunction.

“everyone's genius is right next to their dysfunction”Graham Duncan

Principle

Commercial beats transactional

Seek "commercial" operators who create more value than they capture, not "transactional" ones who try to capture it all.

Duncan ties this to an abundance mentality and to the aggression-plus-integrity signature; commercial people tend to clump together.

Commercial operators cluster; transactional ones burn the network.

“commercial as people give off this vibe of I'm gonna create more value than I capture here. Versus somebody who's more transactional is trying to capture all the value.”Graham Duncan

Principle

Slugging percentage beats hit rate

Investing returns come from slugging percentage, sizing winners bigger than losers, far more than from raw hit rate.

Duncan notes a great PM's hit rate is ~65% and 55% will do fine depending on how they size winners versus losers.

Being right less matters than sizing right.

“You wanna size the winners obviously bigger than the losers.”Graham Duncan

Principle

Taste over judgment

Evaluate people with taste, not judgment, because judgment implies one right answer where talent selection has many.

Duncan likens it to picking a certain kind of podcast guest out of a lineup: a coherent aesthetic, not a verdict.

Cultivate taste in people; resist reducing it to pass/fail judgment.

“I like taste over judgment because judgment implies there's one answer.”Graham Duncan

Principle

In investing there are no products, only the mindset of the individual

Investment management has no products, only humans; you are underwriting a mindset and the prospective decisions it will make.

Duncan notes that if a Fidelity PM leaves and a new one arrives, from his perspective it is a completely new decision, not the same product.

A fund is a mind, not a product; re-underwrite when the mind changes.

“The only product is the mindset of the individual and the decisions they're going to make prospectively from here.”Graham Duncan

Principle

Test integrity by behavior in prior moments of stress

Assess integrity by reconstructing how a person behaved in prior moments of stress, not by asking them to self-report it.

Duncan uses 2008 as a canonical stress test that tested LP-GP, partner, and banking relationships all at once.

Character shows up in earthquakes; go find the earthquake in their past.

“what I'm trying to do when I'm trying to understand for instance, someone's integrity is understand how they behaved in prior moments of stress”Graham Duncan

Principle

The meta-identity of a moneymaker carries no baggage

The best managers identify as moneymakers, not as being right, so they can go long or short the same name with no baggage.

Duncan contrasts this with managers whose identity is tied to being smart or to a specific thesis, which makes reversal painful.

Ego in the position is the enemy of changing the position.

“the meta identity is I'm a moneymaker.”Graham Duncan

Principle

Never feel pressure to put money to work

A disciplined allocator is always comfortable holding cash and never feels pressure to deploy assets.

Duncan contrasts this with the dangerous "I need to put out assets" mindset that leads people to invest in products.

Deployment pressure is how standards die; stay comfortable in cash.

“you always want to feel comfortable holding cash and there should be no pressure but money to work.”Graham Duncan

Principle

Hire people who are better than you at the thing

The allocator's edge is hiring people better at a thing than you are and defining the excellence you cannot personally perform.

Duncan traces this to being 24, hiring and firing people older and more experienced than him; the imposter feeling forced him to articulate what constitutes excellence in each role.

Your inability to do the job is an asset: it makes you specify what great looks like.

“I try to find people who are better at doing a thing than I am at doing that thing.”Graham Duncan

Principle

Get people into positive feedback loops

The job of a talent allocator is to design seats so people do what they would want to do anyway, unlocking self-reinforcing performance.

Duncan frames organization design as constructing positive feedback loops so he does not have to micromanage; it is "right next to" his own laziness.

Put people where they are already compulsive; the loop does the managing.

“my mission in life is to get as many people as possible into positive feedback loops.”Graham Duncan

Frameworks

Reusable systems and operating models — including when they help and when they break.

Framework

Subject-to-object: adult development as making assumptions visible

Growth is the process of moving assumptions from subject (invisible, controlling you) to object (visible, choosable).

Duncan applies Bob Keegan's model with his coach Caroline Coughlin, who surfaces the assumptions that hold him rather than him holding them.

Make your assumptions object; the ones you can't see run you.

“the process of moving to adulthood is one of increasing your mental complexity and increasing the number of things that used to be you used to be subject to assumptions you had, you couldn't see and making them object over time.”Graham Duncan

Framework

Credibility = proven competence + relationships + integrity

Assess credibility as proven competence plus relationships plus integrity, then use that assessment to decide how much decision space to give someone.

Duncan borrows the formula from Chris Fussel and Stanley McChrystal and says it is exactly how he thinks about his own work.

Credibility is a three-part formula that sets how much rope you give.

“credibility equals proven competence plus relationships, plus integrity.”Graham Duncan

Framework

Orchids versus dandelions: match talent to context

Classify talent as orchids (world-class only in the right context) or dandelions (lower-maintenance, robust anywhere), and place each accordingly.

Duncan calls his approach "wild gardening": don't force a dandelion where it doesn't want to be, and give orchids their specific light and temperature.

Talent is context-dependent; the allocator's job is context design.

“orchids can be so beautiful but they need just the right context, they need the right light, temperature and dandelions are just lower maintenance.”Graham Duncan

Framework

Finite versus infinite games in investing

Distinguish finite players (hit a number and leave) from infinite players (play for the game's own sake); back the infinite players.

Duncan draws on James Carse and Kwame Appiah; a finite investment manager targets a net-worth number, whereas infinite players keep playing and recruit others.

Decide which game you're playing, then partner with people playing the long one.

“it's not how well you play the game, it's deciding what game you wanna play.”Graham Duncan

Framework

The paradoxical talent signature: aggression plus humility

The recurring talent signature is a paradoxical pairing of aggression and humility that top evaluators all name in different words.

Duncan notes YC asks if someone is "an animal," Buffett says passion and common sense, Dalio says assertiveness and open-mindedness, all pointing at the same tension.

Great evaluators all describe one paradox: fierce and humble at once.

“Buffet uses passion and common sense. Ray Dalio uses assertiveness and open-mindedness. But I think everybody's honing in on this paradoxical tension”Graham Duncan

Framework

The river with two banks: order versus chaos

Model careers as a river between an order bank (apprenticeship, refining reality) and a chaos bank (originality, asserting reality); stay in the middle and oscillate.

Duncan adapts Dan Siegel's mental-health model; he places Musk and Jobs right next to the chaos bank, brilliant if they keep reality feedback, crazy if they lose it.

Apprentice on the order bank, then swim toward chaos without leaving the water.

“I like the image of a river with two banks. One side is chaos in the other side is order”Graham Duncan

Signals

What appears to be shifting, for whom it matters, and what happens if you ignore it.

Signal

The next crisis will re-sort behavior

Future crises are coming and will again act as natural experiments that separate who behaves well from who behaves badly under stress.

Duncan frames this as an opportunity: each crisis regenerates the integrity signal he most wants on managers.

Downturns are recurring character audits; watch who stands up.

“I'm sure we will have additional crises soon that will again enable us to separate those who behave well from those who behave badly in those periods of stress.”Graham Duncan

Signal

Market games decay, forcing continual reinvention

In financial markets the games that worked five years ago stop working, so staying top 1% demands continual innovation and adaptability.

Duncan cites event-driven merger-arbitrage strategies as an example of a game that has mainly gone away.

No strategy is durable; the durable skill is reinvention.

“the games that were working five years ago no longer work.”Graham Duncan

Opportunities

Only included where there is a buyer, a real wedge, and a plausible revenue path — not vague idea theater.

Opportunity

Customizable life-calendar photo poster

There is an unfilled product: a customizable life-in-weeks poster where you select a photo for each week, ideally partnered with Tim Urban.

Duncan bought Urban's life-calendar poster and fills in circles for milestones; he explicitly calls out the photo-per-week version as a product gap for graphic designers.

A graphical life-in-weeks with photos is a named, unbuilt consumer product.

“a good product, if there are any graphic designers out there, would be to partner with Tim and allow you to create a huge poster and select a photo for each week.”Graham Duncan

Opportunity

Unleash investment genius trapped inside large orgs

A durable opportunity is identifying investment craftsmen trapped inside large organizations and backing them to launch, unlocking genius their prior context suppressed.

Duncan's firm meets over a thousand teams a year partly to find and seed managers who can be unleashed from constraining institutions.

Trapped talent inside big orgs is a recurring, backable seam.

“those craftsmen, those investment managers who were trapped inside large organizations, start their own firms on their own or where they're help and you, you unleash all this extra investment energy in genius that they weren't accessing in their prior context.”Graham Duncan

Lessons still worth keeping

Useful takeaways that did not fully clear the bar for durable principle status.

Lesson

The Solar City short: laughing when the trade blew up

A manager short Solar City laughed the morning Tesla announced it was buying it, modeling ego-free equanimity and flexibility under a painful loss.

Duncan respected him more for the lightness; he trusts that this manager could be short the same position the following month with no qualms.

The manager who laughs at his own blown trade has the right identity.

“he was laughing, he was enjoying the audacity.”Graham Duncan

Lesson

The hour-long reference that demanded specifics

Another allocator refused to let Duncan give a vague positive reference, pressing for specifics for an hour, which modeled reference rigor and became a close friendship.

Duncan entered the call ready to give a mild endorsement; being forced to specify changed how he now runs references.

The best references are earned by refusing the easy endorsement.

“the guy sat on me for an hour and would not allow me to basically give a positive endorsement without specifics.”Graham Duncan

Lesson

"What the fuck do you do here?" revealed his real skill

A sales hire asking Duncan what he even did there crystallized his real skill: putting people into the seats where they build the business.

Duncan, 24 and full of imposter syndrome, realized he had reseated Robert from analyst to salesperson, and Robert built their entire business.

The uncomfortable question named the skill: reseating people into their genius.

“Graham, what the fuck do you do here?”Graham Duncan

Lesson

Traders are the nurses of finance

An OB-GYN's advice to ask the nurses generalized into a rule: the proximate observers of someone's pressure decisions hold the best signal, and in finance that is the traders.

Duncan derived from his own OB-GYN search that nurses know who has ice in their veins, and mapped it onto traders as the receivers of PM decisions.

Ask the nurses; in your field, find who plays that role.

“the nurses know who freaks out during tricky situations and who's got ice in their veins and who makes good decisions”Graham Duncan

The Plays

Try these this week

Verb-first executable actions — each one tied to a stated outcome in the episode.

Underwrite a live deal together to resolve ambiguity

Outcome: When a person is ambiguous, get out of the conference room and underwrite a live investment together to see how they really think.

Context: Duncan will look at a live hotel or SPV with a prospective partner, and also builds context on intense trips, because a pitch across a table reveals little.

“If you look at something that's live with somebody, you learn so much about how they're actually underwriting it, the mental models they're using.”
— Graham Duncan

weeks of shared work per

Before you start

· access to a real live deal or problem
· willingness to spend real time

talentinvesting

Ask "what criteria would you use to hire for this?"

Outcome: Ask candidates and experts what criteria they would use to hire someone for the role; it exposes their definition of success and improves your own.

Context: Duncan used it to choose an OB-GYN and gets nuance he could not have generated himself, then reuses the sharpened criteria on the next interview.

“if you were hiring somebody to do this, what criteria would you use to hire that?”
— Graham Duncan

single meeting, compounding across a search per

Before you start

· a genuine curiosity stance
· a role or decision to anchor the question

talenthiring

Run references as the whole interview, past the script

Outcome: Treat references as the entire assessment: use on-list and off-list references, push past the ten-minute script, and stop around reference four when patterns repeat.

Context: Duncan finds even on-list references bizarrely high-signal because scripts run out; he aggregates net promoter scores and calls ~20 people.

“after 10 minutes on a reference, people run out of like nobody wants to lie.”
— Graham Duncan

days to weeks per hire per

Before you start

· a curiosity, non-gotcha stance
· access to off-list references

talenthiring

Ask "what happened in 2008?"

Outcome: Ask experienced people what happened in 2008 to test how they behaved under maximal stress across every relationship.

Context: Duncan treats 2008 as an event that tested LP-GP, partner, and bank relationships simultaneously, and expects future crises to serve the same function.

“So if they've been around a while, to me an interesting question is what happened in oh eight?”
— Graham Duncan

single interview plus reference cross-check per

Before you start

· candidate has relevant crisis-era tenure

talentinvesting

Make the reference feel safe so they open up

Outcome: Explicitly make references safe: tell them you are calling many people and passing nothing back, and hold a curiosity, not-catching-anyone stance.

Context: Duncan says the reference picks up on the vibe; his goal is to help the candidate land in the best context, not to catch them.

“I'm calling 20 people. I'm not passing any of that information back to the candidate unless you ask me to. So you need to make it safe for them.”
— Graham Duncan

first 60 seconds of each call per

Before you start

· genuine intent to help the candidate
· discipline to not weaponize what you hear

talenthiring

Ask who their complementary partner should be

Outcome: Ask what a partner hired to complement this person should be good at; the answer maps their weaknesses and the team to build around them.

Context: Duncan then uses the answer to help the person assemble the right complementary partner, in the investment context another investing partner.

“if I were gonna hire a partner for this person to compliment their strengths and weaknesses, what would that person be good at?”
— Graham Duncan

single conversation, ongoing team design per

Before you start

· a specific person to anchor on

talenthiring

Force a 1-10 endorsement and ban the seven

Outcome: When scoring references or options 1-10, forbid the seven so people cannot hide in a safe, mild endorsement.

Context: Ferriss learned it from Kyle Maynard (via a successful CEO) and uses it for references and restaurants; Duncan uses net promoter scoring the same way.

“it was the rank X could be anything from one to 10, but you can't use a seven.”
— Tim Ferriss

seconds within a call per

Before you start

· a defined subject to rate

talenthiring

Pit a long against a short to test for disconfirmation

Outcome: Introduce a long and a short on the same name and watch the exchange to see who actually seeks disconfirming evidence versus defends an ideology.

Context: Duncan warns Tesla is an extreme, ideological example, but the extremity makes the disconfirmation-seeking behavior easier to observe.

“sometimes I'll, I'll hook up someone who is long Tesla with someone who is short Tesla. And when you're witnessing that exchange, you can see whether the person who's long Tesla and the person who's short Tesla, are they actively seeking disconfirming evidence or is it an ideological thing?”
— Graham Duncan

a single facilitated conversation per

Before you start

· access to opposing credible investors
· a genuinely contested thesis

talentinvesting

Decision Moments

Actual decisions, real outcomes

Specific decisions narrated in the episode with their outcomes and transferable lessons.

At 24, running an independent research firm and full of imposter syndrome, Duncan had hired a man named Robert as an analyst, but Robert did not light up about analysis and instead relentlessly stalked and called Duncan for four weeks during the interview process.

Did: Instead of managing Robert as an analyst, Duncan reseated him as the firm's salesperson, matching the role to the compulsive behavior Robert had shown rather than the job he applied for.Outcome: Robert built the firm's entire business; years later a former prospect still had cold-sweat dreams of Robert chasing him for a meeting. The episode crystallized Duncan's core skill of putting people into the right seat.

Match people to the seat their behavior reveals, not the role they applied for; the applicant's compulsions are the signal.

Part of an emerging decision pattern across multiple episodes

Duncan and his wife needed to choose an OB-GYN for the birth of their child, a decision with severe, low-probability downside and where every candidate had already cleared many credentialing hurdles.

Did: Rather than assess doctors directly, Duncan asked each what criteria they would use to hire for the role; one answered that you maximize for the downside via NICU affiliation with a teaching hospital and by asking the nurses, who know who has ice in their veins.Outcome: Duncan adopted the NICU-plus-nurses criteria and generalized it into a durable reference heuristic: seek out the proximate observers of pressure decisions, which in finance are the traders.

The criteria question surfaces nuance you did not know to ask about and upgrades your own criteria for every future evaluation.

Part of an emerging decision pattern across multiple episodes

Duncan's business partner observed that Duncan held a too-tight grip on "futurism" - a paranoia about disruption that made him over-weight climate and technology-disruption risk when evaluating hotels and value investments.

Did: Rather than dropping the disruption lens, Duncan made the assumption object: he kept the paranoia as a useful lens but flagged his awareness of the bias whenever it surfaced in an investment evaluation.Outcome: Duncan retained a lens he considers valuable while reducing its distortion of his judgment, converting an invisible assumption he was subject to into an object he can examine and choose.

When you find an over-gripped belief, make it object rather than abandoning it; awareness of the bias preserves the lens without letting it run you.

Tensions surfaced

Contradictions and trade-offs the episode raises — judgment calls a thoughtful operator has to navigate.

Tension

Originality versus triage

You want people who generate many ideas and who ruthlessly kill them, but idea-fertility and triage rarely coexist; the best hold both.

Duncan cites Dean Simonton on genius and volume, then notes the best PMs can force-rank their portfolio and kill their own ideas with high situational awareness.

Both true: have many ideas, and be willing to let most of them die.

“when people have lots of ideas, they're usually not that good at killing them.”Graham Duncan

Tension

Tolerating ambiguity versus deciding

Evaluating talent requires tolerating contradictory signals long enough to make sense of them, which can look like indecision but preserves information.

Duncan's wife and business partner can experience him as indecisive; his resolution is to spend more time and create other contexts, and if it never resolves, do nothing.

Both true: you must eventually decide, and premature decision destroys signal.

“holding those contradictory perspectives as long as you need to to make sense of them.”Graham Duncan

Tension

Aggression versus integrity

Great operators need maximal aggression and genuine integrity at once, a tension resolved by seeking "commercial" people who balance both.

Duncan uses David Tepper as the platonic ideal: ruthless in sizing a position, but never overly aggressive against a counterparty; commercial people who hold this tension clump together.

Both true: be fierce in the position and fair to the counterparty.

“the tension between someone's intensity and integrity or if you double click on each of those aggression and humility”Graham Duncan

Tension

Refining reality versus asserting reality

Careers require both refining reality (apprenticeship) and asserting reality (originality); health is oscillating in the middle, not living on either bank.

Duncan places Musk and Jobs next to the chaos bank: brilliant if they keep reality feedback, crazy if they lose it; most should swim back and forth.

Both true: learn the rules on the order bank, break them near the chaos bank.

“healthy integration is swimming in the middle of the river in between those two.”Graham Duncan

Corpus connection

Where this episode fits for retrieval

What kinds of decisions this briefing is best pulled into.

Primary decisions

• hire
• allocate-capital
• partner-selection