TAC Method for Fitting Exponential Autoregressive Models and Others: Applications in Economy and Finance

There are a couple of purposes in this paper: to study problem approximation with exponential functions and show its relevance for economic science. The solution the first is as conclusive it can be: working max-norm, we determine which datasets have best by means exponentials form f(t)=b+aexp(kt), give necessary sufficient condition some a,b,k∈R be coefficients that approximation, limits when dataset cannot approximated an exponential. For usual case, also been able approximate approximation. As second purpose, how models science (this only applying R-package nlstac) use autoregressive models, another well-established model science, utilizing same tools: numerical algorithm fitting patterns without initial guess designed authors implemented nlstac. We check one more time robustness successfully two very distant areas economy: demand curves nonlinear series. This shows utility TAC (Spanish CT scan) highlights what extent useful.