返回题库

最大化夏普比率的最优权重

Maximizing Sharpe Ratio

专题
Finance / 金融
难度
L4

题目详情

有两只相互独立的股票 A、B,其收益分别为 RAN(μ,σ2)R_A\sim\mathcal{N}(\mu,\sigma^2)RBN(μ,2σ2)R_B\sim\mathcal{N}(\mu,2\sigma^2)。两者期望收益相同,但 B 更“风险大”。

你构造组合:在 B 上权重为 ww,在 A 上权重为 1w1-w。假设无风险利率为 0,求使组合夏普比率最大化的最优 ww

You have two independent stocks, A and B. Their returns are RAN(μ,σ2)R_A \sim \mathcal{N}(\mu, \sigma^2) and RBN(μ,2σ2)R_B \sim \mathcal{N}(\mu, 2\sigma^2). They have the same expected return but B is riskier. You form a portfolio with weight ww in Stock B and (1w)(1-w) in Stock A. Assuming a zero risk-free rate, what is the optimal weight ww that maximizes the portfolio's Sharpe ratio?

解析

组合收益 Rp=(1w)RA+wRBR_p=(1-w)R_A+wR_B

期望收益

E[Rp]=(1w)μ+wμ=μ,\mathbb{E}[R_p]=(1-w)\mu+w\mu=\mu,

ww 无关。

由于独立,方差为

Var(Rp)=(1w)2σ2+w22σ2=σ2[(1w)2+2w2].\mathrm{Var}(R_p)=(1-w)^2\sigma^2+w^2\cdot 2\sigma^2=\sigma^2\big[(1-w)^2+2w^2\big].

夏普比率(无风险利率为 0)为

S(w)=μσ(1w)2+2w2.S(w)=\frac{\mu}{\sigma\sqrt{(1-w)^2+2w^2}}.

因此只需最小化 f(w)=(1w)2+2w2=3w22w+1f(w)=(1-w)^2+2w^2=3w^2-2w+1

f(w)=6w2=0w=13.f'(w)=6w-2=0\Rightarrow w=\frac{1}{3}.

最优为 w=13\boxed{w=\frac{1}{3}}(即 A: 2/32/3,B: 1/31/3)。