You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Define root FinancialMeasure trait, which should facilitate balance sheet modeling implementations.
Flow
Balance
Cashflow <: Flow
All value related functions should use Cashflow{U,T} instead of allowing time and cashflows to be input separately (mostly affects ActuaryUtilities?)
Handling Dates
Ideally proper date handling would be part of this revamp. Timepoint already exists but the use is not clear downstream.
Core Functionality
Instead of present_value, the core function would be value which would have an inner call for all contracts to push the model through the contract to handle any dependencies, e.g. something like:
This would allow for more natural handling of contracts which depend on the economic conditions. Further, if the model is, e.g. stochastic, adding a method for value for the associated model would allow for the logic to synthesize into a scalar value.
How to better expose sensitivity analysis with this? duration is currently solely focused on a scalar discount rate or a single curve, but not if the economic parameters are a basket of curves. Is there some way to use accessors to tell a sensitivity function which parameter to sensitize, with yields just being a special case?
The text was updated successfully, but these errors were encountered:
Define root
FinancialMeasure
trait, which should facilitate balance sheet modeling implementations.Flow
Balance
Cashflow
<:Flow
All value related functions should use
Cashflow{U,T}
instead of allowing time and cashflows to be input separately (mostly affectsActuaryUtilities
?)Handling Dates
Ideally proper date handling would be part of this revamp.
Timepoint
already exists but the use is not clear downstream.Core Functionality
Instead of
present_value
, the core function would bevalue
which would have an inner call for all contracts to push the model through the contract to handle any dependencies, e.g. something like:This would allow for more natural handling of contracts which depend on the economic conditions. Further, if the model is, e.g. stochastic, adding a method for
value
for the associated model would allow for the logic to synthesize into a scalar value.Proof of concept:
Sensitivity Analysis
How to better expose sensitivity analysis with this?
duration
is currently solely focused on a scalar discount rate or a single curve, but not if the economic parameters are a basket of curves. Is there some way to use accessors to tell a sensitivity function which parameter to sensitize, with yields just being a special case?The text was updated successfully, but these errors were encountered: