Prediction error of the expected claims development result in the chain ladder method

Using the chain ladder method we estimate the total ultimate claim amounts at time I and time I + 1 (successive best estimate predictions for the ultimate claims when updating the information from time I to time I + 1). The claims development result is then defined to be the difference of these two best estimators. We analyze the volatility of this updating procedure for both the process variance and the estimation error in the parameters of the chain ladder model.