0

I have a data frame which could be approximated by the following example:

                 firm       <-seq(c(rep(6,253), rep(7, 252), rep(8,254), rep(9,251))
                 LTD05      <- seq(1, 1010, 1)
                 cap        <- seq(2,1011,1)
                 riskfree   <- c(rep(1,253), rep(2, 252), rep(3,254), rep(4,251))
                 equityvol  <- seq(0.1, 0.101, 0.1)
                 df <- data.frame(firm,LTD05,cap,riskfree,equityvol)

LTD 05 is the liabilities of the company, cap is market equity value, riskfree is the risk free return, equityvol is the equity volatility of the company. I want to estimate asset value and asset volatility for each observation in my sample using numerical approach in the Merton model. I run for loop.

 for( i in 1:nrow(df)) {
 fnewton <- function(x){
 y <- numeric(2)
 d1 <- (log(x[1]/df$LTD05[i])+(df$riskfree[i]+x[2]^2/2))/x[2]
 d2 <- d1-x[2]
 y[1] <- df$cap[i]- (x[1]*pnorm(d1) - exp(-df$riskfree[i])*df$LTD05[i]*pnorm(d2))
 y[2]<-df$equityvol[i]*df$cap[i]-pnorm(d1)*x[2]*x[1]
 y
                      }

 xstart <- c(df$cap[i]+df$LTD05[i],df$equityvol[i]) 

 data$asset_new_value[i]<-nleqslv(xstart, fnewton, method="Newton")$x[1]
 data$asset_new_volat[i]<-nleqslv(xstart, fnewton, method="Newton")$x[2]


 i=i+1
  }

This question has been discussed already : here . However, my problem is that I get negative value for asset volatility (which is not case in reality) when I run this loop. I suspect that I may have bugs in for loop. Could you please give any feedback?

Community
  • 1
  • 1
user3618375
  • 33
  • 7
  • 1
    As a side note: I think you could place the function definition of `fnewton` outside the loop and just pass the current `i` to it inside the loop. This may speed up the loop and make it easier to trace errors. Also, I think your `i=i+1` is not necessary since it is a `for` loop. – talat May 12 '14 at 20:45
  • Your code doesn't run: in `fnewton(...)` you refer to `df$rfabsol[i]` but there is no such column in `df`. – jlhoward May 12 '14 at 21:28
  • @ beginneR Thank you @jhoward Sorry , I edited. – user3618375 May 12 '14 at 21:37
  • I found mistake if anyone will want to use these codes. The problem is that I have not used proper starting value for asset volatility). It should be chosen properly when one deals with numerical approach of Merton model. – user3618375 May 17 '14 at 15:18

0 Answers0