Fitting Tweedie's Compound Poisson Model to Insurance Claims Data: Dispersion Modelling

By Gordon K. Smyth, Bent Jørgensen. ASTIN Bulletin: The Journal of the International Actuarial Association, 2002

217 citations recorded by the source corpus.

Subfields: Demography, Economics and Econometrics, Management Science and Operations Research

View Paper: https://www.semanticscholar.org/paper/936a26249e2f07bf9622d13354013814ba6c0532

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Demography

1. Mario V. Wüthrich
2. Enrique ter Horst
3. Abel Rodríguez
4. Gareth William Peters
5. Pavel V. Shevchenko
6. Peter K. Dunn
7. Claudia Czado
8. Wagner Hugo Bonat

In Economics and Econometrics

1. Mario V. Wüthrich
2. Gareth William Peters
3. Pavel V. Shevchenko
4. Wagner Hugo Bonat
5. Enrique ter Horst
6. Abel Rodríguez
7. Peter K. Dunn
8. Claudia Czado

In Management Science and Operations Research

1. Mario V. Wüthrich
2. Gareth William Peters
3. Pavel V. Shevchenko
4. Enrique ter Horst
5. Abel Rodríguez
6. Peter K. Dunn
7. Claudia Czado
8. Bent Jørgensen

Citations in the corpus

• Nested generalized linear mixed models: an orthodox best linear unbiased predictor approach (2007, background): “In addition, it may be applied to positive continuous data with a positive probability component at zero, such as car insurance data (Jørgensen and Souza, 1994; Smyth and Jørgensen, 2002) or customer expenditure data.”
• Extended Poisson–Tweedie: Properties and regression models for count data (2016, background): “The compound Poisson distribution is a frequent choice for the modelling of non-negative data that has a probability mass at zero and is highly right-skewed (Smyth and Jørgensen, 2002).”
• Bayesian dynamic density estimation (2008, methodology): “In the context of actuarial science applications, Bayesian parametric and nonparametric methods have proven useful in modeling insurance claim distributions (Reiss and Thomas 1999; Smyth and Jorgensen 2002; Fellingham and Kottas 2007).”
• Spatial modelling of claim frequency and claim size in non-life insurance (2007, background): “Another approach, which also does not include a separate analysis of claim size and frequency, is given by Jørgensen & de Souza (1994) and Smyth & Jørgensen (2002).”
• Series evaluation of Tweedie exponential dispersion model densities (2005, background): “These distributions have been called " compound Poisson " by Bar-Lev and Stramer (1987), Feller (1968, Section XII.2), Jørgensen and Paes de Souza (1994) and Smyth and Jørgensen (2002) and " compound gamma " by Johnson and Kotz (1970).”
• Flexible Tweedie regression models for continuous data (2016, background): “Following Bonat and Jørgensen [10] and Jørgensen and Knudsen [11], the Pearson estimating function for the dispersion parameters has the form Uqλ(λ,β) = ( n∑ i=1 W iδ[(yi − μi)2 − Ci] , n∑ i=1 W ip[(yi − μi)2 − Ci] ) , withW iδ = −∂C−1i /∂δ andW ip = −∂C−1i /∂p.”
• Model Uncertainty in Claims Reserving within Tweedie's Compound Poisson Models (2009, methodology): “Among the most popular methods there is the chain-ladder method, the Bornhuetter-Ferguson method and the generalized linear model methods.”
• Claims Reserving Using Tweedie's Compound Poisson Model (2003, methodology): “Choosing a suitable parametrization, we estimate the parameters of our model within the framework of generalized linear models (see Jørgensen-de Souza [2] and Smyth-Jørgensen [8]).”