Portfolio Management interview preparation

Questions

100

Traced to a firm

40

Firms

24

Updated

September 2026

Asked at

All firmsBLBlackRock4Vanguard4WMWellington Management4Amundi3ACAQR Capital Management3Neuberger Berman3SCSchroders3Man Group2MSCI2Northern Trust2AllianceBernstein1Apollo Global Management1Blackstone1BMBNY Mellon1Carlyle Group1Fidelity Investments1Goldman Sachs1Invesco1Millennium Management1MSMorgan Stanley1NUNuveen1PIMCO1SSState Street1TPTPG1

Topic

All topicsPortfolio theory5Factor models8Asset allocation11Rebalancing3Portfolio construction7Benchmarks and tracking error5Performance measurement8Risk management6Fixed income and LDI5Currency and global3Implementation and costs5Active versus passive6India markets7Brainteasers5Career and fit16

Level

AnyCoreIntermediateHard

Type

AnyTechnicalCaseMarket viewFitBrainteaser

1. 002Draw me the efficient frontier and tell me what every part of that picture means.Portfolio theoryCorephone / first roundAsset managementMulti-asset

   Say this

   Volatility on the x axis, expected return on the y axis. The cloud of possible portfolios has an upper-left edge, and that edge is the frontier. Anything below it is dominated, because you can get the same return for less risk.

   Then walk it

   1. Plot every combination of your assets. The set is bounded on the left by the minimum variance portfolio, which is the leftmost point on the curve.
   2. The efficient frontier is only the part above that point. Below the minimum variance portfolio the curve bends back, and those portfolios are strictly worse, more risk for less return.
   3. Add a risk-free asset and you get a straight line from the risk-free rate that is tangent to the frontier. That is the capital allocation line, and the tangency point is the maximum Sharpe ratio portfolio.
   4. That is the important bit, because it says everyone should hold the same risky portfolio and simply vary how much cash or leverage they put against it. Risk appetite changes the position on the line, not the mix.
   5. The frontier moves out when you add a genuinely new asset class with low correlation, which is the real argument for adding alternatives, and it moves in when you add constraints.
   6. The caveat I would say unprompted: the frontier is drawn from estimates. Redraw it with five years of different data and the shape moves a lot, so nobody trades the tangency point literally.

   Where candidates lose it

   Drawing the whole ellipse and calling all of it the frontier. Only the upper branch, above the minimum variance portfolio, is efficient. Also, get the axes the right way round: volatility is horizontal, return is vertical.

   Expect next

   • Where does a 60/40 portfolio sit on that picture?
   • What moves the frontier outward?
   • Why does everyone not just hold the tangency portfolio?
2. 025What is your rebalancing policy, and why that one?RebalancingCorephone / first roundMulti-assetWealth management

   Say this

   Tolerance bands with an annual review, rather than pure calendar rebalancing. Bands trade only when the portfolio has actually drifted, which is when rebalancing matters, and they avoid the pointless turnover of trading every quarter because the date changed.

   Then walk it

   1. Rebalancing exists to control risk, not to add return. Left alone, a 60/40 portfolio drifts toward equity because equity compounds faster, so after a long bull run you are running far more risk than the policy you signed.
   2. Calendar rebalancing, monthly or quarterly, is simple and auditable but it trades when nothing has changed and it does not trade when something has. Its worst feature is that it is blind to the size of the drift.
   3. Tolerance bands, say plus or minus 5 percentage points on equities or 25 percent of the sleeve weight, trade only on material drift. Vanguard's own research on this finds the choice between sensible rules barely changes returns but changes costs a lot, so I would optimise for cost and governance.
   4. The practical hybrid most institutions actually run: check monthly, trade only if a band is breached, and rebalance back to the edge of the band rather than all the way to target, which cuts turnover again.
   5. Then the free rebalancing. Use cash flows first: direct new contributions and coupons to the underweight sleeve, and take withdrawals from the overweight. In a portfolio with regular flows that does most of the job at zero cost.
   6. And use derivatives for the fast part. Equity futures can restore the policy beta in a day while the underlying sleeves are traded slowly, which is how large funds rebalance without paying market impact on billions.

   Where candidates lose it

   Saying 'rebalance annually' with no reasoning, or claiming rebalancing raises returns. Its primary job is risk control. If you claim a return benefit, be ready to explain the rebalancing premium properly, because that is the follow-up and a vague answer there unwinds the whole response.

   Expect next

   • Does rebalancing actually add return?
   • What band width would you use?
   • How would you rebalance a taxable portfolio?
3. 035How do you choose a benchmark for a mandate, and what makes a benchmark bad?Benchmarks and tracking errorCorephone / first roundAsset managementInstitutional asset management

   Say this

   A good benchmark is investable, unambiguous, specified in advance, and it represents the manager's actual opportunity set. A bad one is either unachievable, like a fixed 8 percent hurdle, or so different from the portfolio that the excess return measures style rather than skill.

   Then walk it

   1. The standard criteria, from Bailey: unambiguous, investable, measurable, appropriate to the manager's style, reflective of current investment opinion, specified in advance, and owned by the manager in the sense that they accept it.
   2. Investable is the one most often breached. A benchmark including names that cannot be bought in size, or an index with a 15 percent single-stock weight that breaches the fund's diversification rules, sets the manager an impossible task.
   3. Appropriate to the style matters most for attribution. Measure a small cap value manager against a broad large cap index and the excess return is mostly the size and value factor, not the manager. You will fire them for style and hire the next one at the wrong point in the cycle.
   4. Bad benchmark type one, absolute hurdles: 'cash plus 5 percent' is fine as an objective but useless as a benchmark, because it gives no information about whether the manager did well in the environment they faced.
   5. Bad benchmark type two, peer group medians. They are not investable, they are survivorship biased, they are only known after the fact, and they encourage herding. Useful as context, wrong as a benchmark.
   6. In practice I would use a market index matched to the opportunity set, with a custom or blended index where the mandate spans regions or asset classes, and I would state the currency hedging convention in the benchmark definition. Hedged versus unhedged is worth several percent a year and it is astonishing how often that is left vague.

   Where candidates lose it

   Listing the textbook criteria without saying which ones actually get breached. Concrete failures are what earn the marks: peer medians, absolute hurdles, style mismatch, and an unspecified currency hedging convention. That last one is a detail interviewers in global mandates notice immediately.

   Expect next

   • How would you benchmark a multi-asset fund with no natural index?
   • Should the benchmark be hedged or unhedged?
   • What is wrong with a peer group benchmark?
4. 040What is the Sharpe ratio, and where does it mislead you?Performance measurementCorephone / first roundAsset management

   Say this

   Excess return over the risk-free rate divided by the volatility of that excess return. It is the standard measure of return per unit of total risk. It misleads whenever the return distribution is not symmetric, when returns are smoothed, or when the strategy sells tail risk.

   Then walk it

   1. Compute it on excess returns, not raw returns, and annualise properly: multiply the mean by the number of periods and the standard deviation by the square root of that number. A monthly Sharpe times twelve is wrong by a factor of about 3.5.
   2. Failure one, skew. A strategy that sells options or credit protection has a lovely Sharpe ratio right up to the day it does not. Volatility does not see the left tail, so option selling and carry trades look better than they are.
   3. Failure two, smoothing. Illiquid or appraisal-priced assets have autocorrelated returns, which suppresses measured volatility. A private credit fund reporting a Sharpe of 2 is often reporting the marking policy, not the risk. Correct for it by using the Lo adjustment or by summing lagged betas.
   4. Failure three, the horizon and the sample. Sharpe ratios are noisy: distinguishing a Sharpe of 0.5 from 1.0 with confidence takes many years of data, so ranking managers on three-year Sharpe is close to ranking noise.
   5. Failure four, it ignores whether the risk was systematic. A levered index fund has a decent Sharpe and no skill in it, which is why Sharpe is the wrong tool for evaluating an active manager inside a benchmarked mandate. Information ratio is the right one there.
   6. So I would quote Sharpe as a summary, then show Sortino for asymmetry, maximum drawdown for path, and the return distribution's skew and kurtosis. And if the assets are illiquid, I would say explicitly that the volatility is understated.

   Where candidates lose it

   Forgetting the risk-free rate in the numerator, or annualising by multiplying the ratio rather than scaling mean and standard deviation separately. On the substance, the failure interviewers most want to hear is that Sharpe rewards selling tail risk. A candidate who praises a hedge fund with a Sharpe of 3 and no questions about skew has failed the question.

   Expect next

   • How would you adjust Sharpe for illiquidity?
   • Sharpe or information ratio for an active equity manager?
   • What Sharpe ratio would make you suspicious?
5. 046Sharpe ratio or information ratio, which matters more for an active manager?Performance measurementCorephone / first roundAsset managementInstitutional asset management

   Say this

   Information ratio, if the manager has a benchmark. It measures excess return per unit of active risk, which is exactly what the client is paying for, whereas Sharpe includes the market exposure the client could have bought for a few basis points.

   Then walk it

   1. Definitions side by side: Sharpe is excess return over cash divided by total volatility. Information ratio is excess return over the benchmark divided by tracking error.
   2. The reason IR is the right one for a benchmarked mandate: a long-only equity manager's Sharpe is dominated by the equity risk premium, so in a good decade every manager looks skilled and in a bad one every manager looks useless. IR strips the market out.
   3. Typical scale is worth knowing so you can talk about it credibly. An IR of 0.5 over a full cycle is genuinely good and roughly top quartile, 0.75 is excellent, and anything above 1 sustained over a long period is rare and worth questioning.
   4. Sharpe is the right measure when there is no benchmark: absolute return funds, multi-asset total return, and anything where the client's alternative is cash.
   5. The connection to portfolio construction: IR is what the fundamental law predicts, skill times root breadth, and it is what a tracking error budget is denominated in. If I know a manager's IR I can say how much excess return I should expect for the tracking error I am giving them, which is the arithmetic of hiring them.
   6. The shared caveat: both are estimated from short samples and both are noisy, so I would look at IR alongside consistency, the proportion of rolling periods above benchmark, rather than treating a single point estimate as a fact.

   Where candidates lose it

   Treating them as interchangeable or picking Sharpe for a benchmarked fund. The decisive point is that Sharpe rewards market beta the client could buy for nothing. Also know realistic magnitudes; a candidate who says an IR of 3 is achievable has never looked at real manager data.

   Expect next

   • What information ratio would you expect from a good manager?
   • When is Sharpe the right measure?
   • How does IR link to the tracking error budget?
6. 063What are the components of transaction cost?Implementation and costsCorephone / first roundAsset managementPortfolio implementation

   Say this

   Explicit costs you can see on the ticket, commission, taxes, exchange fees, and implicit costs you have to measure, spread, market impact, delay and opportunity cost. The implicit ones are usually several times larger than the explicit ones, which is why cost control is a portfolio construction question.

   Then walk it

   1. Explicit: brokerage commission, exchange and clearing fees, and transaction taxes. In India that means STT, stamp duty and GST on brokerage, which together make high-turnover strategies structurally more expensive than in the US.
   2. Spread: you buy at the offer and sell at the bid, so half the spread each way is a cost even for a tiny order. In a liquid large cap that is a couple of basis points; in a small cap it can be 50.
   3. Market impact: your own order moves the price. It scales roughly with the square root of order size relative to average daily volume, so trading 20 percent of a day's volume is far more than four times as expensive as trading 5 percent.
   4. Delay and opportunity cost: the price drift between decision and execution, and the alpha lost on unexecuted quantity. These are invisible in a broker report and often the largest components for an active manager.
   5. Then the structural ones people forget: the cost of crossing the spread on the rebalance of an index at the reconstitution date, when everyone trades the same way at the same time, and the tax cost of realising gains in a taxable portfolio.
   6. Practically I would measure all of it as implementation shortfall against the decision price, break it down by strategy and market, and use it as an input to how fast and how often I am willing to trade. The right target is not zero cost, it is maximum alpha net of cost.

   Where candidates lose it

   Listing only commission and spread. Market impact and opportunity cost are the ones that matter, and the square root relationship between size and impact is the detail that shows you understand why capacity is limited. Giving Indian specifics, STT and stamp duty, is a cheap way to show you know the market you are being hired for.

   Expect next

   • How does impact scale with order size?
   • Which costs get worse in a stressed market?
   • How would you trade a large order in an illiquid stock?
7. 068What is the difference between an ETF and a mutual fund?Active versus passiveCorephone / first roundVanguardGeneralist · Malvern · 2026

   Say this

   Both are pooled vehicles, but an ETF trades on an exchange all day at a market price while a mutual fund transacts once a day at NAV. The structural consequence that matters is the creation and redemption mechanism, which makes ETFs more tax efficient and shifts trading costs onto the person doing the trading.

   Then walk it

   1. Dealing: mutual fund orders are aggregated and struck at one NAV per day. An ETF trades continuously at a price that can sit at a premium or discount, with an authorised participant arbitraging the gap by creating or redeeming baskets.
   2. The in-kind mechanism is the real difference. Redemptions are met by delivering securities to the authorised participant rather than selling them, so the fund does not realise capital gains. In the US that makes ETFs materially more tax efficient than mutual funds, which must distribute realised gains.
   3. Cost incidence: in a mutual fund, one investor's redemption forces the fund to trade and all remaining holders pay the cost. In an ETF, the seller crosses the spread themselves, so long-term holders are insulated. That is a genuine fairness advantage.
   4. Where mutual funds are better: automatic investment plans and fractional amounts, no bid-offer spread for regular small contributions, and no risk of trading at a discount in a stressed market. For a monthly SIP investor a mutual fund is often the better instrument even if the ETF's expense ratio is lower.
   5. In India the differences are sharper. ETF liquidity is thin outside the Nifty and Sensex trackers, so tracking difference and impact cost can exceed the expense ratio saving, and index funds rather than ETFs are usually the better passive vehicle for a retail investor. Institutional flows, particularly EPFO, dominate Indian ETF assets.
   6. And the caveat on stressed markets: an ETF's price is a real-time price, so in a dislocation it can trade well below the stale NAV of an illiquid bond portfolio. That is the ETF telling the truth faster, not the ETF failing, and it is worth being able to say that clearly.

   Where candidates lose it

   Answering only 'ETFs trade intraday'. The substance is the in-kind creation and redemption mechanism and who bears trading costs. And do not claim ETFs are always better; for a regular small contribution plan, and in India where ETF liquidity is thin, an index fund is frequently the right answer.

   Expect next

   • Why is an ETF more tax efficient?
   • What happens when an ETF trades at a discount to NAV?
   • ETF or index fund for an Indian retail investor?

   Reported by candidates at Vanguard (Generalist, Malvern, 2026). Source: Wall Street Oasis.

8. 075What is a PMS, and how does it differ from a mutual fund and an AIF?India marketsCorephone / first roundIndian asset managementIndian wealth management

   Say this

   A PMS runs a separate account in the client's own name with a 50 lakh rupee minimum, so the client owns the securities directly and is taxed on each transaction. A mutual fund is a pooled vehicle taxed at the investor's exit. An AIF is a pooled private fund with a one crore minimum and far wider freedom on strategy and leverage.

   Then walk it

   1. PMS: discretionary or non-discretionary, minimum 50 lakh, securities held in the client's own demat account. Because there is no pooling, every trade the manager makes creates a taxable event for that client, and performance is reported net of the fees that client actually paid.
   2. Mutual fund: pooled, heavily regulated on concentration, liquidity and disclosure, daily NAV, minimum investment of a few hundred rupees, and no tax at the fund level, so gains are only taxed when the investor redeems. That deferral is a real, quantifiable advantage over a PMS for a high-turnover strategy.
   3. AIF: three categories. Category I for venture and infrastructure, Category II for private equity and private credit, which is the largest, and Category III for hedge-fund-like strategies including long-short and leverage. Minimum one crore, taxation depends on category and structure, and Category III has been the fastest growing.
   4. The construction freedom runs the other way to the regulation. A mutual fund has tight single-issuer and single-stock limits. A PMS can run 15 concentrated positions. A Category III AIF can be long-short and levered. So concentration and leverage rise as the investor's ticket size rises.
   5. Fee models differ too: mutual funds have a capped total expense ratio and no performance fee, PMS commonly charges a fixed fee plus a performance fee over a hurdle with a catch-up, and AIFs use private-fund style fees with carry.
   6. The practical comparison I would make to a client is after-tax and after-fee. A PMS with a 20 percent performance fee and full annual taxation of realised gains needs to beat a mutual fund by a significant margin before the client is better off, and the sales pitch rarely presents it that way.

   Where candidates lose it

   Describing PMS as just 'a mutual fund for rich people'. The two structural differences that matter are direct ownership, which changes the tax treatment completely, and the freedom to concentrate. And be ready with the after-tax comparison, because that is the question a real client asks and most candidates have never done the arithmetic.

   Expect next

   • Why does the tax treatment favour a mutual fund for high turnover?
   • What can a Category III AIF do that a mutual fund cannot?
   • Which would you recommend to a client with five crore?
9. 081Two assets each have twenty percent volatility and a correlation of a half. What is the volatility of an equally weighted portfolio?BrainteasersCoretechnicalAsset managementRisk management

   Say this

   About 17.3 percent. Portfolio variance is 0.25 times 400 plus 0.25 times 400 plus 2 times 0.25 times 0.5 times 400, which is 100 plus 100 plus 100, so 300. The square root of 300 is about 17.3.

   Then walk it

   1. Set it up in variance terms, always. Each asset's variance is 400 in percent-squared units, and the covariance is the correlation times the two volatilities, so 0.5 times 20 times 20, which is 200.
   2. Portfolio variance equals w1 squared times var1 plus w2 squared times var2 plus 2 w1 w2 times covariance. That is 0.25 times 400 twice, plus 2 times 0.5 times 0.5 times 200, giving 100 plus 100 plus 100 equals 300.
   3. Square root: 17.3 percent. So combining two identical-risk assets at 0.5 correlation cut risk by about 13 percent of its original level, which is the whole diversification benefit in one number.
   4. Know the boundary cases cold, because they are the follow-up. Correlation 1 gives 20 percent, no benefit at all. Correlation 0 gives 20 over root 2, about 14.1 percent. Correlation minus 1 gives zero, a perfect hedge.
   5. The general result worth having memorised: for n equally weighted assets with equal volatility sigma and common correlation rho, portfolio variance is sigma squared times rho plus 1 minus rho over n. As n goes to infinity, volatility tends to sigma times the square root of rho.
   6. That limit is the useful part. With 20 percent volatility assets at 0.3 average correlation, no amount of diversification gets you below about 11 percent. That is the systematic floor, and it is why diversification stops helping.

   Where candidates lose it

   Averaging the volatilities, or adding them and forgetting the covariance term is multiplied by two. Work in variance, then take the square root at the end. And have the asymptotic result ready, sigma times root rho, because the follow-up is almost always 'and with a hundred assets?'

   Expect next

   • What if correlation were zero, or minus one?
   • What is the limit with a hundred such assets?
   • How does that connect to the diversification floor?
10. 082A fund is up fifty percent one year and down fifty percent the next. What is its average annual return, and what did the investor actually get?BrainteasersCorephone / first roundAsset managementPerformance analysis

    Say this

    The arithmetic average is zero, but the investor is down 25 percent. A hundred goes to 150 then to 75. The gap is volatility drag, and it is the reason geometric return is the only one that describes what an investor experienced.

    Then walk it

    1. Compute it: 1.5 times 0.5 equals 0.75, so terminal wealth is 75 percent of the start. The geometric return is the square root of 0.75 minus one, which is about minus 13.4 percent a year.
    2. The general relationship: geometric return is approximately arithmetic return minus half the variance. Here volatility is enormous, so the drag is enormous, and the approximation is only rough at this size of move.
    3. Practical consequence one: a fund can advertise a positive average annual return while every investor lost money. That is why performance reporting standards require compounded, annualised figures.
    4. Practical consequence two, and this is the portfolio management point: reducing volatility raises compounded return even if you do not change the average. That is the whole mathematical case for risk control, rebalancing and diversification, rather than just a comfort argument.
    5. Put a realistic number on it so it does not sound like a trick. A portfolio with a 7 percent arithmetic return and 20 percent volatility compounds at roughly 5 percent. Cut volatility to 12 percent and it compounds at about 6.3 percent. Same expected return, 130 basis points more wealth every year.
    6. And the asymmetry to name: recovering from a 50 percent loss needs a 100 percent gain. Losses and gains are not symmetric in wealth terms, which is why drawdown control matters more than chasing the last few percent of upside.

    Where candidates lose it

    Saying the average is zero and stopping, or getting the recovery arithmetic backwards. The interviewer is testing whether you instinctively think in compounded terms. Tie it to the portfolio conclusion, that lowering volatility raises compounded wealth for the same average return, or you have answered a maths question rather than an investment one.

    Expect next

    • What return do you need to recover from a 50 percent loss?
    • So how much is volatility worth in compounded terms?
    • Which return would you show a client?

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.