search

INET-Complexity / isle

11 stars 20 forks source link

Total collapse of the insurance market #48

Open jsabuco opened 6 years ago

jsabuco commented 6 years ago

With a high probability in the runs of the model there will be a total collapse of the insurance market. The way of recovery of the market seems unrealistic in the model. Basically after every total collapse there is a transient period until the insurance market is completely recovered since the market entry of insurance firms is made at random every time step. The recovery time is not fixed and depends of the parameter settings.

All this might introduce some noise when aggregating over several realizations of the model.

x0range commented 6 years ago

As discussed earlier, I can see 4 ways to deal with this none of which is very good:

  1. Leave it as it is and allow the simulation to reset (Problem: contamination of ensemble simulations with unrealistic data)
  2. Stop simulation run and remove from ensemble (Problem: biased sampling)
  3. Stop simulation run, leave in ensemble but only up to collapse (Problem: declining sample size with time)
  4. Stop simulation run, remove from ensemble, but record number of removed runs (Problem: 1. complicated to implement, 2. may lead to either empty ensembles or infinite loops if failed runs are replaced)