0
$\begingroup$

Say I have a portfolio of 3 stocks $A,B,C$ with $\mu_A = 5%$, $\mu_B = 10%$, $\mu_C = 15%$ and volatility $\sigma_A = 10%$, $\sigma_B = 15%$, and $\sigma_C = 25%$. Let us also say that correlations are $\rho_{AC} = 0.7$, $\rho_{AB} = 0.3$, and $\rho_{BC} = -0.1$. Say total portfolio value is 1 and it is composed of $A,B,C$ equally by value. How would I calculate the corresponding risk exposure that I have to each of the three underlying securities?

Portfolio $\mu_{total} = \frac{1}{3} \times \mu_A + \frac{1}{3} \times \mu_B + \frac{1}{3} \times \mu_C$.

Portfolio $\sigma_{total} = \sqrt{\frac{1}{9}(\sigma_A^2+\sigma_B^2+\sigma_C^2 + 2\rho_{AC}\sigma_A\sigma_C+2\rho_{AB}\sigma_A\sigma_B+2\rho_{BC}\sigma_B\sigma_C)}$

How would you divide up $\sigma_{total}$ or is it not possible?

$\endgroup$
2
$\begingroup$

You can do 2 things:

  1. incremental risk: Calculate the volatility with the asset and with the asset replaced by cash. The difference gives you the (non-linear) incremental risk contribution of the asset. They don't sum up to $\sigma$.

  2. contributions to volatility (Euler allocation)

As $\sigma = \sigma^2/\sigma$ you can define risk contributions by $$ \frac{w_i cov(r_i,r)}{\sigma}, $$ where $w_i$ is the weight of asset $i$, $r_i$ is its return (as a random variable) and $r$ is the portfolio return (with the asset $i$ weighted by $w_i$). If you add these quantities: $$ \sum_{i=1 }^n \frac{w_i cov(r_i,r)}{\sigma} = \frac{\sum_{i=1 }^n w_i cov(r_i,r)}{\sigma} = \frac{cov(\sum_{i=1 }^n w_i r_i,r)}{\sigma} = \sigma^2/\sigma = \sigma. $$ This can also be seen as $w_i \frac{\partial \sigma}{\partial w_i}$ where $\frac{\partial \sigma}{\partial w_i}$ is called the marginal contribution to volatility.

$\endgroup$
  • $\begingroup$ Whats the intuition of using method 1 vs 2? 2 seems more of a mathematical construct such that you get sum of individual risk components add up to overall portfolio volatility $\endgroup$ – bob May 22 '16 at 20:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.