A Bayesian internal model for reserve risk: an extension of the correlated chain ladder

Ercole, Carnevale Giulio; Paolo, Clemente Gian Article A Bayesian internal model for reserve risk: an extension of the correlated chain ladder Risks Provided in Cooperation with: MDPI – Multidisciplinary Digital Publishing Institute, Basel Suggested Citation: Ercole, Carnevale Giulio; Paolo, Clemente Gian (2020) : A Bayesian internal model for reserve risk: an extension of the correlated chain ladder, Risks, ISSN 2227-9091, MDPI, Basel, Vol. 8, Iss. 4, pp. 1-20, https://doi.org/10.3390/risks8040125 This Version is available at: https://hdl.handle.net/10419/258078 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ risks Article A Bayesian Internal Model for Reserve Risk: An Extension of the Correlated Chain Ladder Carnevale Giulio Ercole 1and Clemente Gian Paolo 2,* 1PartnerRe, Hardstrasse 301, 8005 Zürich, Switzerland; giulio.er[email protected] 2Department of Mathematics for Economic, Financial and Actuarial Sciences, Università Cattolica del Sacro Cuore, Largo Agostino Gemelli 1, 20123 Milan, Italy *Correspondence: [email protected] Received: 28 September 2020; Accepted: 9 November 2020; Published: 19 November 2020   Abstract: The goal of this paper was to exploit the Bayesian approach and MCMC procedures to structure an internal model to quantify the reserve risk of a non-life insurer under Solvency II regulation. To this aim, we provide an extension of the Correlated Chain Ladder (CCL) model to the one-year time horizon. In this way, we obtain the predictive distribution of the next year obligations and we are able to assess a capital requirement compliant with Solvency II framework. Numerical results compare the one-year CCL with other traditional approaches, such as Re-Reserving and the Merz and Wüthrich formula. One-year CCL proves to be a legitimate alternative, providing values comparable with the more traditional approaches and more robust and accurate risk estimations, that embed external knowledge not present in the data and allow for a more precise and tailored representation of the risk profile of the insurer. Keywords: Bayesian models; stochastic claims reserving; claims development results 1. Introduction The most recent regulatory requirements for European insurers, known as Solvency II, introduced the possibility of tailored risk modeling for many sources of risks borne by each undertaking. As an alternative to the Standard Formula, according to its own technical capabilities and upon approval of the supervisory authority, each insurer can develop its own internal model. For a specific risk, or a particular set of risks, such a partial internal model is supposed to capture the insurer’s risk profile in a more consistent way than the Standard Formula, equal for all the market participants. To accomplish this task, actuarial literature produced several models. In this paper, we provide a partial internal model for assessing reserve risk’s capital requirement using Bayesian procedures. In the context of tailored risk modeling, a Bayesian approach has indeed a great potential. In fact, Bayesian techniques are able to include in statistical models external knowledge and subjective judgments that are not necessarily deduced by the data, other than the fact that they require the explicit specification of all the assumptions made. If properly used then, these features enable a better description of the risks faced by an insurer. Anyway, the use of techniques with a Bayesian flavor is not something new for actuarial science (see Klugman (1992)) and with respect to claims reserving can be traced back to Bornhuetter and Ferguson (1972). Despite this, Bayesian models often proved to be too complex to be implemented on real world applications as this kind of modeling often leads to mathematically intractable forms. However, this mathematical complexity can be overcome using Markov Chain Monte Carlo (MCMC) simulation techniques that provide empirical approximations when closed formulas are not available. With respect to claims reserving, the first appearance of MCMC procedures is credited to Ntzoufras and Della Portas in 2002 Ntzoufras and Dellaportas (2002), Risks 2020,8, 125; doi:10.3390/risks8040125 www.mdpi.com/journal/risks Risks 2020,8, 125 2 of 20 which has been followed by many other papers (see, e.g., Peters et al. (2017) and Wüthrich (2007) for recent applications in the field of claims reserves). Concerning claim reserving, one of the most popular deterministic methods is the chain ladder (see, e.g., Friedland (2010) and Hindley (2017) for a review of main deterministic methodologies). As well-known , the ultimate cost of each accident year is predicted using run-off triangles of paid and/or incurred losses and assuming that prior patterns of losses persist also in the future (e.g., in terms of settlement speed or behavior of the case reserve). Several stochastic methods have been proposed in the literature in order to measure a variability of the Chain Ladder methodology ( see, e.g., Mack (1993) and Wüthrich and Merz (2007) ). In this context, it is worth to be mentioned the Correlated Chain Ladder (see Meyers (2015); Frees et al. (2016)) that exploits the advantages of Bayesian models and allows to model a correlation between accident years. In this framework, we move from the Correlated Chain Ladder (CCL) and we provide an extension of this approach in order to model the claim development result distribution in a framework compliant with Solvency II. On a one-year time-frame, this proposal represents an alternative of two classical approaches 1 , Re-Reserving Diers (2009) and Merz and Wüthrich formula Wüthrich et al. (2008), widely used in practice when chain ladder is the underlying deterministic method. However, this approach is not a mere third option. Indeed, Bayesian techniques represent a more refined and sophisticated approach to obtain claims reserves variability, and this is especially true in comparison to bootstrap based algorithms 2 (see, e.g., Hastie et al. (2009) ). In the context of modeling regulatory capital, using a fully specified probabilistic framework, able to integrate external information, provides a more advanced approach than methods that simply fit semi-parametrically the variability on a limited set of observations. Typically, when speed and simplicity are critical, bootstrap estimates could be more appropriate, while when accuracy is important, Bayesian modeling could represent a more consistent choice. With respect to stochastic claim reserving, we think Bayesian techniques can have a role in improving current standards where the accuracy of estimates is critical, like internal modeling practice. This paper is organized as follows. In Section 2, the original CCL model, introduced in Meyers (2015) , is described. In Section 3, we provide the extension of the model in a one-year view. In particular, we adapt the idea of a Bayesian update originally found in Meyers (2017) in order to assess the distribution of claims development result. In Section 4, we develop a case study based on motor third-party liability data. In particular, in Section 4.1, we focus on the assumptions and on the calibration procedure. In Sections 4.2 and 4.3, we provide main results comparing our proposal to classical approaches provided in the literature. We show how one-Year CCL can be a viable alternative to assess a capital requirement compliant with Solvency II regulation. Conclusions follow. 2. The Correlated Chain Ladder In Bayesian claims reserving, inference is performed by computing a posterior distribution of the parameters. This distribution is derived via Bayes theorem by combining a prior, representative of external knowledge, and observations. To this extent, we need a data structure that eases the procedure 1 In this framework, the dynamics of the one-year change in P&C insurance reserves estimation has been also studied in Dacorogna et al. (2018) by analyzing the process that leads to the ultimate risk in the case of “fixed-sum” insurance contracts. 2 Statistical literature drew many parallels between Bayesian MCMC and Bootstrap procedures. In statistical inference, these algorithms represent two different ways offered respectively by the Bayesian and frequentist paradigms to assess parameter uncertainty (see, e.g., Efron (2011); Hastie et al. (2009) ). The Bootstrap algorithm, and especially its non-parametric version, generates many samples from the data, from which is possible to obtain the distribution of a statistic or a parameter, mimicking the Bayesian effect of a posterior distribution. This distribution is produced in a quick and simple way, requiring no probabilistic assumptions, no prior elicitation nor MCMC procedures and can be effectively considered an “approximation of a non-parametric, non-informative posterior distribution” (Hastie et al. 2009). Risks 2020,8, 125 3 of 20 of prior parameters specification. Concerning reserving models, we refer to a cross-classified structure. Given a run-off triangle: D(t)={Xij :i=0, . . . , t;j=0, . . . , t;i+j≤t} where i , j , and t represent the accident year, the development year, and the evaluation period respectively, and being Xij whatever claims figure (typically either incremental or cumulative payments). In general, a cross-classified structure has the following form: Xij =f(αi,βj) Claims figures are thought as a function of parameters, where αi acts as a parameter linked to the accident year and βj as a development parameter. 3 This is a potentially superior informative structure that allows us to link the data to phenomena separately related to either the accident or the development year. With respect to Bayesian inference, this structure is clearly best suited to receive prior information about different factors that affect the reserving process.4 2.1. Model Specification The Correlated Chain Ladder, introduced by Meyers (2015), is basically a re-parameterization of the traditional Chain Ladder model, that exploits exactly a cross-classified structure in order to enhance the possibilities of the traditional algorithm and re-interpret it in a Bayesian context. Given a triangle of cumulative payments: D(t)={Cij :i=0, . . . , t;j=0, . . . , t;i+j≤t} data are assumed generated by a log-normal distribution: Cij ∼logNormalµij,σj for i=0, . . . , tand j=0, . . . , t, where: •µ0j=α0+βj •µij =αi+βj+ρ·log(Ci−1,j)−µi−1,jfor i>0 In this framework, a level parameter αi embeds information about accident year i , βj is a sort of new development factor, that captures information about the development of payments and ρ is a feature that models the correlation between different accident years. This last parameter is one of the most remarkable innovations brought by this model, in fact it relaxes the hypothesis of independence between accident years, one of the main assumptions of several stochastic models related to Chain Ladder (see Mack (1993); Wüthrich et al. (2008)). 2.2. Parameter Specification At this point we can elicit prior distribution for each parameter of the model. This step is critical (see Berger et al. (2009)), as we have to choose reasonable distributions and values that are consistent with our data. As a general case, we have the following: αi∼Normallog(Bi) + [log(elri)],eifor i=0, . . . , t(1) 3In general, parameters could be more than two and left up to the needs of the modeler. 4 Anyway this structure is not exclusive of Bayesian models as it is applied also in frequentist models, as for instance the Over-Dispersed Poisson bootstrap Model England and Verrall (1999). Risks 2020,8, 125 4 of 20 where Bi represents known gross premiums related to the accident year i and [log(elri)] is a further random variable representative of the logarithm of the expected loss ratio for each accident year i , while ei is the precision, rather than the variance, in accordance to the parameterization adopted by most inference engines. It is necessary to further elicit a prior distribution for log(elr)i and as a general case we specify a non-informative uniform prior with parameters γi and δi (as defined in Meyers (2015)): log(elri)∼Uni f ormγi,δi. (2) Obviously, the model can be easily adapted by defining different parameterizations or selecting different distributions 5 . By virtue of the parameterization in (1) and (2), the level parameter αi describes the logarithm of the ultimate cost of the accident year i . This hyperparameter is critical since its mean is defined on the log-space and even a slight variation of the parameters highly affects the final result. As a result of this, we should state a reasonable domain of variation of the parameters that will result in values of the claims reserve consistent with our data. The development parameter βjis defined as a uniform random variable on a negative support: βj∼Uni f orm−η, 0for j=0 , . . . , t−1 (3) with η∈R+. According to this assumption, the parameter βjcan be interpreted as the portion of the ultimate cost paid up to the development year j and hence, it is strictly depending on the settlement speed of the portfolio. For j=t , we set βj equal to 0, assuming that all claims are settled until the development year t . The approach can be easily extended in case the estimation of a tail is needed. Regarding the log-variance, we have: σ2 j= t ∑ h=j τh(4) where τh∼Uni f orm(0, 1). As for the correlation parameter, the choice is still left to the modeler. As a general case, it is possible to set a non-informative uniform prior: ρ∼Uni f orm−1, 1(5) 2.3. Posterior and Predictive Distributions After model specification and prior elicitation then we use an inference engine, like STAN or JAGS. to generate samples from the posterior distribution. This step is critical as well, as we need a strong sample that we can deem truly representative of the posterior distribution of parameters. To this extent, we suggest to sample from n different chains that start from n different points, to allow for a sufficiently long warm-up phase and to set an appropriate thin factor. At this point we have Kparameters sets Θ(k), each one representative of a reserving scenario: Θ(k)={αi}t i=0,{βj}t−1 j=0,{σj}t j=0,ρk with k=1, . . . , K. From the posterior distribution, we can obtain the predictive distribution for each cell of the lower triangle, and in particular for the last column. This will allow us to obtain the predictive distribution of the ultimate cost and thus of the claims reserve. For each k , we sample this ultimate cost from a log-normal: C(k) it ∼logNormalµ(k) it ,σ(k) tfor i=1, . . . , t. (6) 5For instance, different distributions may be chosen for different accident years i. Risks 2020,8, 125 5 of 20 As usual, we get the k -th determination of the predictive distribution of the claims reserve by the following formula: R(k)= t ∑ i=1 C(k) it − t ∑ i=1 Ci,t−i. (7) Iterating the procedure K times, we have a Bayesian predictive distribution of the claims reserve in a total run-off framework. 3. An Extension in a One-Year Framework Now we propose a way to adapt the correlated Chain Ladder on the one-year time horizon in order to describe the reserve risk according to Solvency II guidelines. In this context, a solvency capital requirement (SCR) for reserve risk, at the end of time t, can be derived as: SCR0.995 =VaR0.995 t ∑ i=1 Pi,t−i+1+RD(t+1)!v1−RD(t)(8) where Pi,t−i+1 denotes the incremental payments of calendar year t+ 1 for claims incurred in the accident year i , RD(t) is the best estimate at time t and RD(t+1) is the best estimate at time t+ 1 considering only existing claims at time t . Hence, the capital requirement at a confidence level equal to 99.5% is obtained as the difference between the Value at Risk of the distribution of the next year insurer obligations, opportunely discounted with the risk-free discount factor v1 , and the best estimate 6 at time t . Being RD(t) a known amount at the valuation date, we can directly focus on the sum of next year payments for claims already incurred and the residual reserve that will be posted after experiencing one more year of run-off. Adapting a reserving model on a given time horizon means making an assumption about the payments patterns on such a time horizon and then computing a residual reserve in light of this new information. To this extent, the CCL tries to ground this informative update on a Bayesian approach, making the Bayes theorem a starting point for the determination of the residual reserve after one year of run-off. Main idea is to simulate a set of losses given the parameters sets generated with the MCMC procedure and then re-evaluate the probability of the same sets using the Bayes theorem. Each set could be thought as a different scenario for the reserving process and this effectively means to re-weight probabilities of future scenarios in light of latest realizations. The idea of Bayesian update is originally found in Meyers (2017) and Frees et al. (2016). Here, we adapt it in order to provide an algorithm capable of producing a predictive distribution of the next year obligations over the one-year time horizon. Such a distribution is compliant with Solvency II guidelines and is suitable to represent an internal model for reserve risk. Our goal then will be modeling the following random variable: Y1yr = t ∑ i=1 Pi,t−i+1+RD(t+1). (9) To do this, we perform the following steps: 6 According to Article 76 of the Solvency II Directive European Commission (2009), the claims reserve must be equal to the current amount that insurance and reinsurance undertakings would have to pay if they were to transfer their insurance and reinsurance obligations immediately to another insurance or reinsurance undertaking. This definition leads to a claims reserve evaluated as the sum of the best estimate and risk margin. As prescribed by Solvency II, risk margin is not considered in (8) to avoid problems of circularity. Risks 2020,8, 125 6 of 20 1. The starting point is represented by the K parameters sets Θ(k) simulated with the MCMC procedure. Θ(k)={αi}t i=0,{βj}t−1 j=0,{σ2 j}t j=0,ρk with k=1, . . . , K We rearrange them according to the original CCL parametrization in order to proceed with the simulations. For each k, we have: µ0j=α0+βjfor i=0 and j=0, . . . , t µij =αi+βj+ρlog(Ci−1,j) + µi−1,jfor i=1, . . . , tand j=0, . . . , t where, being the cumulative payment Ci−1,j not available in lower triangle (i.e., for i+j>t ), we simulate it from a logNormalµi−1,j,σj. Hence, for each parameters set we have a t×t matrix Mk= [µk ij] containing the posterior log-mean parameters and a vector σk= [σk j]containing posterior log-variance parameters. 2. We simulate the sets of losses starting from the elements of M and σ . Following a one-year approach, we only generate, for each k , S different realizations of the next year future cumulative payments C(s,k) i,t−i+1 (with i= 1, . . . , t , k= 1, . . . , K , s= 1, . . . , S ), hence performing a total of K·S simulations. We end up with an array of dimension K·S where each element is a trapezoid T(s,k) composed of the original triangle D(t) and a further simulated diagonal related to payments in the calendar year t+ 1. In Table 1we provide a visualization to clarify this object: each column stands for a different parameter set k and each row is representative of the s -th batch of simulations over the parameters sets, i.e., a Bayesian predictive distribution. 3. Next year incremental payments reported in Formula (9) can be easily obtained by transforming the array of simulated cumulative values in an array of incremental amounts with the following formula: P(k,s) i,t−i+1=C(k,s) i,t−i+1−Ci,t−i(10) for i= 0, . . . , t . In this step, we obtained K·S diagonals of simulated values of the next year payments. 4. Now, we perform the learning update. We use the information generated by the s-th batch of simulations to update the probability of each parameter set. To this extent, we compute the likelyhood of the observations for each element of the array. Being ΦL(x|µ , σ) the density function of a log-normal random variable, for every element we have: LT(s,k)|Θ(k)=∏ C(s,k) ij ∈T(s,k) ΦLC(s,k) ij µ(s,k) ij ,σ(s,k) j(11) Given the sample of K sets, resampling from the sample guarantees that the probability of sampling a specific set is the same for all the sets. In other words, all the sets initially sampled from the MCMC procedure are equally likely. Then, by means of the Bayes theorem, it is possible to obtain the posterior probability of the k-th parameter set given the simulated losses as: PΘ(k)T(s,k)=LT(s,k)|Θ(k)·P(Θ(k)) ∑K h=1LT(s,h)|Θ(h)·P(Θ(h))=LT(s,k)|Θ(k) ∑K h=1LT(s,h)|Θ(h). (12) In other words, for each simulation, we re-evaluate the probability of each set obtaining a posterior probability distribution. If each set is representative of a possible scenario of the reserving process, Risks 2020,8, 125 7 of 20 we have effectively the posterior distribution of all possible scenarios. By iterating the process we obtain Sdifferent posterior distributions for the parameters. 5. By virtue of the log-normal assumption, for each parameters’ set, the expected cumulative payments at times greater than tcan be computed as: EC(k) ij =expµ(k) ij +σ2(k) j 2(13) for i=0, . . . , t,j=0, . . . , tand i+j>t. The Best Estimate RD(t+1) keasily follows. For each batch of simulations s , we use the posterior distribution computed at step 4 in order to obtain a post run-off re-weighted next year Best Estimate. For every s: ˆ RD(t+1) s= K ∑ k=1 RD(t+1) k·PΘ(k)T(s,k)(14) Thus, we are able to obtain a predictive distribution of the expected values of the residual reserve according to Solvency II standards. 6. The predictive distribution of claims development results can be derived by assessing the distribution of the obligations at the end of year t+ 1. For each simulation, we add to each value of the residual reserve distribution the realized diagonal. For each s , we have at disposal K batches of diagonals over all the parameter sets. Thus, we sample a diagonal and we add it to the value calculated at step 6. In this way, we obtain the s -th realization of the next year obligations: Y1yr s=t ∑ i=1 ˜ P(s) i,t−i+1+ˆ RD(t+1) s(15) Again, by iterating this process over all the S batches of simulations, we finally obtain a predictive distribution of future obligations. Table 1. We provide an immediate visualization of what was described in the step 2. T stands for the trapezoid obtained adding to the original triangle the new diagonal containing simulated payments of first calendar year after the evaluation date. The upper parenthesis locates the simulated diagonals in terms of parameter set kand number of simulations s. Simulation Array Θ(1). . . Θ(k). . . Θ(K) 1T(1,1). . . T(1,k). . . T(1,K) . . . . . . . . . sT(s,1). . . T(s,k). . . T(s,K) . . . . . . . . . ST(S,1). . . T(S,k). . . T(S,K) We can then use this distribution to compute the reserve risk capital charge according to the Solvency II directive by applying (8), i.e., by subtracting the Best Estimate to the discounted 99.5% quantile. Previous steps have been described in order to explain the framework of the algorithm we are providing. When coding it, many short-cuts could be taken. For instance, in Table 1the array is composed of trapezoids. It is not strictly necessary to save the original triangle for each element, not only because of computational speed issues but also because this information will be simplified Risks 2020,8, 125 8 of 20 when computing the learning update. In general, the algorithm could be implemented not necessarily following the outline we provided. We provide also a generalization on an n -year time frame, allowing to obtain a predictive distribution for the next-n-years obligations. Mathematical details are reported in Appendix A. 4. An Application of the Model 4.1. Dataset and Model Calibration In this section, we describe the results obtained applying the correlated chain ladder to a 11 ×11 triangle representative of a Motor third party liability portfolio 7 . The triangle has been obtained simulating cumulative payments from a log-normal distribution starting from an observed triangle to preserve the confidentiality of the data. This allows to check if the parameters of the posterior distributions are consistent with the data used. As described later, the posterior distribution has been obtained using a Hamiltonian Monte Carlo (HMC) procedure. To implement the methodology, we maintained the log-normal hypothesis for the observed cumulative payments in the triangle and we elicited the following prior structure, structured according to our knowledge about the business. In particular, the log-ultimate cost is approximated by the parameter αi, defined as follows: αi∼log(Bi) + [log(elri)] + ui where: -log(Bi)is the natural logarithm of the premium earned for the i-th generation. -[log(elri)] is the natural logarithm of the expected loss ratio, for the i-th generation. -uiis a random noise defined as follows: (ui=0 for i=0, . . . , 3 ui∼Uni f orm(−0.6, 0.6)for i=4, . . . , 10 We deemed the variability of the loss ratio sufficient, given that oldest accident years are almost closed. With respect to the elr variable, we choose a log-normal distribution centered on a prior loss ratio differentiated by accident year, mainly based on the analyses made by the company for underwriting purposes. Values are specified in Table 2. Since a limited claims pattern is observed for recent accident years, a higher volatility is assumed. 7Cumulative payments are reported in Table A1. Risks 2020,8, 125 15 of 20 To conclude, we notice that the one-year CCL is not only in line with the more traditional approaches, but can qualify as a model able to produce more robust estimates of the tail of our interest, and a third alternative approach when the other two do not provide a conclusive evidence. Finally, we provided in Figure 4a plot of the one-year distribution generated by this adapted CCL. Figure 4. Distribution of the claims reserve over a one-year time horizon, obtained from 10,000 simulations. The lines represent the 99.5% quantile and the Best Estimate, respectively. By subtracting the Best Estimate from the quantile, we obtain the Solvency II capital requirement that represents the hypothetical SCR for reserve risk returned by this internal model. As reported in Table 6, this value is equal to 52,766.51. Amounts are in thousands of euros. 4.4. Important Remarks The one-year CCL proved to be computationally intensive. In fact, for all the 10,000 sets of parameters, we generated 10,000 batches of simulations of the first diagonal, each composed of other 10 simulations. The algorithm resulted in approximately 10 9 simulations to get 10,000 realizations relative to an 11 × 11 triangle. The method took a while to produce these results and it is advised both to implement it on a fast language as C and to exploit the possibilities of parallel computing to split the work on many different cores, or distributed computing, to split the work on different machines. In this study, we implemented the model in R, exploiting the possibilities offered by the package parallel that allowed us to split the calculations on three different cores of a laptop. The code to reproduce these results is available on the github profile ’GiulioErcole’. 5. Conclusions At end of this study, the one-year CCL performed well and proved to be a legitimate alternative to the other well known models to quantify reserve risk. We sense that this model has something more to offer with respect to Re-Reserving, Merz–Wüthrich formula and the market wide parameter as it captures far more information about the variability of loss liabilities and provide a more truthful representation of the risks connected to the claims reserve. Despite the computational issues, the Bayesian nature of this model allows for embedding prior and external knowledge that could potentially lead to more Risks 2020,8, 125 16 of 20 accurate and tailored representations of the risk profile of the insurer and that are even more in line with the principles behind the possibilities of structuring internal models. Thus we are convinced that more in general a Bayesian approach has still a lot of potential in light of the Solvency II framework. Anyway, it is to be kept in mind that this feature could be a double edged sword as this approach includes some sort of subjectivity that poses the risk that a company could distort purposefully their estimates in order to relieve its capital charges. As a result of this, we want to stress that it is extremely important that priors are elicited truthfully in order to obtain sound and plausible estimates. Author Contributions: All authors contributed equally to this manuscript. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Conflicts of Interest: The authors declare no conflict of interest. Legal Disclaimer: The opinions expressed in the paper are solely those of the author. This paper is for general information, education and discussion purposes only. It does not constitute legal or professional advice and does not necessarily reflect, in whole or in part, any corporate position, opinion or view of PartnerRe or its affiliates. Appendix A. An N-Years Generalization The algorithm shown above could be easily generalized on an n -year time frame, allowing to obtain a predictive distribution for the next-n-years obligations: ˜ Yn−yrs = n ∑ h=1t ∑ i=h Pi,t−i+h+RD(t+n)(A1) From this distribution it is possible to obtain a hypothetical capital charge on such a time frame by applying a risk measure. A longer time period can be useful for instance for own risk assessment purposes as provided by the second pillar of Solvency II. What follows is an extension of the Solvency II guidelines on an n-years time horizon: 1. Starting from the point 2 of the outline of the algorithm structured in the previous section, we simulate a K·S array in which every element is composed by a number n of diagonals, where again K is the number of parameter sets and S is the number of simulations for each parameter set. The s , k -th set of diagonals represent an hypothetical development of the triangle on which we will build a recursive update. In the one-year model, we had for the first diagonal: PΘ(k)T(s,k) 1=LT(s,k) 1|Θ(k)·P(Θ(k)) ∑K h=1LT(s,h) 1|Θ(h)·P(Θ(h)) =LT(s,k) 1|Θ(k) ∑K h=1LT(s,h) 1|Θ(h) where T1 stands for the trapezoid obtained by adding 1 future diagonal to the triangle. We can then use the probabilities obtained in this way to update our knowledge after a second year of run-off: PΘ(k)T(s,k) 2=LT(s,k) 2|Θ(k)·PΘ(k)T(s,k) 1 ∑K h=1LT(s,h) 2|Θ(h)·PΘ(h)T(s,k) 1 Then, recursively obtain the probabilities of each scenario after n years: PΘ(k)T(s,k) n=LT(s,k) n|Θ(k)·PΘ(k)T(s,k) n−1 ∑K h=1LT(s,h) n|Θ(h)·PΘ(h)T(s,k) n−1(A2) Iterating over S, we have S probability distributions that we will use as before to compute S determinations of the residual reserve after n years. 2. At this point, for each parameters’ set we calculate the Best Estimate at time t+n for claims incurred at time t. As for the one-year version, we compute the expected values of the lower triangle: EC(k) ij =expµ(k) ij +σ2(k) j 2(A3) Risks 2020,8, 125 17 of 20 for i+j>t+n . Again, from these values we can obtain the expected values of incremental amounts and the new Best Estimate for a given parameters’ set RD(t+n) k. 3. By iterating step 2 on all parameters sets we obtain K values of the next-n-years Best Estimate. For each batch of simulations s we use the posterior distribution of parameters’ sets computed at step 1 in order to reweight the Best Estimates in light of the realized simulations. For every swe obtain one value: ˆ RD(t+n) s= K ∑ k=1 RD(t+n) k·PΘ(k)T(s,k) n(A4) 4. As before we use the simulated cumulative developments stored in the array to obtain a new array of simulated incremental developments: P(s,k) i,t−i+h=C(s,k) i,t−i+h−C(s,k) i,t−i+h−1, for h=2, . . . , n(A5) P(s,k) i,t−i+1=C(s,k) i,t−i+1−Ci,t−i for i=0, . . . , t. 5. Having at disposal simulated diagonals and the nyears Best Estimates it is possible to obtain the predictive distribution of the next-n-years obligations. For every value of the Best Estimate we add a realized development of payments for the first nyears. For the s -th value of the Best Estimate we have K simulated scenarios of developments. Hence, we sample a scenario and we add it to the s -th realization of the nyears Best Estimate: Yn−yrs s=n ∑ h=1t ∑ i=h P(s,w) i,t−i+h+ˆ RD(t+n) s(A6) where w is a random number between 1 and K. Again, by iterating this process over all the S batches of simulations, we finally obtain a predictive distribution of the next-n-years obligations. Clearly the methodology described in the n -years time frame is a generalization of the one-year approach reported in the previous section. This extension can be useful for risk margin purposes because it is possible to modify a bit Equation (A6) in order to catch the pattern of reserve risk capital requirement until total run-off. Additionally, analyses on a longer time frame are useful in the context of capital management and decision making. For instance, the main purpose of the forward looking assessment of the undertaking’s own risks, is to ensure that the undertaking engages in the process of assessing all the risks inherent to its business and determines the corresponding capital needs. As specified in the Solvency II regulation, the results of the own risk solvency assessment process need to be developed including medium term capital management. Appendix B. Data Here, we provide the triangle and the premium that have been used as a starting point for the analyses contained in this paper. The triangle covers 11 years and is representative of a motor third party liability portfolio. To preserve confidentiality of the data, claims have been generated from a log-normal distribution. Risks 2020,8, 125 18 of 20 Table A1. Triangle of cumulative payments and earned premiums. Amounts in thousands of euros. Accident Development Years Earned Years 0 1 2 3 4 5 6 7 8 9 10 Premium 0 50,145.22 108,869.70 118,157.58 123,434.78 128,075.39 128,620.06 133,727.32 137,249.55 139,652.12 140,224.86 140,668.36 187,498.55 1 66,529.63 120,628.35 135,607.54 138,325.18 141,986.84 143,254.48 148,625.10 151,619.72 153,318.71 154,132.17 209,638.07 2 67,249.55 120,410.05 132,236.67 139,283.38 143,759.42 146,514.73 148,870.33 153,126.08 155,180.52 217,899.50 3 71,335.57 127,456.02 140,645.27 147,157.83 147,993.28 150,819.76 152,306.83 155,879.16 218,391.25 4 76,200.45 146,032.65 160,291.64 168,785.03 171,834.53 172,940.93 176,259.91 234,357.93 5 75,407.41 155,886.72 174,502.23 181,683.61 189,903.49 192,026.62 243,614.50 6 60,923.30 115,047.56 122,880.04 131,293.91 136,295.32 216,966.57 7 60,214.31 120,050.52 132,031.54 137,061.78 197,976.27 8 51,171.99 100,917.33 113,701.60 192,253.76 9 61,167.95 112,561.89 211,541.73 10 70,564.48 247,891.00 Table A2. Parameter µof the log-normal distributions used to simulate values. Accident Development Years Years 0 1 2 3 4 5 6 7 8 9 10 0 10.88 11.60 11.68 11.72 11.76 11.76 11.80 11.83 11.85 11.85 11.85 1 11.07 11.70 11.81 11.84 11.86 11.87 11.91 11.93 11.94 11.95 2 11.13 11.70 11.79 11.84 11.88 11.90 11.91 11.94 11.95 3 11.15 11.76 11.85 11.90 11.90 11.92 11.93 11.96 4 11.25 11.89 11.99 12.04 12.05 12.06 12.08 5 11.23 11.96 12.07 12.11 12.15 12.17 6 11.01 11.65 11.72 11.79 11.82 7 10.98 11.70 11.79 11.83 8 10.84 11.52 11.64 9 11.01 11.63 10 11.12 Risks 2020,8, 125 19 of 20 Table A3. Expected value of the parameter µof the a-posteriori distribution. Accident Development Years Years 0 1 2 3 4 5 6 7 8 9 10 0 10.95 11.60 11.70 11.74 11.77 11.78 11.81 11.83 11.85 11.85 11.85 1 11.04 11.70 11.80 11.84 11.87 11.88 11.90 11.93 11.94 11.94 2 11.05 11.71 11.81 11.85 11.88 11.89 11.91 11.94 11.95 3 11.07 11.73 11.83 11.87 11.90 11.91 11.93 11.96 4 11.23 11.88 11.98 12.02 12.05 12.06 12.09 5 11.32 11.98 12.08 12.12 12.15 12.16 6 10.99 11.64 11.74 11.79 11.81 7 11.03 11.69 11.79 11.83 8 10.88 11.53 11.63 9 10.98 11.64 10 11.17 Table A4. Comparison between loss ratio a priori and posterior value. Years Prior Loss Ratio Posterior Loss Ratio 1 0.73756 0.7359 2 0.7178 0.7163 3 0.731 0.7282 4 0.806 0.8045 5 0.85 0.8487 6 0.685 0.6838 7 0.787 0.7863 8 0.705 0.7053 9 0.72 0.7209 10 0.74 0.7407 References Berger, James O., José M. Bernardo, and Dongchu Sun. 2009. The formal definition of reference priors. The Annals of Statistics 37: 905–38. [CrossRef] Bornhuetter, Ronald L., and Ronald E. Ferguson. 1972. The actuary and IBNR. In Proceedings of the Casualty Actuarial Society. Arlington: Casualty Actuarial Society, pp. 181–95. Dacorogna, Michel, Alessandro Ferriero, and David Krief. 2018. One-Year Change Methodologies for Fixed-Sum Insurance Contracts. Risks 6: 75. [CrossRef] Diers, Dorothea. 2009. Stochastic re-reserving in multi-year internal models—An approach based on simulations. Presented at Astin Colloquium, Helsinki, Finland, June 1–4. Efron, Bradley. 2011. The Bootstrap and Markov Chain Monte Carlo. The Journal of Biopharmaceutical Statistics 21: 1052–62. [CrossRef] [PubMed] England, Peter D., and Richard Verrall. 1999. Analytic and bootstrap estimates of prediction errors in claims reserving. Insurance Mathematics and Economics 25: 281–93. [CrossRef] European Commission. 2009. Directive 2009/138/EC of the European Parliament and of the Council of 25 November 2009 on the Taking-Up and Pursuit of the Business of Insurance and Reinsurance (Solvency II). Luxembourg: European Commission. Frees, Edward W., Richard A. Derrig, and Glenn Meyers, eds. 2016. Predictive Analytics Applications in Actuarial Science. Cambridge: Cambridge University Press. Friedland, Jacqueline. 2010. Estimating Unpaid Claims Using Basic Techniques. Casualty Actuarial Society. Available online: https://www.casact.org/press/index.cfm?fa=viewArticle&articleID=816 (accessed on 19 November 2020). Hastie, Trevor, Robert Tibshirani, and Jerome Friedman. 2009. The Elements of Statistical Learning, Data Mining Inference and Prediction, 2nd ed. Stanford: Stanford University. Hindley, David. 2017. Deterministic Reserving Methods. In Claims Reserving in General Insurance (International Series on Actuarial Science). Cambridge: Cambridge University Press, pp. 40–145. Risks 2020,8, 125 20 of 20 Hoffman, Matthew D., and Andrew Gelman. 2014. The No-U-Turn Sampler: Adaptively Setting Path Lengths in Hamiltonian Monte Carlo. Journal of Machine Learning Research 15: 1351–81. Klugman, Stuart A. 1992. Bayesian Statistics in Actuarial Science. Boston: Kluwer. Mack, Thomas. 1993. Distribution Free calculation of the standard error of Chain Ladder reserve estimates. ASTIN Bulletin The Journal of the IAA 23: 213–25. [CrossRef] Meyers, Glenn. 2017. A Cost of Capital Risk Margin Formula For Non-Life Insurance Liabilities. Variance. Meyers, Glenn. 2015. Stochastic Loss Reserving Using Bayesian MCMC Models. CAS monograph series. Arlington : Casualty Actuarial Society. Ntzoufras, Ioannis, and Petros Dellaportas. 2002. Bayesian Modelling of Outstanding Liabilities Incorporating Claim Count Uncertainty. North American Actuarial Journal 6: 113–28. [CrossRef] Peters, Gareth W., Rodrigo S. Targino, and Mario V. Wüthrich. 2017. Full bayesian analysis of claims reserving uncertainty. Insurance Mathematics and Economics 73: 41–53. [CrossRef] Wüthrich, Mario V., and Michael Merz. 2007. Stochastic Claims Reserving Methods in Insurance. Hoboken: Wiley. Wüthrich, Mario V., Michael Merz, Hans Bühlmann, Massimo De Felice, Alois Gisler, and Franco Moriconi. 2008. Modelling the Claims Development Result for Solvency Purpose. CAS E-Forum. Wüthrich, Mario V. 2007. Using a Bayesian Approach for Claims Reserving. Variance 1: 292–301. Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. c 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).