Accurate computation of the right tail of the sum of dependent log-normal variates

We study the problem of the Monte Carlo estimation of the right tail of the distribution of the sum of correlated log-normal random variables. While a number of theoretically efficient estimators have been proposed for this setting, using a few numerical examples we illustrate that these published proposals may not always be useful in practical simulations. In other words, we show that the established theoretical efficiency of these estimators does not necessarily convert into Monte Carlo estimators with low variance. As a remedy to this defect, we propose a new estimator for this setting. We demonstrate that, not only is our novel estimator theoretically efficient, but, more importantly, its practical performance is significantly better than that of its competitors.

[1]  Jens Ledet Jensen,et al.  Exponential Family Techniques for the Lognormal Left Tail , 2014, 1403.4689.

[2]  W. Stahel,et al.  Log-normal Distributions across the Sciences: Keys and Clues , 2001 .

[3]  Quang Huy Nguyen,et al.  New efficient estimators in rare event simulation with heavy tails , 2014, J. Comput. Appl. Math..

[4]  Søren Asmussen,et al.  On the Laplace Transform of the Lognormal Distribution , 2014, Methodology and Computing in Applied Probability.

[5]  Søren Asmussen,et al.  Asymptotics of sums of lognormal random variables with Gaussian copula , 2008 .

[6]  Sandeep Juneja,et al.  Efficient simulation of tail probabilities of sums of correlated lognormals , 2011, Ann. Oper. Res..

[7]  Søren Asmussen,et al.  Conditional Monte Carlo for sums, with applications to insurance and finance , 2018, Annals of Actuarial Science.

[8]  Zdravko I. Botev,et al.  Fast and accurate computation of the distribution of sums of dependent log-normals , 2017, Ann. Oper. Res..

[9]  D. Kortschak,et al.  Second Order Asymptotics of Aggregated Log-Elliptical Risk , 2014 .

[10]  S. Posner,et al.  Asian Options, The Sum Of Lognormals, And The Reciprocal Gamma Distribution , 1998 .

[11]  P. Holgate.  Lognormal Distributions: Theory and Applications , 1989 .

[12]  Michel Mandjes,et al.  Tail distribution of the maximum of correlated Gaussian random variables , 2015, 2015 Winter Simulation Conference (WSC).

[13]  D. Dufresne.  The log-normal approximation in financial and other computations , 2004, Advances in Applied Probability.

[14]  Robert E. Kass,et al.  Second‐Order Asymptotics , 2011 .

[15]  Mohamed-Slim Alouini,et al.  Unified Importance Sampling Schemes for Efficient Simulation of Outage Capacity Over Generalized Fading Channels , 2016, IEEE Journal of Selected Topics in Signal Processing.

[16]  Emmanuel Bacry,et al.  Log-normal continuous cascade model of asset returns: aggregation properties and estimation , 2013 .

[17]  E. Crow,et al.  Lognormal Distributions: Theory and Applications , 1987 .

[18]  Dominik Kortschak,et al.  Efficient simulation of tail probabilities for sums of log-elliptical risks , 2013, J. Comput. Appl. Math..

[19]  Søren Asmussen,et al.  Error Rates and Improved Algorithms for Rare Event Simulation with Heavy Weibull Tails , 2015 .

[20]  Piet Van Mieghem,et al.  Lognormal Infection Times of Online Information Spread , 2013, PloS one.

[21]  Dirk P. Kroese,et al.  Handbook of Monte Carlo Methods , 2011 .

[22]  B. Tuffin.  Bounded normal approximation in simulations of highly reliable Markovian systems , 1999 .

[23]  Dirk P. Kroese,et al.  Improved algorithms for rare event simulation with heavy tails , 2006, Advances in Applied Probability.

[24]  Sergio Ortobelli Lozza,et al.  Asymptotic stochastic dominance rules for sums of i.i.d. random variables , 2016, J. Comput. Appl. Math..

[25]  Jens Ledet Jensen,et al.  Approximating the Laplace transform of the sum of dependent lognormals , 2015, Advances in Applied Probability.