Skip to main content
This page collects equity derivatives questions for quantitative interviews, covering option pricing, Greeks, the volatility surface and forward pricing.
Status of this page. Last checked on 15 November 2026. These answers use the standard textbook framework (Black-Scholes and its extensions). They are not tied to any market’s or exchange’s contract rules. For market details see Hong Kong listed derivatives.

Pricing relationships

For European options with no dividends: C − P = S − K e^(−rT).
  • Continuous dividend yield q: C − P = S e^(−qT) − K e^(−rT)
  • Discrete dividends: replace S with S − PV(dividends)
If it fails, you can lock in a riskless profit with a conversion or reversal. For example, if C − P is above the right-hand side, sell the call, buy the put, buy the stock and borrow cash. In practice, borrow cost, funding rates and bid-ask spreads mean parity holds as a band rather than an equality. Conversely, you can back out the implied borrow cost or implied dividend from call and put prices.
With a continuous dividend yield: F = S e^((r − q) T). With discrete dividends: F = (S − PV(D)) e^(rT).If F is above this value: buy the spot, sell the future and deliver at expiry (cash-and-carry). If F is below: short the spot and buy the future (reverse cash-and-carry). The basis, F minus S, comes from financing cost net of expected dividends and converges to zero at expiry.In practice, borrow cost, funding spreads, dividend uncertainty and margin all widen the no-arbitrage band. This is also the basis for Delta-1 arbitrage between index futures and ETFs. See Delta-1: hedging with listed derivatives.
An option can be replicated with the underlying and cash, and the replication cost is the no-arbitrage price. The replication strategy does not depend on your view of the stock’s drift μ, so μ is replaced by the risk-free rate r in the price.Under the risk-neutral measure Q, price equals the discounted expected payoff: V = e^(−rT) E_Q[payoff]. This does not mean market participants are risk neutral. It is a computational device. Volatility σ stays in the formula because the replication cost depends on how much the path moves.
Call, no dividends: Never exercise early. Doing so gives up time value and the interest on the strike, and the call is worth more than its intrinsic value.Call, with dividends: It can be optimal to exercise just before an ex-dividend date, when the dividend exceeds the remaining time value plus the interest earned on the strike.Put: Early exercise can be optimal, especially when deep in the money and rates are high, because receiving the strike early earns interest while the remaining time value is small.
Higher expected dividends lower call prices and raise put prices, because the stock drops by the dividend on the ex-date and the option holder does not receive it. This also shows up as a lower forward price, and it is why S is replaced by S e^(−qT) in put-call parity.For trading: dividend expectations and dividend risk are separate risk factors, and they remain after delta hedging. See the Dividends row in Delta-1: hedging with listed derivatives.

Black-Scholes and its limits

Assumptions: The stock follows geometric Brownian motion (lognormal) with constant drift and volatility, a constant risk-free rate, continuous trading, no transaction costs, unlimited short selling and no arbitrage.Most commonly violated:
  • Constant volatility: Real markets show a volatility smile or skew, and volatility itself moves randomly over time.
  • Continuous paths: Real prices jump (earnings, events, gaps at the open), and delta hedging cannot fully remove jump risk.
  • No transaction costs and continuous hedging: Discrete hedging and costs leave residual P&L noise.
This is why the market quotes in implied volatility rather than treating Black-Scholes as the true model.
Approximately: P&L ≈ (1/2) Γ S² [ (ΔS / S)² − σ_implied² × dt ].You make money when the realized squared return exceeds the implied variance for that day, that is, when realized volatility is higher than implied volatility. Long gamma loses in the opposite case. This is the core of volatility trading: trading an option is essentially trading the gap between realized and implied volatility.This is an approximation. Vega P&L (the move in implied volatility itself) is separate, and how gamma is distributed across strikes and maturities affects the actual result.
  • Gamma: Largest at the money (ATM). Near expiry, ATM gamma rises sharply (pin risk) while in-the-money and out-of-the-money gamma goes to zero.
  • Vega: Largest near the money and increasing with time to expiry (vega is roughly proportional to √T). Vega goes to zero near expiry.
  • Theta: ATM options decay fastest close to expiry.
So short-dated ATM options are high gamma and low vega, while long-dated options are low gamma and high vega.

Volatility surface

Downside strikes (puts) have higher implied volatility than upside strikes. Common explanations:
  • Hedging demand: Institutions buy puts to protect holdings, which pushes put prices up.
  • Leverage effect: When the stock falls, leverage rises and so does volatility, giving a negative correlation between price and volatility.
  • Crash risk premium: Investors pay a premium for downside tail protection.
Skew shapes differ by underlying, and some single stocks have flatter or even upward-sloping skew.
  • Sticky strike: When the underlying moves, the implied volatility at each fixed strike stays unchanged.
  • Sticky delta (sticky moneyness): Implied volatility moves with moneyness, so when the underlying rises, the implied volatility at a given strike changes with it.
With skew, the two imply different effective deltas. Under sticky delta, a spot move also shifts implied volatility, so the effective delta equals the Black-Scholes delta plus vega × (∂σ / ∂S). Your choice of assumption affects hedge ratios and P&L attribution.
Two basic conditions:
  • Calendar spread: At a fixed strike, total variance σ² × T must increase with maturity, otherwise a calendar spread is an arbitrage.
  • Butterfly: At a fixed maturity, option price must be a convex function of strike, which means the implied risk-neutral density cannot be negative.
Also check that call prices decrease in strike and that the call spread does not exceed the discounted strike difference. When fitting a surface (for example with an SVI-type parameterization), build these constraints into the fit.