internet. For the institution I work for, I need to investigate economic data. What I was asked for is that based on any historical economic data of the country, I need to build "base","pessimistic" and "optimistic" scenarios and assign them probabilities to occur based on data. For example, I used quarterly gdp growth values for 1999-2023 and then used the random picking method 10.000 times. I said if GDP growth is below 0.5 this is a "pessimistic" scenario and has % a 7.5 probability of occurring, if growth is between 0.5 and 4 this is the base scenario and has %85 probability. If it is greater than 4 it is an "optimistic" scenario. Basically, I need to divide the economy into 3 sections with probabilities. However, with assigned probabilities numbers will change, plausibly. These probabilities will pave the way for the forecast of the country.
My boss wants me to include other variables such as CDS, interest rates, industrial production, etc. But how can I build a multivariate simulation model and then say if CDS>x and Growth>y this is a pessimistic scenario and it has the probability of "Y" occurring? Or do I need to use monte carlo simulation, can there be any other statistical methods?
You can use Monte Carlo/bootstrap, data fitting, QQ-plots, density estimation tests, kernel estimation, etc. There are a lot of methods but some will work better than others depending on what your data looks like. Do you have different data for the other variables that need to be included? I would suggest to work through them one by one.