Korea Financial Risk Manager flashcards

32 cards. Read the term and the definition here. The deck at the top of this page is optional.

Coherent risk measure

A risk measure satisfying monotonicity, translation invariance, positive homogeneity, and subadditivity. Expected shortfall is coherent; VaR generally fails subadditivity when combining portfolios.

Systematic risk

Risk that cannot be diversified away because it affects all assets, such as interest rate or recession risk. It is priced in equilibrium and measured by beta.

Sharpe ratio

Portfolio excess return divided by total standard deviation of returns. It measures reward per unit of total risk and suits undiversified portfolios.

Jensen's alpha

Realized portfolio return minus the CAPM-predicted return from beta times the market risk premium. Positive alpha indicates outperformance after adjusting for systematic risk.

Risk appetite vs tolerance

Risk appetite is the aggregate amount of risk a firm is willing to accept to pursue objectives. Risk tolerance is the acceptable deviation around that appetite.

Portfolio variance formula

For two assets: w1^2*sigma1^2 + w2^2*sigma2^2 + 2*w1*w2*rho*sigma1*sigma2. Correlation drives the diversification benefit; lower rho reduces portfolio variance.

Heteroskedasticity consequence

Non-constant error variance leaves OLS coefficient estimates unbiased and consistent but makes conventional standard errors invalid, so robust or White standard errors are used.

Adjusted R-squared

A goodness-of-fit measure that penalizes additional regressors by degrees of freedom. Unlike R-squared, it can fall when an irrelevant variable is added.

Omitted variable bias

Bias in estimated coefficients when a relevant regressor is excluded and is correlated with an included regressor. The bias direction depends on both correlations.

Unit root

A property of a non-stationary series whose shocks persist indefinitely. Differencing once typically produces a stationary I(0) series.

Central limit theorem

The sampling distribution of the mean approaches normality as sample size grows, regardless of the population distribution, provided variance is finite.

Forward versus futures

Forwards are customized OTC contracts settled at maturity with counterparty risk. Futures are standardized, exchange-traded, cleared, and marked to market daily.

Put-call parity

For European options on non-dividend stocks, C - P = S - K*e^(-r*T). Violations permit arbitrage using the underlying and a risk-free bond.

Interest rate swap

An agreement exchanging fixed for floating interest payments on a notional principal. Its value equals the difference between the present values of the two legs.

Repo transaction

A collateralized short-term borrowing where the seller transfers securities and repurchases them at a higher price, the difference being implied interest.

Credit default swap

A contract transferring default risk from protection buyer to protection seller. The buyer pays periodic premiums; the seller pays upon a credit event.

Cost-of-carry model

Forward price F = S*e^((r - q + c)*T), where q is dividend or convenience yield and c is storage cost. It links spot and forward prices by arbitrage.

Binomial option pricing

Prices options by risk-neutral discounting over an up and down move. The risk-neutral probability p = (e^(r*T) - d)/(u - d).

Black-Scholes-Merton assumptions

Lognormal returns, constant volatility and rates, no arbitrage, continuous trading, and frictionless markets. The call formula is C = S*N(d1) - K*e^(-r*T)*N(d2).

Option delta and gamma

Delta is the change in option value per unit change in underlying price. Gamma is the change in delta per unit change in price, largest near the money.

Historical simulation VaR

Reprices the current portfolio under past risk-factor changes. It captures fat tails and non-linearity but depends heavily on the chosen historical window.

GARCH(1,1) model

Models conditional variance as omega + alpha*epsilon^2 + beta*sigma^2. It captures volatility clustering, with persistence measured by alpha plus beta.

Monte Carlo simulation

Generates random paths for risk factors from assumed distributions and reprices the portfolio. It is flexible for non-linear instruments but computationally intensive.

Value at risk

The loss threshold not exceeded with a stated probability over a stated horizon. A 99% one-day VaR of 10 means losses exceed 10 on about 1% of days.

Expected shortfall

The probability-weighted average loss beyond the VaR threshold. It is a coherent risk measure and captures tail severity that VaR ignores.

Basel VaR multiplier

Internal-model capital uses the greater of previous-day and 60-day average VaR, multiplied by at least 3, at 99% confidence over 10 days.

Backtesting traffic light

Basel classifies 250-day VaR exceptions into green, yellow, and red zones. More exceptions raise the capital multiplier and trigger supervisory action.

Square-root-of-time scaling

Scales VaR by the square root of the horizon, for example one-day to ten-day VaR by sqrt(10). It assumes independent and identically distributed returns.

Expected loss formula

Expected loss equals probability of default times loss given default times exposure at default. Unexpected loss is the volatility around this expected value.

Merton structural model

Treats equity as a call option on firm assets with debt as the strike. Default occurs when asset value falls below the default barrier at maturity.

Distance to default

Measures how many standard deviations asset value is from the default barrier. It equals (ln(V/D) + (mu - 0.5*sigma^2)*T)/(sigma*sqrt(T)).

Wrong-way risk

Occurs when exposure to a counterparty rises as the counterparty's credit quality deteriorates, so losses cluster in default scenarios.