Hedge Funds interview preparation

Questions

100

Traced to a firm

39

Firms

16

Updated

September 2026

Asked at

All firmsMan Group10Balyasny Asset Management7Bridgewater Associates3DED.E. Shaw3Apollo Global Management2KKR2Oaktree Capital Management2Point722SCSquarepoint Capital2ACAQR Capital Management1BGBaupost Group1Coatue Management1HPS Investment Partners1Northern Trust1Viking Global Investors1Wolverine Trading1

Topic

All topicsStrategy taxonomy8Stock pitch10Short selling6Portfolio construction8Risk and drawdown8Performance and alpha7Event-driven and merger arb8Distressed and credit5Fund structure and economics7Financing, NAV and operations6Compliance and research process5Quant and systematic6India and Category III AIFs5Career and fit11

Level

AnyCoreIntermediateHard

Type

AnyTechnicalMarket viewBrainteaserCaseFit

1. 009Pitch me a stock.Stock pitchIntermediateevery roundMan GroupEquity Hedge · London · 2016Apollo Global ManagementInvestments · Remote · 2021

   Say this

   Trade first, then the business, then the variant view, then the catalyst, then the risk and what would make you wrong. Ninety seconds. At a hedge fund the variant view and the catalyst are the only parts that get you hired; everything else is table stakes.

   Then walk it

   1. Open with the position, not the company. 'Long X at 62, target 85, roughly 35 percent upside over twelve to eighteen months, and I would size it at 4 percent of the book.' Sizing in the opening line is what separates a fund pitch from a research note.
   2. Two sentences on the business. What it sells, to whom, and the one operating metric that drives the P&L.
   3. The variant view, quantified. 'Consensus has 9 percent revenue growth next year. I think it is 14, because the two contracts announced in April are not in the sell-side models yet and they are worth 5 points of growth.' Name the number consensus has and the number you have.
   4. The catalyst and the clock. What makes the market agree, and when. A quarterly print, a capacity ramp, a contract renewal, an index event, a capital markets day. Without a dated catalyst it is an opinion, not a position.
   5. The bear case with a price on it. 'If the contracts slip a year I lose about 15 percent.' Then the asymmetry: 35 up against 15 down justifies the position even at even odds.
   6. Close with the falsifier and the hedge. The one disclosure you would watch, and how you would express it, whether outright long or paired against a competitor to strip out the sector move.

   Where candidates lose it

   Pitching a mega-cap with a thesis from the financial press. If the reason is in the newspaper it is in the price. Also, never pitch without a number for the bear case and a sizing view. A hedge fund interviewer is testing whether you think in positions, not in recommendations.

   Expect next

   • How would you hedge it?
   • What is the bear case, and what does the stock do in it?
   • Who is on the other side of this trade and why are they wrong?

   Reported by candidates at Man Group (Equity Hedge, London, 2016); Apollo Global Management (Investments, Remote, 2021). Source: Wall Street Oasis.

2. 010What is your variant view on that name, and why is the market wrong?Stock pitchIntermediatesuperdayLong-short equityMulti-manager platforms

   Say this

   A variant view is a specific, numerical disagreement with consensus plus a reason the disagreement exists. Two parts, and people forget the second one. If you cannot say why the market has not already worked it out, you probably do not have an edge, you have a summary.

   Then walk it

   1. State it as two numbers. Consensus 2027 EBITDA is 940 million; my number is 1.15 billion. That gap is the position.
   2. Then the source of the gap. There are only a few legitimate ones: you have better information, you have done work others have not, you have a longer horizon than the marginal holder, or you read the same facts differently.
   3. The structural reasons a gap persists are the most defensible: the stock is under-covered, the disclosure is buried in a segment note, the shareholder base is index and cannot act, or the payoff sits beyond the two-year window most sell-side models run to.
   4. Test it against the price. Run a reverse DCF or an implied-multiple check: what does the current price have to assume? If the market is discounting 6 percent growth in perpetuity and the installed base alone gets you 8, you have located the disagreement rather than asserted it.
   5. Then the falsifier. What observable would tell you consensus is right and you are wrong, and when do you see it? If nothing can, it is a belief, not a thesis.
   6. Be honest about the weakest variant view: 'the market is short-termist' is not an edge, it is a hope. The strongest is a measurable fact you found that is not yet in the numbers.

   Where candidates lose it

   Restating the bull case louder. A variant view has to differ from consensus in a quantified way, which means you must actually know what consensus is. Look up the sell-side number before the interview. Candidates who cannot say what the street has next year lose this question in one sentence.

   Expect next

   • What is consensus for next year?
   • Why has the market not figured this out?
   • What is the one data point that would prove you wrong?
3. 013What is your favourite telecom stock?Stock pitchIntermediatetechnicalBalyasny Asset ManagementGeneralist · New York · 2020

   Say this

   Pick one, commit to it, and make the answer about the sector's economics rather than the company's story. Telecom is a capital intensity and pricing-power question: the winner is whoever earns a return above cost of capital on the network they have already built.

   Then walk it

   1. Frame the sector first, in one line. Telecom is a high fixed cost, low marginal cost, heavily regulated oligopoly where the swing variables are subscriber pricing, capex intensity and spectrum cost.
   2. Then the metrics that matter, which are not the ones from other sectors. ARPU, churn, subscriber net adds, capex as a percent of sales, EBITDA less capex, and net debt to EBITDA. Leverage is structurally high, so the equity is a levered bet on ARPU.
   3. Then the actual pick with a number. For instance, a market where three players have replaced four and tariffs are rising: the operator with the lowest cost per gigabyte and the most spectrum gets disproportionate incremental margin because every new subscriber drops through at near-zero marginal cost.
   4. The India angle is genuinely the best telecom case study going, and worth using if you know it. Tariff repair after consolidation moved ARPU up materially, and the equity story became entirely about whether that pricing held while capex rolled off.
   5. Say the bear case in the same breath. Spectrum auctions are a recurring, unavoidable capital call; a price war resets the whole thesis in a quarter; and regulated markets can hand a windfall to the consumer at any point.
   6. Then the hedge, since this is a hedge fund question. Long the share gainer, short the subscale operator with the same spectrum costs and worse coverage. Same regulatory risk, opposite unit economics, and the trade isolates the operating gap.

   Where candidates lose it

   Answering with a household name and a vague 5G story. The interviewer is testing whether you know the sector's unit economics. If you cannot say ARPU, churn and capex intensity for the name you picked, pick a different sector. Also do not say 'I do not follow telecom' and stop; name what you do follow and offer that instead.

   Expect next

   • What is ARPU doing in that market and why?
   • How would you short telecom?
   • How do you value spectrum?

   Reported by candidates at Balyasny Asset Management (Generalist, New York, 2020). Source: Wall Street Oasis.

4. 018How would you perform a whitespace analysis?Stock pitchIntermediatecase studyViking Global InvestorsQuantitative Research · New York · 2024

   Say this

   Whitespace analysis maps where a company could sell but currently does not, and then asks how much of that gap is actually reachable. It is a growth-runway tool: define the full addressable grid, mark what is already penetrated, and size the remainder with an honest win rate.

   Then walk it

   1. Build the grid explicitly. Two or three axes that matter: customer segment by geography by product. Cells are combinations, and every cell gets a size and a current penetration.
   2. Get the penetration from disclosure and bottom-up data rather than management's TAM slide. Customer counts, store counts, licence counts, segment revenue, and where possible third-party data like app downloads or job postings.
   3. Then subtract what is not really reachable. Cells owned by an entrenched incumbent with switching costs, cells where the product does not fit without heavy investment, cells where regulation blocks entry. That haircut is the analytical content of the exercise.
   4. Then convert it to a number the model can use. Reachable whitespace times a realistic win rate over a defined period equals incremental revenue, and I would keep the win rate low enough to be defensible, often 10 to 20 percent rather than a third.
   5. Cross-check against history. If the company has been adding 400 customers a year and your whitespace implies 2,000 a year, the whitespace is not the constraint, execution is. Reconcile the two or drop the analysis.
   6. The limitation to say out loud: whitespace tells you the ceiling, not the path. Plenty of companies with enormous whitespace never grow, because the gating factor is sales capacity or the economics of serving the marginal customer, not the size of the opportunity.

   Where candidates lose it

   Producing a large TAM number and calling it a thesis. TAM slides are marketing. A credible whitespace analysis is mostly about what you exclude and the win rate you apply, and it must reconcile with the company's demonstrated rate of expansion.

   Expect next

   • How would you validate the penetration numbers independently?
   • What win rate would you use, and why?
   • Where does this analysis mislead you?

   Reported by candidates at Viking Global Investors (Quantitative Research, New York, 2024). Source: Wall Street Oasis.

5. 019Pitch me a short.Short sellingIntermediatesuperdayLong-short equityShort-biased funds

   Say this

   Same structure as a long, but three extra things have to be in the pitch: why the market is wrong in a way that resolves on a clock, what the borrow costs, and what blows you up. Shorts are timing trades, not valuation trades, so name the catalyst before the valuation.

   Then walk it

   1. Lead with the trade and the constraint. 'Short X at 40, target 25, borrow is 3 percent annualised and there is ample availability, and I would cap it at 2 percent of the book because short interest is already 9 percent of float.'
   2. Then the thesis, and pick a category. The good short buckets are structural decline being extrapolated as cyclical, accounting that overstates earnings quality, a broken unit economic that growth is masking, and a balance sheet that needs to refinance into a worse market.
   3. Then the catalyst with a date, because time works against a short. A refinancing, a covenant test, a lock-up expiry, a competitor's capacity coming online, a guidance reset. Valuation alone does not close a short.
   4. Then the carry. Borrow cost, dividends you owe, and the interest you earn on the proceeds. A 12 percent borrow means you need the thesis to work within months, not years, and saying that shows you have actually shorted something.
   5. Then the blow-up risk, explicitly. Float, short interest as a percent of float and days to cover, retail interest, index events, and whether the company could do something reflexive like a buyback, a raise or getting acquired. A takeout is the classic way a good short thesis loses 40 percent overnight.
   6. Close with sizing and stop. Shorts get smaller as they go against you in risk terms because the position grows, so I would run a hard stop and a smaller starting size than an equivalent-conviction long.

   Where candidates lose it

   Pitching an expensive stock. 'It trades at 60 times earnings' is not a short thesis, it is an observation, and the last decade has been brutal to people who thought otherwise. Also, candidates forget the borrow and the squeeze risk entirely, which tells a PM you have never actually been short anything.

   Expect next

   • What is the borrow on that name?
   • What is short interest as a percentage of float?
   • What would make you cover?
6. 045How do you attribute a fund's P&L?Performance and alphaIntermediatetechnicalMulti-manager platformsRisk management

   Say this

   Split it into the pieces that correspond to decisions someone made. Market, sector, style factor, then idiosyncratic stock selection, then financing and trading costs. The residual after the systematic pieces is the only part that is evidence of skill.

   Then walk it

   1. Start with market: beta-adjusted net exposure times the index return. That is the part you would have earned with no stock selection at all.
   2. Then sector or industry: the net weight in each industry times that industry's return relative to the index. This catches the PM who is really making a sector call and calling it stock picking.
   3. Then style factors from the risk model: growth, value, momentum, size, quality, volatility. Each has a net loading and a factor return, so each has a P&L line.
   4. Then the residual, which is stock selection. Split it long and short, because a book that makes all its money on the long side in a rising market has not demonstrated a short-selling capability and that matters for how much gross it should run.
   5. Then the costs that are frequently ignored: borrow fees, dividends paid on shorts, financing spread on the leverage, and realised trading cost. On a 300 percent gross book these can be well over a hundred basis points a year.
   6. The caveat to state: attribution is model dependent and the pieces do not add up cleanly. There is always an interaction and residual term, factor returns are estimated, and a PM can dispute the classification of a name. So I would use it to ask questions rather than to settle them.

   Where candidates lose it

   Attributing to positions rather than to risks. Listing the top five winners and losers is a P&L report, not an attribution. The point is to isolate whether the money came from decisions the PM is paid for. Remember to include financing and borrow costs; candidates almost always leave them out and they are large on a levered book.

   Expect next

   • What if all the alpha is on the long side?
   • How would you attribute a macro book instead?
   • What are the biggest cost lines on a levered long-short book?
7. 077Walk me through your research process on a new name.Compliance and research processIntermediatetechnicalLong-short equityMulti-manager platforms

   Say this

   I work backwards from the question that decides the stock. Understand the business and what the price implies, find the one or two variables the thesis turns on, then spend almost all the time on those. The goal is not to know everything, it is to have an edge on the thing that matters.

   Then walk it

   1. Day one is the price. What does today's valuation imply about growth, margin and duration? A reverse DCF or an implied-multiple check tells me what I have to disagree with, which stops me doing a month of work on a stock that is fairly priced.
   2. Then the primary documents: the last three annual reports, the segment notes, the accounting policies, and the last eight quarterly transcripts read back to front so I can see which promises were kept. Filings before sell-side notes, always.
   3. Then the industry structure. Who are the competitors, where does the profit pool sit, who has pricing power, what are the barriers, what is the customer's alternative. This is where the durability question gets answered and it is what business judgement actually means.
   4. Then identify the crux and state it as a question with a number attached. 'Does gross margin reach 42 percent by 2028?' Then the work plan follows: channel checks, pricing data, competitor disclosure, supplier commentary, whatever bears on that number specifically.
   5. Then build the model to the drivers, put consensus next to my numbers, and write a one-page thesis with the variant view, the catalyst, the bear case with a price, the falsifiers and the sizing. If I cannot write it in one page, I have not found the crux.
   6. Then the falsification step, which is the part that separates research from advocacy: go and find the best bear argument, ideally from someone short the name, and see if it survives. And say the honest limitation, that time is the constraint, so tier three names get a screen and a model rather than this whole process.

   Where candidates lose it

   Describing a linear process that ends with a recommendation. A hedge fund process starts with what the price implies and converges on one crux. Candidates who say 'read the 10-K, build a model, do comps, make a recommendation' have described a training programme, not a research process. Name the crux and the falsification step.

   Expect next

   • How long does that take and what do you cut when you have two days?
   • What is the crux on a name you are following now?
   • How do you find the best bear argument?

Firm tags come from public, anonymous candidate reports on Wall Street Oasis: strong signal, not sworn testimony. Firms are named as the places a question was reported, not as partners of Fin Maverick. Answers are written for this page to show how to think out loud; they are not scripts to recite.