Interview questions
1,322 questions in the style trading and research firms ask, from probability and brainteasers to options, statistics and coding. Every question is open to read, with its hints. The 50 in the free sample can be answered and marked on the spot, no account needed; open 3 of their worked solutions to try, then all of them with a free account.
496 questions · page 6 of 13 · Clear filters
Random question- 201With orthonormal regressors, the lasso minimising…
AppliedRegression and econometricsCitadelTwo SigmaAQR
- 202With highly collinear regressors, ridge regression often predicts better out of…
AppliedRegression and econometricsCitadelTwo SigmaAQR
- 203An instrument Z has Cov(Z,Y) = 0.6 and Cov(Z,X) = 0.4.
AdvancedRegression and econometricsCitadelTwo SigmaD. E. Shaw
- 204Your first-stage F-statistic is 4. What does that tell you about the IV estimate?
AdvancedRegression and econometricsCitadelTwo SigmaD. E. Shaw
- 205You model trade counts per venue where exposure times differ. What belongs in the model?
AdvancedRegression and econometricsCitadelTwo SigmaD. E. Shaw
- 206Two models are fitted on 100 observations.
AdvancedRegression and econometricsCitadelTwo Sigma
- 207You average 1,000 observations from an AR(1) process with autocorrelation 0.5.
AdvancedRegression and econometricsCitadelTwo SigmaAQR
- 208A design matrix has squared singular values 9, 4 and 1.
AdvancedRegression and econometricsCitadelTwo Sigma
- 209Which of these is closest to being a stationary series?
FoundationTime seriesCitadelTwo Sigma
- 210For white noise, sample autocorrelations have standard error about 1/√T.
FoundationTime seriesCitadelTwo Sigma
- 211When is the AR(1) process xₜ = φxₜ₋₁+εₜ covariance stationary?
FoundationTime seriesCitadelTwo Sigma
- 212A HAR model forecasts tomorrow’s realised volatility as 0.1+0.4 RV_d+0.3 RV_w+0.2 RVₘ.
FoundationTime seriesCitadelTwo Sigma
- 213Your labels are each asset’s return over the next 10 trading days.
FoundationTime seriesCitadelTwo SigmaAQR
- 214You regress one stock’s price on another’s and get R² = 0.95 with a t-statistic of 40.
AppliedTime seriesCitadelTwo SigmaD. E. ShawFree solution
- 215A spread follows Δsₜ = −0.05 sₜ₋₁+εₜ on daily data.
AppliedTime seriesCitadelTwo SigmaPoint72 and CubistFree solution
- 216An AR(1) has φ = 0.8 and innovation variance 1. What is its stationary variance?
AppliedTime seriesCitadelTwo Sigma
- 217Daily equity returns show almost no autocorrelation, but their squares show a lot.
AppliedTime seriesOptiverCitadelTwo Sigma
- 218An augmented Dickey–Fuller test fails to reject its null. What have you learned?
AppliedTime seriesCitadelTwo SigmaD. E. Shaw
- 219For xₜ = εₜ+0.5εₜ₋₁, what is the lag-one autocorrelation?
AppliedTime seriesCitadelTwo SigmaD. E. Shaw
- 220For xₜ = 2+0.6xₜ₋₁+0.3xₜ₋₂+εₜ, what is the unconditional mean?
AppliedTime seriesCitadelTwo SigmaD. E. Shaw
- 221Your prior on a state is 10 with variance 4, and you observe 13 with…
AppliedTime seriesCitadelTwo SigmaJump Trading
- 222What is the leverage effect in equity returns?
AppliedTime seriesCitadelTwo SigmaD. E. Shaw
- 223Using EWMA with λ = 0.94, yesterday’s volatility estimate was 1% and today’s return is 2%.
AppliedTime seriesCitadelTwo SigmaJump Trading
- 224You have ten years of daily data, start with a two-year training window, expand…
AppliedTime seriesCitadelTwo SigmaD. E. Shaw
- 225Daily returns have first-order autocorrelation 0.1 and none beyond.
AppliedTime seriesCitadelTwo SigmaAQR
- 226With T = 100 observations, the sample autocorrelations at lags one and two are 0.2 and 0.1.
AppliedTime seriesCitadelTwo Sigma
- 227An AR(1) process has mean 2 and coefficient 0.8.
AppliedTime seriesCitadelTwo Sigma
- 228For the stationary AR(2) process xₜ = 0.5xₜ₋₁+0.3xₜ₋₂+εₜ, what is the lag-one…
AppliedTime seriesCitadelTwo Sigma
- 229What is the lag-one autocorrelation of the MA(1) process xₜ = εₜ+2εₜ₋₁?
AppliedTime seriesCitadelTwo Sigma
- 230A series’ ACF decays geometrically while its PACF is significant at lags one…
AppliedTime seriesCitadelTwo Sigma
- 231An AR(1) process has coefficient 0.8 and innovation variance 1.
AppliedTime seriesCitadelTwo Sigma
- 232A spread follows an Ornstein–Uhlenbeck process dX = θ(μ−X) dt+σ dW with θ = 5 per year.
AppliedTime seriesJane StreetCitadelTwo Sigma
- 233A daily GARCH(1,1) has ω = 0.000002, α = 0.08 and β = 0.90.
AppliedTime seriesOptiverCitadelTwo Sigma
- 234A GARCH(1,1) has α+β = 0.98.
AppliedTime seriesCitadelTwo Sigma
- 235With the GARCH(1,1) above (ω = 0.000002, α = 0.08, β = 0.90), yesterday’s…
AppliedTime seriesOptiverCitadelTwo Sigma
- 236For a random-walk level observed with noise, what does the Kalman filter’s…
AppliedTime seriesCitadelTwo Sigma
- 237Returns are modelled with a Student-t distribution with 5 degrees of freedom.
AppliedTime seriesCitadelTwo Sigma
- 238Daily returns have volatility 1% and first-order autocorrelation −0.1.
AppliedTime seriesCitadelTwo SigmaAQR
- 239The autocorrelation of absolute daily returns stays positive and decays only…
AppliedTime seriesCitadelTwo Sigma
- 240Out of sample, your return forecast has mean squared error 0.98 while the…
AppliedTime seriesCitadelTwo SigmaAQR