Calculating Portfolio return from Log-returns

Question

Data available : Daily simple returns of 2 stocks for last 500 days. Weights : Equal weighted i.e 0.50 is the weight of each stock in the portfolio.

I want to calculate daily return of the portfolio and compute the historical Value at Risk of the portfolio. The way I do it is

for (i in 1:500)
{
    preturn[i] = 0.50 * log(1 + s1[i]) + 0.50 * log(1 + s2[i]) 
}

VaR = quantile(preturn, 0.05)    # This is the VaR value
VaR = exp(VaR) - 1     # Converting back to simple return

where s1 and s2 are vectors of daily simple returns of stock 1 and stock 2

My question is whether my approach is proper in calculating the portfolio return and computing the quantile and then finally converting it to simple return.

Answer 1

Your daily return on the equal-weight portfolio is

0.5s1 + 0.5s2

And log-return would be the log(1 + above).