HI,

I am working with a maximum likelihood that I estimated using maxlik and bootstrap using Maxboot but my end game is to bootstrap elasticities at the mean value.

My question is : how do I bootstrap elasticities calculated from maximum likelihood function estimates that also take account each re sample changes the mean value?

## 1 Answer

0

Without some change in MaxBoot's source code, it won't be able to produce the elasticities you want. You could modify MaxBoot's code, but it might be easier to write your own code for this. Use the function extsmpl() in a loop to extract a random subsample, estimate the coefficients using Maxlik, and in this case you'll have access to the mean of the subsample. It'll look something like this:

results = {};

for i(1,100,1);

call exctsmpl("mData","sData",100);

z = loadd("sData");

{ x,f,g,cov,ret } = maxlik(z,0,&lfct,x0);

// x is vector of ML estimates of bootstrap sample

mn = meanc(z); // means of bootstrap sample

// compute bootstrap elasticity

// and concatenate to results

endfor;

// compute mean and covariance matrix of

// bootstrapped elasticities from results

## Your Answer

## 1 Answer

Without some change in MaxBoot's source code, it won't be able to produce the elasticities you want. You could modify MaxBoot's code, but it might be easier to write your own code for this. Use the function extsmpl() in a loop to extract a random subsample, estimate the coefficients using Maxlik, and in this case you'll have access to the mean of the subsample. It'll look something like this:

results = {};

for i(1,100,1);

call exctsmpl("mData","sData",100);

z = loadd("sData");

{ x,f,g,cov,ret } = maxlik(z,0,&lfct,x0);

// x is vector of ML estimates of bootstrap sample

mn = meanc(z); // means of bootstrap sample

// compute bootstrap elasticity

// and concatenate to results

endfor;

// compute mean and covariance matrix of

// bootstrapped elasticities from results