`Regression Anytime' with Brute-Force SVD Truncation

We propose a new least-squares Monte Carlo algorithm for the approximation of conditional expectations in the presence of stochastic derivative weights. The algorithm can serve as a building block for solving dynamic programming equations, which arise, e.g., in non-linear option pricing problems or in probabilistic discretization schemes for fully non-linear parabolic partial differential equations. Our algorithm can be generically applied when the underlying dynamics stem from an Euler approximation to a stochastic differential equation. A built-in variance reduction ensures that the convergence in the number of samples to the true regression function takes place at an arbitrarily fast polynomial rate, if the problem under consideration is smooth enough.

[1]  Charles R. Johnson,et al.  Matrix analysis , 1985, Statistical Inference for Engineers and Data Scientists.

[2]  Alessandro Balata,et al.  Regress-Later Monte Carlo for optimal control of Markov processes , 2017, 1712.09705.

[3]  Lars Stentoft,et al.  A simulation-and-regression approach for stochastic dynamic programs with endogenous state variables , 2011, Comput. Oper. Res..

[4]  Adam Krzyzak,et al.  A Distribution-Free Theory of Nonparametric Regression , 2002, Springer series in statistics.

[5]  Christian Bender,et al.  Least-Squares Monte Carlo for Backward SDEs , 2012 .

[6]  Paul Glasserman,et al.  Simulation for American Options: Regression Now or Regression Later? , 2004 .

[7]  Warren B. Powell,et al.  Approximate Dynamic Programming - Solving the Curses of Dimensionality , 2007 .

[8]  Joel M. Vanden Exact Superreplication Strategies for a Class of Derivative Assets , 2006 .

[9]  A. Krzyżak,et al.  Optimal global rates of convergence for interpolation problems with random design , 2013 .

[10]  Matthew S. Maxwell,et al.  Approximate Dynamic Programming for Ambulance Redeployment , 2010, INFORMS J. Comput..

[11]  Christian Bender,et al.  Pathwise Dynamic Programming , 2018, Math. Oper. Res..

[12]  Eduardo S. Schwartz Patents and R&D as Real Options , 2003 .

[13]  E. Gobet,et al.  Rate of convergence of an empirical regression method for solving generalized backward stochastic differential equations , 2006 .

[14]  Paul Glasserman,et al.  Monte Carlo Methods in Financial Engineering , 2003 .

[15]  R. Bellman Dynamic programming. , 1957, Science.

[16]  A. Pelsser,et al.  The difference between LSMC and replicating portfolio in insurance liability modeling , 2015, European Actuarial Journal.

[17]  Carlos Vázquez,et al.  Stratified Regression Monte-Carlo Scheme for Semilinear PDEs and BSDEs with Large Scale Parallelization on GPUs , 2016, SIAM J. Sci. Comput..

[18]  T. Inglot,et al.  INEQUALITIES FOR QUANTILES OF THE CHI-SQUARE DISTRIBUTION , 2010 .

[19]  J. Schoenmakers,et al.  Robust Multiple Stopping -- A Pathwise Duality Approach , 2020, 2006.01802.

[20]  S. Menozzi,et al.  On Some non Asymptotic Bounds for the Euler Scheme , 2010, 1001.1347.

[21]  Golub Gene H. Et.Al Matrix Computations, 3rd Edition , 2007 .

[22]  HansenPer Christian The truncated SVD as a method for regularization , 1987 .

[23]  B. Bouchard,et al.  Discrete-time approximation and Monte-Carlo simulation of backward stochastic differential equations , 2004 .

[24]  Adam Krzyzak,et al.  Nonparametric estimation of a function from noiseless observations at random points , 2017, J. Multivar. Anal..

[25]  Nicola Secomandi,et al.  Comparison of Least Squares Monte Carlo Methods with Applications to Energy Real Options , 2017 .

[26]  M. Avellaneda,et al.  Pricing and hedging derivative securities in markets with uncertain volatilities , 1995 .

[27]  Pierre Henry-Labordere,et al.  Uncertain Volatility Model: A Monte-Carlo Approach , 2010 .

[28]  Nizar Touzi,et al.  A Probabilistic Numerical Method for Fully Nonlinear Parabolic PDEs , 2009, 0905.1863.

[29]  Francis A. Longstaff,et al.  Valuing American Options by Simulation: A Simple Least-Squares Approach , 2001 .

[30]  Christian Bender,et al.  Iterative Improvement of Lower and Upper Bounds for Backward SDEs , 2017, SIAM J. Sci. Comput..

[31]  Cornelis W. Oosterlee,et al.  The Stochastic Grid Bundling Method: Efficient Pricing of Bermudan Options and their Greeks , 2013, Appl. Math. Comput..

[32]  John N. Tsitsiklis,et al.  Regression methods for pricing complex American-style options , 2001, IEEE Trans. Neural Networks.

[33]  Marco Avellaneda,et al.  Reducing variance in the numerical solution of BSDEs , 2013 .

[34]  Xiaolu Tan,et al.  Discrete-time probabilistic approximation of path-dependent stochastic control problems , 2014, 1407.0499.

[35]  A. Cohen,et al.  Optimal weighted least-squares methods , 2016, 1608.00512.

[36]  Emmanuel Gobet,et al.  Preliminary control variates to improve empirical regression methods , 2013, Monte Carlo Methods Appl..

[37]  Albert Cohen,et al.  On the Stability and Accuracy of Least Squares Approximations , 2011, Foundations of Computational Mathematics.

[38]  Cornelis W. Oosterlee,et al.  Stochastic grid bundling method for backward stochastic differential equations , 2019, Int. J. Comput. Math..

[39]  Joel A. Tropp,et al.  User-Friendly Tail Bounds for Sums of Random Matrices , 2010, Found. Comput. Math..

[40]  Terry Lyons,et al.  Uncertain volatility and the risk-free synthesis of derivatives , 1995 .

[41]  Jianfeng Zhang A numerical scheme for BSDEs , 2004 .

[42]  Mark Broadie,et al.  A Primal-Dual Simulation Algorithm for Pricing Multi-Dimensional American Options , 2001 .

[43]  Ben Zineb Tarik,et al.  Preliminary control variates to improve empirical regression methods , 2013 .

[44]  Emmanuel Gobet,et al.  Error expansion for the discretization of backward stochastic differential equations , 2006, math/0602503.

[45]  Sean R Eddy,et al.  What is dynamic programming? , 2004, Nature Biotechnology.

[46]  Huyên Pham,et al.  A numerical algorithm for fully nonlinear HJB equations: An approach by control randomization , 2013, Monte Carlo Methods Appl..

[47]  Antoon Pelsser,et al.  Fast Convergence of Regress-Later Estimates in Least Squares Monte Carlo , 2013, 1309.5274.