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. 004Correlations rise in a crisis. What does that do to a portfolio built on historical correlations?Portfolio theoryHardsuperdayMulti-assetRisk management

   Say this

   It means your risk model understates exactly the scenario you care about. A portfolio that looks like it has 10 percent volatility in normal times can behave like a levered single bet in a stress event, because the diversification you paid for disappears when you need it.

   Then walk it

   1. The mechanism is common ownership and common funding. When investors deleverage, they sell what is liquid regardless of fundamentals, so cross-asset correlations converge toward one for days or weeks.
   2. The quantitative effect is non-linear. Push a 0.3 correlation to 0.8 in a two-asset 50/50 portfolio with 15 percent vols and volatility goes from about 12 percent to about 14 percent. Do it across ten sleeves and the risk number is out by far more.
   3. So the fix is not a better historical estimate. It is to run correlation stress scenarios explicitly: what does this portfolio lose if everything except cash and government bonds goes to 0.9 correlated?
   4. Practically that means holding some assets whose diversification is structural rather than statistical. Cash, short-dated government bonds, and options have a payoff profile that does not depend on a correlation estimate holding.
   5. It also means sizing on the stressed correlation, not the average one. If your equity and credit sleeves behave as one thing in a crisis, size them as one thing.
   6. And a caveat the other way: correlations also mean-revert. Selling diversifiers after a crisis because they failed in the crisis is how investors end up buying the next one back at a worse price.

   Where candidates lose it

   Treating the correlation matrix as a fact rather than an estimate with a regime. The expected answer names the funding and deleveraging mechanism, and gives one structural hedge whose payoff does not depend on the matrix. Saying 'I would use a longer history' misses the point, because the average includes the calm periods you are not worried about.

   Expect next

   • Which diversifiers actually worked in 2022?
   • How would you build a correlation stress test?
   • Does gold diversify?
2. 005Is volatility the right measure of risk for a long-term investor?Portfolio theoryHardsuperdayAsset managementPension and endowment investing

   Say this

   It is measurable, comparable and mathematically tractable, which is why the whole apparatus is built on it. But it is the wrong risk for most real investors, whose actual risk is failing to meet an obligation, and whose actual pain is drawdown and its path.

   Then walk it

   1. What volatility gets right: it is the only input that makes portfolio mathematics work, it is estimable from short samples, and over long horizons it does correlate with the losses that hurt.
   2. First problem: it is symmetric. Upside surprises raise measured risk, which is nonsense for an investor who only minds losing money. That is the case for Sortino and semi-variance.
   3. Second problem: it ignores the path. Two funds with identical volatility and identical annualised return can have a 15 percent and a 45 percent maximum drawdown, and only one of those gets redeemed by clients.
   4. Third problem: it is blind to illiquidity and to smoothed valuations. Private credit marked quarterly reports a beautiful volatility number, and that number is an artefact of the marking process, not of the risk.
   5. The institutional answer is that risk is mission failure: for a pension it is the funding ratio falling, for an endowment it is the spending rule becoming unsustainable, for an individual it is running out of money. All of those are liability-relative, and volatility of assets alone does not measure them.
   6. So in practice I would report volatility because everyone speaks it, and manage to drawdown, funded status and liquidity, because that is what actually ends mandates.

   Where candidates lose it

   Answering only 'no, drawdown matters more' without conceding what volatility does well. The strong answer is that volatility is the tractable proxy that makes the maths work, then names three specific failures, and lands on liability-relative risk. Also do not say private assets are low risk because their reported volatility is low; interviewers are listening for that exact mistake.

   Expect next

   • So what would you report to a pension trustee board?
   • How do you measure the risk of an illiquid asset?
   • Does a long horizon actually make equities safer?
3. 008If CAPM is right, why does the low-volatility anomaly exist?Factor modelsHardsuperdayFactor investingQuantitative asset management

   Say this

   Because most investors cannot use leverage, so they buy high beta instead. That bids up risky stocks and leaves low beta ones cheap, which flattens the security market line and hands a risk-adjusted premium to whoever can lever a low-vol portfolio.

   Then walk it

   1. The empirical fact first: portfolios of low volatility or low beta stocks have delivered similar or better absolute returns than high beta ones, with far lower risk, in most markets over long samples.
   2. Explanation one, leverage constraints. A long-only fund that needs 10 percent expected return cannot lever a defensive portfolio, so it buys high beta. Frazzini and Pedersen's betting-against-beta trade is exactly this: long low beta levered up, short high beta.
   3. Explanation two, benchmark-relative incentives. A manager judged against an index measures risk as tracking error, and holding low beta stocks in a rising market is career risk. So the natural buyer of defensives is absent.
   4. Explanation three, behavioural. Lottery preference. Retail flows chase high volatility, high attention names, which is a persistent bid on the expensive side of the trade.
   5. The catch for implementation: much of the apparent premium is a bet on interest rates and on the sector composition, because low-vol screens load on staples, utilities and quality. Unhedged, you own a duration proxy, which is why low-vol struggled in 2022.
   6. So I would run it sector-neutral and rate-aware, and I would size it knowing the premium is mostly risk reduction rather than higher absolute return. That is still a good trade for an investor who cares about drawdown.

   Where candidates lose it

   Describing the anomaly and stopping. The question asks why, and the answer the interviewer is listening for is leverage constraints plus benchmark-relative incentives. Also flag the hidden rates and sector exposure, because a candidate who thinks low-vol is free risk reduction has not looked at 2022.

   Expect next

   • How would you neutralise the sector exposure?
   • Why did low volatility fail in 2022?
   • Is this arbitraged away now that everyone knows about it?
4. 009How would you define the quality factor, and why is it harder to pin down than value?Factor modelsHardsuperdayFactor investingQuantitative asset management

   Say this

   Quality is a composite, usually profitability, stability and low leverage, sometimes with accruals and payout added. It is harder to pin down than value because there is no single accounting ratio that defines it, so every provider builds it differently and the definitions disagree.

   Then walk it

   1. The most defensible single component is gross profitability, revenues minus cost of goods over assets, from Novy-Marx. It is the version with the strongest out-of-sample record and it is nearly orthogonal to value.
   2. A standard composite adds return on equity, earnings stability, low accruals, low financial leverage and low net issuance. AQR's quality-minus-junk uses profitability, growth, safety and payout.
   3. Why it is slippery: value has one axis, price relative to a fundamental. Quality has no price in it at all, it is a description of the business, so you can construct it a dozen ways and get low correlation between the versions.
   4. That creates a real problem for an allocator. Two quality ETFs can have a correlation of active returns below 0.6, so 'I have quality exposure' does not tell you what you own until you look at the definition.
   5. The economic case is that markets underprice persistence. High-return businesses fade more slowly than the market assumes. The counter-case is that quality is partly just the residual of a long bull market in growth names.
   6. The interaction that matters most: quality without a valuation constraint is how you end up paying 45 times earnings for consumer staples. Quality at a reasonable price, essentially profitability combined with value, is the more robust construction and it is what most serious factor books run.

   Where candidates lose it

   Giving one ratio, usually ROE, as the answer. ROE alone is leverage-contaminated, which is why safety and low leverage sit in the composite. The stronger answer names gross profitability specifically and then says the honest thing, that quality definitions disagree with each other.

   Expect next

   • Why is gross profitability preferred to net margin?
   • How do you stop a quality screen becoming expensive growth?
   • Would quality work in India, where promoter-held companies dominate?
5. 010Explain how you would construct a factor, and then how you would optimise that construction.Factor modelsHardtechnicalACAQR Capital ManagementInvestment Research · New York · 2021

   Say this

   Build the raw signal first, rank the universe, neutralise the exposures you do not want, then form the long-short portfolio. Optimising it is mostly about the same premium surviving costs: slower rebalancing, buffer zones, better weighting and trading the cheap end of the signal.

   Then walk it

   1. Start with signal hygiene: point-in-time data with no look-ahead, lagged fundamentals by the real reporting delay, winsorise outliers, and standardise to z-scores within the universe.
   2. Neutralise deliberately. Sector, country, size and beta neutrality strip out exposures you are not trying to own. A naive value screen in India is mostly a bet on PSUs and metals.
   3. Then the weighting. Quintile sorts are the academic default but they throw away information. A signal-weighted or optimised portfolio captures more of the spread for the same turnover, so long as you cap single-name weights.
   4. Optimisation is largely cost optimisation. Buffer the rebalance so names near the boundary do not churn, trade over multiple days, net the new signal against the existing book rather than liquidating, and rebalance on a schedule that matches the signal's decay. Momentum decays in weeks so it must be traded fast; value decays over years so daily rebalancing is pure cost.
   5. Then combine signals before trading, not after. Running separate value and momentum sleeves means you buy and sell the same name twice. Integrating the scores into one target portfolio cuts turnover materially, often by a third.
   6. And measure the optimisation honestly. Every knob you turn is another degree of freedom, so I would judge the final construction on out-of-sample and out-of-region performance net of realistic costs, not on the backtest Sharpe.

   Where candidates lose it

   Describing quintile sorts as if that were the whole job. At a firm like this the interesting content is implementation: turnover, buffering, signal integration and cost. Also, do not forget point-in-time data. A candidate who backtests on restated fundamentals has produced a number that cannot be traded.

   Expect next

   • How much would costs eat from a monthly rebalanced momentum factor?
   • How do you decide the rebalancing frequency?
   • How do you avoid overfitting when you have tuned this many parameters?

   Reported by candidates at AQR Capital Management (Investment Research, New York, 2021). Source: Wall Street Oasis.

6. 011Factors decay after publication. How do you decide whether a factor is still investable?Factor modelsHardsuperdayQuantitative asset managementFactor investing

   Say this

   Separate three explanations for weak recent performance: it was never real, it was real and has been arbitraged, or it is real and simply cheap right now. The test is whether the economic mechanism still exists and whether the spread has widened or narrowed.

   Then walk it

   1. First, the base rate. Published anomaly returns fall by something like half after publication in McLean and Pontiff's work, and part of that decay is just the original result being overfitted.
   2. Was it ever real? Check whether it replicates out of sample, in other regions, and with sensible construction choices. If the premium only exists in US small caps before 1990 with one specific definition, it was a data artefact.
   3. Has it been arbitraged? Look at the money in it and at the crowding: assets in the strategy, the correlation of factor returns with flows, and whether the short leg has become expensive to borrow. Arbitrage shows up as lower premium and higher correlation across implementations.
   4. Or is it just cheap? This is the crucial distinction. The value spread, the valuation gap between the cheap and expensive legs, was at extreme wides in 2020 and value then delivered strongly for three years. Poor past returns with a wide spread is a buying condition, not evidence of decay.
   5. Then the mechanism test. A premium that is compensation for a risk or that exists because of a structural constraint on other investors, like leverage limits or benchmark-relative mandates, is much more durable than one with only a behavioural story.
   6. The practical conclusion: I would keep the small number of factors with strong priors, value, momentum, quality, low risk, carry, size them modestly, and refuse to time them aggressively, because factor timing on valuation spreads has a poor live record even though it looks good in backtests.

   Where candidates lose it

   Treating recent underperformance as proof of decay. The single most valuable distinction here is between a factor that is dead and a factor that is cheap, and the evidence for that is the valuation spread between the legs. Candidates who cannot make that distinction would have sold value in 2020.

   Expect next

   • Would you time factors on their valuation spread?
   • How would you measure crowding?
   • How many factors would you actually run?
7. 012You regress a fund's returns on factors and the OLS assumptions are violated. How would you fix it?Factor modelsHardtechnicalACAQR Capital ManagementInvestments · Greenwich · 2022

   Say this

   Diagnose which assumption broke, because the fix differs. Heteroskedasticity and autocorrelation do not bias the coefficients, only the standard errors, so you correct the errors. Omitted variables and endogeneity bias the coefficients themselves, and that needs a change of specification.

   Then walk it

   1. Heteroskedasticity, which is guaranteed in financial returns because volatility clusters: keep OLS coefficients and use White or Newey-West robust standard errors. Or model the variance directly with GARCH if you care about the conditional risk.
   2. Autocorrelated residuals, common when the fund holds illiquid or stale-priced assets: Newey-West with a sensible lag, and separately run the Dimson or Scholes-Williams correction with lagged market returns, because stale marks understate beta. Summing the lagged betas is often the real finding.
   3. Multicollinearity between factors, for example value and profitability after 2015: coefficients stay unbiased but become unstable and standard errors blow up. Orthogonalise the factors, drop one, or use ridge rather than pretending the loadings are precise.
   4. Omitted variable, which is the dangerous one: a fund with apparent alpha against a three factor model often loses all of it against a five factor model with momentum. That is bias, not inefficiency, so the fix is a better specification, not better standard errors.
   5. Non-normal residuals and fat tails: OLS is still consistent, but inference on short samples is unreliable, so I would bootstrap the confidence intervals rather than trusting t-statistics from 36 monthly observations.
   6. And the structural one nobody tests for: parameter instability. A fund's betas shift with regime, so I would run rolling windows or a Kalman filter rather than assuming one constant loading over ten years. A single full-sample regression on a manager who changed style is a meaningless number.

   Where candidates lose it

   Reaching straight for robust standard errors as a universal fix. Robust errors do nothing about omitted variables or endogeneity, which is where the real inference error lives in fund regressions. Separate 'the coefficient is wrong' from 'the standard error is wrong' out loud, and mention stale pricing if the fund holds anything illiquid.

   Expect next

   • How would you detect stale pricing in a fund's returns?
   • What would you do with only 36 monthly observations?
   • How do you test whether the betas are stable?

   Reported by candidates at AQR Capital Management (Investments, Greenwich, 2022). Source: Wall Street Oasis.

8. 016Does tactical asset allocation add value?Asset allocationHardsuperdayMulti-assetAsset allocation

   Say this

   On average, no. The evidence on discretionary market timing is poor and the fee and cost drag is certain. Where it has some support is systematic, valuation and momentum based tilts, run at small size around a strategic policy, with a hard discipline on when the view expires.

   Then walk it

   1. The problem is breadth. A tactical allocator makes a handful of independent bets a year, so even with a genuinely good hit rate, the fundamental law says the information ratio will be small. An equity manager making hundreds of decisions has a structural advantage.
   2. The evidence: most tactical funds underperform a static policy mix of the same risk, and dispersion between them is wide, which is what luck looks like. GTAA as a category has not delivered a persistent premium.
   3. What has some support: long-horizon valuation signals, CAPE-style, with a five to ten year horizon rather than a twelve month one; cross-asset momentum and trend following, which has a real and well-documented premium; and carry.
   4. Implementation matters more than the signal. Tactical shifts are cheapest expressed in futures and overlays, not by trading the underlying sleeves, and the cost of moving 5 percent of a large portfolio through cash equities can eat the whole expected edge.
   5. Governance is the thing that actually kills it. A committee that takes a view, sees it go against them for two quarters and reverses is guaranteed to lose money. Pre-commit to the horizon, the size and the exit condition.
   6. So my answer: keep the tactical range narrow, plus or minus 5 percentage points, run it systematically where possible, budget it explicitly against tracking error, and measure it separately so you can see whether it has earned anything. Most of the time the honest finding is that it has not.

   Where candidates lose it

   Enthusiastically saying yes and describing how you would read the macro. Interviewers in multi-asset seats have seen the attribution and know tactical is usually a small negative. The credible answer concedes the base rate first, then names the specific systematic signals that have evidence, then talks about governance and implementation cost.

   Expect next

   • What signals would you actually use?
   • How would you size a tactical tilt?
   • How would you measure whether the tactical overlay has added value?
9. 029Why is mean-variance optimisation so unstable in practice, and what do you do about it?Portfolio constructionHardsuperdayMulti-assetQuantitative asset management

   Say this

   Because it is an error maximiser. The optimiser takes your most uncertain input, expected returns, and deliberately loads up on whichever asset has the highest estimate, so estimation error gets amplified rather than diversified. Small input changes produce enormous weight changes.

   Then walk it

   1. The mechanics of the failure: the solution involves inverting the covariance matrix, and with correlated assets that matrix is near-singular, so tiny differences in expected returns produce huge long-short positions in similar assets.
   2. Scale of the problem: Michaud's error maximisation and Chopra and Ziemba's work put the damage from expected return error at roughly ten times that of variance error and a hundred times that of covariance error. So the input you know least is the one that matters most.
   3. Fix one, better inputs. Shrink expected returns toward the mean or toward equilibrium, shrink the covariance matrix, impose a factor structure, and use longer histories for the covariance and shorter for nothing.
   4. Fix two, change the objective. Minimum variance, maximum diversification and risk parity avoid expected returns entirely, which removes the worst input at the cost of implicitly assuming returns are proportional to risk.
   5. Fix three, constrain and resample. Position bounds and asset class ranges cap the damage. Resampled efficiency, running the optimisation on many bootstrapped input draws and averaging the weights, produces far more stable portfolios than the point solution.
   6. Fix four, start from the market. Black-Litterman reverse-engineers the returns implied by market weights and then only tilts where you have a view with a confidence attached. That is the cleanest answer because it makes the default a sensible portfolio rather than a corner solution.
   7. The practical endpoint: most serious multi-asset houses run an optimiser as a diagnostic and a sense check, and set the actual policy mix with judgement, constraints and scenario testing. Saying that is more credible than claiming you trust the solver.

   Where candidates lose it

   Saying 'garbage in, garbage out' and leaving it there. The interviewer wants the specific mechanism, error maximisation via matrix inversion, the relative importance of the inputs, and at least two named remedies. Naming Black-Litterman or resampling turns a textbook answer into a practitioner's one.

   Expect next

   • Explain Black-Litterman then.
   • Why does minimum variance behave better out of sample?
   • What does resampling actually do to the weights?
10. 030Explain Black-Litterman and what problem it solves.Portfolio constructionHardsuperdayMulti-assetQuantitative asset management

    Say this

    It fixes the expected return input. Instead of forecasting returns from scratch, you reverse-engineer the returns implied by market capitalisation weights, treat those as the neutral prior, then blend in your own views with an explicit confidence. The output is a portfolio that only deviates where you actually have a view.

    Then walk it

    1. Step one, reverse optimisation. Take market weights, the covariance matrix and a risk aversion parameter, and solve backwards for the expected returns that would make market weights optimal. Those are the equilibrium returns.
    2. Step two, state views as portfolios, not as point forecasts of everything. A view can be absolute, 'EM equity returns 7 percent', or relative, 'European equities beat Japanese by 2 percent', and you attach a confidence, effectively a variance, to each.
    3. Step three, Bayesian blend. The posterior expected returns are a precision-weighted average of the equilibrium prior and your views. High confidence pulls the posterior toward your view; low confidence leaves it near equilibrium.
    4. Step four, optimise on the posterior. Because you started from market weights, no view means you end up at market weights, and one view produces a tilt concentrated in that view rather than a corner solution across twenty assets.
    5. Why that matters practically: it converts 'I am mildly bullish Japan' into a defined, sized, explainable deviation. Every position in the output can be traced to a view or to equilibrium, which is enormously easier to govern than an unconstrained optimiser's output.
    6. The limitations, which I would name: the tau and confidence parameters are subjective and the answer is sensitive to them, market cap weights as equilibrium are questionable for asset classes like bonds and private markets, and it still needs a covariance matrix. It makes the fragile input manageable rather than removing it.

    Where candidates lose it

    Describing it vaguely as 'combining views with the market'. The two mechanical steps that must be there are reverse optimisation from market weights to get the prior, and views expressed with a confidence that determines how far the posterior moves. Also concede that tau is a fudge factor, because interviewers who have implemented it know it is.

    Expect next

    • How do you set the confidence on a view?
    • What is the equilibrium portfolio for a bond allocator?
    • How is this different from just constraining the optimiser?

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.