Abstract
Monte-Carlo based valuation in life insurance involves the simulation of various components of the balance sheet: portfolios, guarantees, assets mix, market conditions and many other specific risk factors; which can be time-consuming. In view of the time needed to achieve this task, insurers are then facing a trade-off of balancing the number of simulations against the uncertainty surrounding the estimated quantity. In the current paper, we propose a variance-reduction methodology using a machine learning technique. It roots from the unsupervised learning literature in conjunction with the quantization of random processes. The goal is to reduce the number of simulated Brownian paths, using auxiliary scenarios that can be seen as path clusters, which efficiently and accurately approximate the initial ones with regards to an adequate measure of distance. Moreover, we introduce penalty to accommodate for various inputs of the initial conditions of risk factors. By doing so, we implicitly assign labels to the scenarios and thus advocate using a semi-supervised learning to enhance the performance of the scenarios reduction by imposing an additional impurity condition on the clusters based on these labels. This is made possible thanks to a decomposition property of the insurers cash-flows, which allows to disentangle the initial conditions of risk factors from the Brownian motions driving their dynamics. The training of the proposed learning algorithm is based on an adaptation of the well-known k-means algorithm. An intensive numerical study is carried out over a range of simulation setups to compare the performances of the proposed methodology. We show, numerically, that the proposed methodology outperforms some classical variance reduction approach. Also, using a real-life dataset, we show that our methodology outperforms some conventional variance reduction used by life insurance practitioners.
Similar content being viewed by others
Notes
Undertakings for the Collective Investment of Transferable Securities
References
Aloise, D., Deshpande, A., Hansen, P., & Popat, P. (2009). Np-hardness of Euclidean sum-of-squares clustering. Machine Learning, 75(2), 245–248.
Bally, V., & Pages, G. (2003). Error analysis of the optimal quantization algorithm for obstacle problems. Stochastic Processes and Their Applications, 106(1), 1–40.
Bauer, D., Bergmann, D., & Kiesel, R. (2010). On the risk-neutral valuation of life insurance contracts with numerical methods in view. ASTIN Bulletin: The Journal of the IAA, 40(1), 65–95.
Bauer, D., Bergmann, D., & Reuss, A. (2010b). Solvency II and nested simulations: A least-squares Monte Carlo approach. In Proceedings of the 2010 ICA congress.
Bignon, A., Ndjeng-Ndjeng, A., Salhi, Y., & Thérond, P.-E. (2019). A reduced-form model for a life insurance’s net asset value. Bankers, Markets & Investors, 157, 1–13.
Borel-Mathurin, F. & Vedani, J. (2019). Market-consistent valuation: A step towards calculation stability.
Cathcart, M. & Morrison, S. (2009). Variable annuity economic capital: The least-squares Monte Carlo approach. Life & Pensions, pp. 36–40.
Chauvigny, M., Devineau, L., Loisel, S., & Maume-Deschamps, V. (2011). Fast remote but not extreme quantiles with multiple factors: Applications to solvency ii and enterprise risk management. European Actuarial Journal, 1(1), 131–157.
Conze, A. (1991). Path dependent options: The case of lookback options. The Journal of Finance, 46(5), 1893–1907.
Davidson, I. & Ravi, S. (2005). Clustering with constraints: Feasibility issues and the k-means algorithm. In Proceedings of the 2005 SIAM international conference on data mining, pp. 138–149. SIAM.
Demiriz, A., Bennett, K. P., & Embrechts, M. J. (1999). Semi-supervised clustering using genetic algorithms. Artificial Neural Networks in Engineering, pp. 809–814.
Durrett, R. & Durrett, R. (1984). Brownian motion and martingales in analysis. Wadsworth Advanced Books & Software California.
Felix, J.-P. (2016). Cash flow projection models. In Modelling in life insurance: A management perspective, pp. 63–79. Springer.
Floryszczak, A., Le Courtois, O., & Majri, M. (2016). Inside the Solvency 2 black box: Net asset values and solvency capital requirements with a least-squares monte-carlo approach. Insurance: Mathematics and Economics, 71, 15–26.
Glynn, P. W. (1994). Efficiency improvement techniques. Annals of Operations Research, 53(1), 175–197.
Gray, R. (1984). Vector quantization. IEEE Assp Magazine, 1(2), 4–29.
Hainaut, D., Devolder, P., & Pelsser, A. (2018). Robust evaluation of SCR for participating life insurances under Solvency II. Insurance: Mathematics and Economics, 79, 107–123.
Hilli, P., Koivu, M., Pennanen, T., & Ranne, A. (2007). A stochastic programming model for asset liability management of a finnish pension company. Annals of Operations Research, 152(1), 115–139.
Liu, Y., Pagès, G., et al. (2020). Characterization of probability distribution convergence in wasserstein distance by \( l^p\)-quantization error function. Bernoulli, 26(2), 1171–1204.
MacQueen, J. et al. (1967). Some methods for classification and analysis of multivariate observations. In Proceedings of the fifth Berkeley symposium on mathematical statistics and probability, vol. 1, pp. 281–297. Oakland, CA, USA.
Pollard, D. (1982). Quantization and the method of k-means. IEEE Transactions on Information Theory, 28(2), 199–205.
Sculley, D. (2010). Web-scale k-means clustering. In Proceedings of the 19th international conference on World wide web, pp. 1177–1178.
Thérond, P.-E. (2016). About market consistent valuation in insurance. In Modelling in life insurance: A management perspective, pp. 43–60. Springer.
Vedani, J., El Karoui, N., Loisel, S., & Prigent, J.-L. (2017). Market inconsistencies of market-consistent European life insurance economic valuations: Pitfalls and practical solutions. European Actuarial Journal, 7(1), 1–28.
Villani, C. (2008). Optimal transport: Old and new, vol. 338. Springer
Zenios, S. A. (1995). Asset/liability management under uncertainty for fixed-income securities. Annals of Operations Research, 59(1), 77–97.
Zhang, T., Ramakrishnan, R., & Livny, M. (1996). Birch: An efficient data clustering method for very large databases. ACM SIGMOD Record, 25(2), 103–114.
Acknowledgements
The authors would like to thank the anonymous referees for their constructive comments that improved the quality of this paper.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
A Appendices
A Appendices
1.1 A.1 Analytical variance of the call option
We will derive the theoretic variance of the cash flow \(f(Z,x)=e^{-rT}(x\exp ((r-\sigma ^2/2)T+\sigma \sqrt{T}Z)-C)^+\), where Z is a random Gaussian variable. First, we consider
where \(d_2=\left( \log (C/x)+(r-\sigma ^2/2)T\right) /\sigma \sqrt{T}\). Hence, we have the classic Black & Scholes analytic formula
Hence, letting \(u=y+\sigma \sqrt{T}\), we come up the well-celebrated Black & Scholes formula
where \(d_1=d_2+\sigma \sqrt{T}\), and \(\phi \) is the standard normal cumulative distribution function. In order to come with the variance of cash flows, we need to compute the following quantity
Using the same arguments as the above formula up to a change of variable, we can write
where \(d_3=d_2+2\sigma \sqrt{T}\). Hence, the variance of the cash-flows can be given explicitly, which will be used to quantify the variance underlying the Monte-Carlo simulations as follows
1.2 A.2 Variance reduction with antithetic variables technique
The antithetic variable techinque estimates the expected value of f(Z, x) by
The variance of this estimator is given by:
where
and
with \(d_4=-d_2+\sigma \sqrt{T}\).
1.3 A.3 Analytical variance of the fixed strike lookback call option
Consider a fixed strike Lookback Call Option that starts at \(t=0\). Define \(X = \max (C,x)\) then
Rights and permissions
About this article
Cite this article
Jimenez, M., Salhi, Y. A semi-supervised learning approach for variance reduction in life insurance. Ann Oper Res 334, 157–184 (2024). https://doi.org/10.1007/s10479-022-04845-7
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10479-022-04845-7