On Prediction of Future Insurance Claims When the Model Is Uncertain

Predictive modeling is arguably one of the most important tasks actuaries face in their day-to-day work. In practice, actuaries may have a number of reasonable models to consider, all of which will provide different predictions. The most common strategy is to first use some kind of model selection tool to select a “best model,” and then use that model to make predictions. However, there is reason to be concerned about the use of the classical distribution theory to develop predictions because these ignore the selection effect. Since accuracy of predictions is crucial to the insurer’s pricing and solvency, care is needed to develop valid prediction methods. In this paper, we undertake an investigation of the effects of model selection on the validity of classical prediction tools and make some recommendations for practitioners.

[1]  Alan J. Miller,et al.  leaps: Regression Subset Selection. , 2004 .

[2]  Robert Tibshirani,et al.  Post‐selection inference for ℓ1 ‐penalized likelihood models , 2016, The Canadian journal of statistics = Revue canadienne de statistique.

[3]  W. Loh,et al.  Consistent Variable Selection in Linear Models , 1995 .

[4]  Ian Duncan,et al.  Testing Alternative Regression Frameworks for Predictive Modeling of Healthcare Costs , 2015 .

[5]  Edward W. Frees,et al.  Predictive Modeling Applications in Actuarial Science , 2014 .

[6]  Gareth W. Peters,et al.  Model Uncertainty in Claims Reserving within Tweedie's Compound Poisson Models , 2009, ASTIN Bulletin.

[7]  H. Leeb,et al.  CAN ONE ESTIMATE THE UNCONDITIONAL DISTRIBUTION OF POST-MODEL-SELECTION ESTIMATORS? , 2003, Econometric Theory.

[8]  B. Efron,et al.  Second thoughts on the bootstrap , 2003 .

[9]  B. Efron.  Estimation and Accuracy After Model Selection , 2014, Journal of the American Statistical Association.

[10]  R. Tibshirani,et al.  A SIGNIFICANCE TEST FOR THE LASSO. , 2013, Annals of statistics.

[11]  Andreas Tsanakas,et al.  Parameter Uncertainty and Residual Estimation Risk , 2016 .

[12]  John A. Nelder,et al.  Generalized linear models. 2nd ed. , 1993 .

[13]  Paul Kabaila,et al.  The Effect of Model Selection on Confidence Regions and Prediction Regions , 1995, Econometric Theory.

[14]  Boualem Djehiche,et al.  Regression modeling with actuarial and financial applications by Edward W. Frees , 2011 .

[15]  Todd A. Kuffner,et al.  On overfitting and post‐selection uncertainty assessments , 2017, 1712.02379.

[16]  Robert Tibshirani,et al.  The Elements of Statistical Learning: Data Mining, Inference, and Prediction, 2nd Edition , 2001, Springer Series in Statistics.

[17]  R. Tibshirani.  Regression Shrinkage and Selection via the Lasso , 1996 .

[18]  Hannes Leeb,et al.  Conditional predictive inference post model selection , 2009, 0908.3615.

[19]  B. M. Pötscher,et al.  MODEL SELECTION AND INFERENCE: FACTS AND FICTION , 2005, Econometric Theory.

[20]  H. Akaike,et al.  Information Theory and an Extension of the Maximum Likelihood Principle , 1973 .

[21]  G. Schwarz.  Estimating the Dimension of a Model , 1978 .

[22]  A. Buja,et al.  Valid post-selection inference , 2013, 1306.1059.

[23]  Jonathan Taylor,et al.  Statistical learning and selective inference , 2015, Proceedings of the National Academy of Sciences.

[24]  Y. Benjamini,et al.  False Discovery Rate–Adjusted Multiple Confidence Intervals for Selected Parameters , 2005 .

[25]  B. M. Pötscher,et al.  CAN ONE ESTIMATE THE UNCONDITIONAL DISTRIBUTION OF POST-MODEL-SELECTION ESTIMATORS? , 2007, Econometric Theory.

[26]  Andrew J. G. Cairns,et al.  A discussion of parameter and model uncertainty in insurance , 2000 .

[27]  Shujuan Huang,et al.  MODEL SELECTION AND AVERAGING OF HEALTH COSTS IN EPISODE TREATMENT GROUPS , 2015, ASTIN Bulletin.

[28]  Brian Hartman,et al.  Model Selection and Averaging in Financial Risk Management , 2013 .

[29]  David Hinkley,et al.  Bootstrap Methods: Another Look at the Jackknife , 2008 .