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. 041What is the difference between alpha and beta, and how do you actually separate them?Performance and alphaIntermediatetechnicalMulti-manager platformsAsset management

   Say this

   Beta is the return you get for taking market risk, which anyone can buy for a few basis points. Alpha is what is left after you strip out every systematic exposure you were paid to take. You separate them by regressing the return stream on the factors and looking at the intercept.

   Then walk it

   1. The single-factor version: regress fund returns on market returns. The slope is beta, the intercept is alpha, the residual is unexplained. A fund with 0.6 beta in a market up 20 percent earned 12 points from beta before any skill.
   2. The problem is that a single factor flatters almost everyone. Add size, value, momentum, quality and low volatility and a lot of claimed alpha becomes a known factor tilt. That is the entire point of the Fama-French and Carhart extensions.
   3. For hedge funds you go further and add the exposures specific to the strategy: credit spreads, volatility, term premium, and for merger arb a factor that is essentially short the market's tail. Fung and Hsieh style models exist precisely because hedge fund returns are non-linear.
   4. The practical version inside a fund is attribution rather than regression. Decompose today's P&L into market, sector, style and residual, and only the residual counts as stock picking.
   5. The subtlety worth saying: alpha is model dependent. It is defined as what your factor model cannot explain, so a better model shrinks it. Alpha today is often just a factor nobody has named yet, which is roughly the history of quantitative finance.
   6. And beware short samples. With three years of monthly data the standard error on alpha is large enough that a 3 percent annual alpha is statistically indistinguishable from luck. Saying that is what separates a candidate who has run the regression from one who has read about it.

   Where candidates lose it

   Defining alpha as 'outperformance versus the index'. That is beta plus a benchmark choice. Alpha is residual to a factor model, which means it depends on which factors you include, and any honest answer says so. Also mention the sample-size problem: it is the fastest way to sound like you have handled real return data.

   Expect next

   • How many years of data do you need to prove an alpha is real?
   • Is a small-cap tilt alpha or beta?
   • How would you decompose a long-short fund's returns?
2. 042What is the Sharpe ratio, and what are its limitations?Performance and alphaCorephone / first roundAsset managementMulti-manager platforms

   Say this

   Excess return over the risk-free rate, divided by the volatility of that excess return. It is return per unit of risk, and the limitation is that it defines risk as volatility, which is the wrong definition for anything with a skewed or illiquid payoff.

   Then walk it

   1. Rough benchmarks to have in your head: a long-only equity index sits around 0.4 to 0.5 over the long run, a decent hedge fund 0.8 to 1.2, a platform at the fund level 2 or more because of diversification across pods, and anything claiming 4 over a long period needs explaining.
   2. Annualisation matters and gets fumbled. Multiply the monthly mean by 12 and the monthly standard deviation by the square root of 12. That scaling assumes independent returns, which is exactly what fails for illiquid books.
   3. Limitation one, symmetry. Volatility punishes upside surprise as much as downside. A fund whose good months are huge looks worse than a fund grinding out the same return, which is backwards for an investor.
   4. Limitation two, and this is the big one: a strategy that sells tail risk has a beautiful Sharpe until the tail arrives. Writing out-of-the-money options or running a levered convergence trade manufactures a high Sharpe by hiding the risk in the third and fourth moments.
   5. Limitation three, smoothing. Illiquid positions marked on stale prices have artificially low measured volatility, which inflates the ratio. The tell is high autocorrelation in monthly returns, and I would check that before believing any private-credit or distressed Sharpe.
   6. So I would look at Sharpe alongside skew, kurtosis, worst drawdown, time to recover and return autocorrelation. Sortino and Calmar cover part of the gap, and neither fixes the fundamental point that one number cannot describe a return distribution.

   Where candidates lose it

   Forgetting the risk-free rate in the numerator, or annualising by multiplying volatility by 12. Then, on limitations, giving only the symmetry point. The tail-selling and the stale-marks problems are what an allocator actually worries about, and naming autocorrelation as the diagnostic is the detail that lands.

   Expect next

   • How would you detect a fund that is selling tail risk?
   • What does high autocorrelation in monthly returns tell you?
   • What is the difference between Sharpe and Sortino?
3. 043What is the difference between the Sharpe ratio and the information ratio?Performance and alphaIntermediatetechnicalAsset managementQuantitative hedge funds

   Say this

   Sharpe measures excess return over cash divided by total volatility. Information ratio measures excess return over a benchmark divided by tracking error. Sharpe asks whether the whole portfolio was worth owning; information ratio asks whether the active decisions added value.

   Then walk it

   1. The numerator differs in what you subtract. Sharpe subtracts the risk-free rate. IR subtracts the benchmark, so all the market return is stripped out and only the active bet remains.
   2. The denominator differs too. Sharpe uses the volatility of total excess return; IR uses tracking error, the volatility of the difference from the benchmark.
   3. For a genuinely market-neutral hedge fund the two converge, because cash is effectively the benchmark. For a long-only manager they are very different: a fund can have a strong Sharpe from owning equities and a terrible IR from poor stock selection.
   4. The fundamental law of active management lives here: IR is roughly the information coefficient times the square root of breadth. It tells you a manager can raise IR either by being more skilful per decision or by making more independent decisions, which is the entire argument for systematic investing.
   5. Practical benchmark: a long-only manager with an IR of 0.5 sustained over ten years is genuinely good. Most active managers do not clear it after fees, which is why the fee compression happened.
   6. The shared weakness: both use volatility as risk and both need long samples to say anything. And IR is hostage to the benchmark choice, so a manager can improve their IR by picking a benchmark they are structurally tilted away from. Always ask what the benchmark is before reading the number.

   Where candidates lose it

   Saying they are basically the same thing, or getting tracking error and volatility confused. The distinction is what you subtract and what you divide by, and the good answer adds the benchmark-gaming point. For a market-neutral fund saying they converge shows you understand the definitions rather than memorising two formulas.

   Expect next

   • For a market-neutral fund, which would you quote and why?
   • What is a good information ratio over ten years?
   • How can a manager game their information ratio?
4. 046What is a good hit rate, and why is that not the right question?Performance and alphaIntermediatetechnicalLong-short equityGlobal macro

   Say this

   Around 55 percent is genuinely good for a fundamental book, and many excellent macro managers run below 50. It is the wrong question on its own because P&L is hit rate times average win against miss rate times average loss. Expectancy, not accuracy, is what pays.

   Then walk it

   1. Write the identity out: expectancy equals p times average win minus one minus p times average loss. You can be right 40 percent of the time with a three to one win-loss ratio and run a very good book.
   2. That is exactly the trend-following and macro profile. Lots of small losses cut quickly, a few large winners held. The discipline is entirely in the size of the losses, not in the frequency of the wins.
   3. Fundamental long-short is the other shape: higher hit rate, smaller average win, because positions are researched and sized before entry rather than added to as they work.
   4. So the diagnostic questions are better than the hit rate. What is my average winner versus average loser, do I cut losers faster than winners, and am I adding to winners? A high hit rate with a losing book means you take profits early and let losses run, which is the most common retail-shaped mistake in a professional seat.
   5. Sample size again. Twenty positions tells you nothing about hit rate. You need hundreds of independent decisions before the number means anything, which is another reason systematic books can measure themselves and discretionary books mostly cannot.
   6. The honest limitation: hit rate is easy to game by how you define a decision. Was a position you added to three times one call or four? Without a consistent definition, the number is a story rather than a statistic.

   Where candidates lose it

   Quoting a high hit rate as though it were the goal. Interviewers use this to see whether you think in expectancy. Also watch the trap in the other direction: saying hit rate does not matter at all. It matters, it just has to be read with the win-loss ratio alongside it.

   Expect next

   • What is your win-loss ratio on your own investing?
   • Why do macro managers tolerate a sub-50 percent hit rate?
   • How would you know if you were cutting winners too early?

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.