State Space Models

All state space models are written and estimated in the R programming language. The models are available here with instructions and R procedures for manipulating the models here here.

Tuesday, February 18, 2025

World-System (1960-2010) Mexico Takes a Random Walk

 


In an earlier post (here) I used the MXL20 model to forecast eight futures for Mexico in the 21st Century. The best model was the Business-As-Usual (BAU) Model, however, in the short-run (year-to-year) the Random Walk (RW) actually had the best Akaike Information Criterion (AIC).


The shocks to the state variable MX1 (the Growth Shocks) are presented above. Mostly, the shocks are positive, except for around the turn of the Century (after 2000)*. In the short-run, year-to-year, the Mexican Economy has been able to continuously adjust to positive growth shocks. 



Notes

* The Growth Shocks are the residuals from the MX1_RW model. You should notice that the residuals are not normally distributed as are the residuals in the simulated MX1_RW presented at the beginning of the post. This is why I compute Bootstrap rather than normal distribution statistics.

AIC Statistics





Wednesday, February 5, 2025

World System: Generalizing the Malthusian Controller

 


In an earlier post (here), I presented Leibenstein's Malthusian Error Correcting Controller (ECC). It can be generalized to the dominant ECCs in most theoretical economic models (above).