LDF Curve Fitting is a stochastic reserving method. It assumes that the age to ultimate loss development is
under a function (could be an Inverse Power function) with the ultimate loss. Then the model utilizes some
interpolation and extrapolation to smooth out the LDF curve, especially the "tail side", as you can see from
the sample picture below. The LDF Curve Fitting model may help people to avoid possible over-Parameterizations. But,
please use with caution that, it is critical for you to choose the right growth function to fit your business.
ReserveMaster incorporates this method by implementing Nelder-Mead Maximum Likelihood Fitting engine.
The system provides this methodology as a "side-comparison" to its three estimation-rich packages.
LDF Curve Fitting detail theories, methods and Excel samples can be found from
LDF Curve Fitting paper by Dave Clark (2003).
Here is some screen outputs from ReserveMaster.