Quant Research & Quant Trading

The probability core quant interviews lean on — distributions, conditioning and expected value under uncertainty.

• ▸Conditional expectation and the law of total expectation
• ▸Common distributions and when each appears on the desk
• ▸Reasoning about variance, covariance and independence

Black-Scholes intuition and the Greeks as risk sensitivities you can explain, not just recite.

• ▸From no-arbitrage to the Black-Scholes formula
• ▸Delta, gamma, vega and theta as desk risk levers
• ▸Put-call parity and replication arguments

Implied vs realised volatility, skew and term structure, and what the surface tells a trader.

• ▸Implied volatility and the smile/skew
• ▸Term structure and what shifts it
• ▸Estimating realised vol from returns

How orders, books and execution actually work — the mechanics behind microstructure questions.

• ▸Limit order books and queue priority
• ▸Spread, depth and market impact
• ▸Adverse selection and signalling

NumPy and SciPy patterns for pricing, simulation and statistics under interview time pressure.

• ▸Vectorised pricing and Monte Carlo
• ▸scipy.stats for distributions and fits
• ▸Writing clear, testable numerical code

Loading, joining and aggregating trade and market data the way an analyst would.

• ▸Joins, groupby and rolling windows
• ▸Resampling and time-series alignment
• ▸Spotting and handling bad data

Query patterns for prices, trades and reference data, including window functions.

• ▸Window functions for running metrics
• ▸Joining trades to reference data
• ▸Aggregations that match desk reports

An optional track for dev-leaning quant roles: structuring a small pricer in modern C++.

• ▸Headers, types and numerical care
• ▸A binomial tree from scratch
• ▸Where C++ matters on quant desks

Black-Scholes Pricer

Finish a European option pricer, derive d1/d2 and return the Greeks for a given parameter set.

Vol Surface Analysis

Clean a quotes file, build an implied-vol surface and comment on skew and term structure.

VaR Backtest Starter

Implement a historical VaR estimate and backtest exceptions against realised PnL.