FRM笔记Notes..docVIP

STYLEREF 标题 1 \* MERGEFORMAT Module 2 Market RiskPAGE PAGE 12Module 1 Quantitative AnalysisVery useful data (one-tailed)P(Zz)5% 2.5%1%0.5%z 1.645σ 1.960σ2.326σ2.575σPart 1 Basic ProbabilitiesP(A∪B)=P(A)+P(B)-P(A∩B)P(A)= P(A∩A1)+ P(A∩A2) +……+P(A∩An)Bayes’s TheoremB1, B2, ……. Bn → A or NOT A, Permutation: Combination: Joint Probability: f(y|x) = f(x,y) / f(x)Chebychev’s Inequality: ≥1-1/k2Var (X)=E[(X-EX)2]=E(X2)-(EX) 2Var(X±Y)=Var(X)+Var(Y) ± 2*Cov(X, Y)(Generally N, rather than N-1)Cov(X, Y)= E[(X-EX)(Y-EY)]=E(XY)-E(X)E(Y)E(XY)= E(X)E(Y)+ Cov(X, Y)Cov(X, Y+Z)= Cov(X, Y) +Cov(X, Z)Understanding basic concepts:Volatility: Not necessarily refer to variance. Whether it refers to variance or sd depends on the unit: $ or $2, % or %2. In GARCH(1,1), it often stands for variance, while stands for sd in VAR calculating.Sd: It can be in the absolute form (in $), or in relative form (in %). In marginal VAR math, it is usually relative, which is commensurate with R calculation.Skewness=Excess kurtosis= (0: leptokurtic; 0: platykurtic)Mode is found first: corresponding to the peakMedian: area to the left = to the right. It is pulled in the direction of the skewMean is affected more than median by the skew, or, Median is always between Mode Mean.MeanVarianceContinuous uniformf(x)=1/(b-a)(a+b)/2(b-a)2/12BinomialnpNpqPoissonSample mean :(in bracket: finite population w/o replacement)μ, Proportions(binomial Population)p, Difference: S1-S2 (of sample statistic from 2 populations)Differences: S is sample meanPortfolio Default ModelFuture loss of any asset i is a random variable denoted by Xi=bi/N, where bi is Bernoulli variable with a probability of pi of being 1 (stands for default). Ebi=p, Var(bi)=pq.Expected total loss = ΣXi = N * p/N = pUnexpected total loss (assume independence) =[Var(ΣXi)]1/2 = [N * pq/N2]1/2 = [pq/N]1/2Supposing a portfolio of 2 assets with default correlation ρ=cov(b1, b2)/pq：Var (X1+X2) = Var(X1)+Var(X2)+2Cov(X1,X2)=pq/4 + pq/4 + 2*ρ*pq/4 = pq (1+ρ)Thu