asitav-sen / ifrsassistant Goto Github PK

Currently all the functions run in the shiny runtime. We should check the possibility to host some of the functions as plumber API and monitor the app performance.

EMI (easy monthly instalment) is currently calculated from balance divided by number months of the loan left. However, this may not always be the case. For e.g. what if balance of a defaulted loan reflects outstanding + normal balance? The only way out is to ask for emi data.

Currently forecasting of macroeconomic data is done using autoarima. However, this may not be the best way. Perhaps use an improved exotic model?

Currently scenario generation/monte-carlo simulation is done to estimate possible variations in change of value of collateral (or similar). The the possible range is asked from the user.
When we forecast macroeconomic data, we already get a range of possible outcomes. Perhaps, we can include these possible macroeconomic forecast to generate additional simulation.
The concern however is the computation power requirement. Is it worth it?