Wright Meets Markowitz: How Standard Portfolio Theory Changes When Assets Are Technologies Following Experience Curves

We consider how to optimally allocate investments in a portfolio of competing technologies using the standard mean-variance framework of portfolio theory. We assume that technologies follow the empirically observed relationship known as Wright's law, also called a "learning curve" or "experience curve", which postulates that costs drop as cumulative production increases. This introduces a positive feedback between cost and investment that complicates the portfolio problem, leading to multiple local optima, and causing a trade-off between concentrating investments in one project to spur rapid progress vs. diversifying over many projects to hedge against failure. We study the two-technology case and characterize the optimal diversification in terms of progress rates, variability, initial costs, initial experience, risk aversion, discount rate and total demand. The efficient frontier framework is used to visualize technology portfolios and show how feedback results in nonlinear distortions of the feasible set. For the two-period case, in which learning and uncertainty interact with discounting, we compare different scenarios and find that the discount rate plays a critical role.

[1]  R. Pindyck,et al.  The Learning Curve and Optimal Production Under Uncertainty , 1987 .

[2]  Joseph B. Mazzola,et al.  A Bayesian Approach to Managing Learning-Curve Uncertainty , 1996 .

[3]  J. Doyne Farmer,et al.  How well do experience curves predict technological progress? A method for making distributional forecasts , 2017, 1703.05979.

[4]  P. Zeppini.  A discrete choice model of transitions to sustainable technologies , 2015 .

[5]  K. Arrow.  The Economic Implications of Learning by Doing , 1962 .

[6]  R. Litzenberger,et al.  SKEWNESS PREFERENCE AND THE VALUATION OF RISK ASSETS , 1976 .

[7]  Peter Thompson,et al.  The Relationship between Unit Cost and Cumulative Quantity and the Evidence for Organizational Learning-by-Doing , 2012 .

[8]  Jan K. Brueckner,et al.  Optimal production with learning by doing , 1983 .

[9]  H. Waelbroeck,et al.  Optimal Execution of Portfolio Transactions with Short‐Term Alpha , 2013 .

[10]  Hua He,et al.  Dynamic Trading Policies with Price Impact , 2001 .

[11]  D. North.  Competing Technologies , Increasing Returns , and Lock-In by Historical Events , 1994 .

[12]  Socrates Kypreos,et al.  Endogenizing R&D and Market Experience in the "Bottom-Up" Energy-Systems ERIS Model , 2004 .

[13]  K. Fisher-Vanden,et al.  Should We Give Up after Solyndra? Optimal Technology R&D Portfolios under Uncertainty , 2015, Journal of the Association of Environmental and Resource Economists.

[14]  Luís M. B. Cabral,et al.  The Learning Curve, Market Dominance and Predatory Pricing , 1994 .

[15]  S. Rosen.  Learning by Experience as Joint Production , 1972 .

[16]  J. Trancik,et al.  Statistical Basis for Predicting Technological Progress , 2012, PloS one.

[17]  R. Cowan.  Tortoises and Hares: Choice among Technologies of Unknown Merit , 1991 .

[18]  Optimal Investment Strategy for Competing Learning Technologies: An Analytical Approach , 2014 .

[19]  K. Riahi,et al.  Risk Hedging Strategies under Energy System and Climate Policy Uncertainties , 2013 .

[20]  J. Bergh.  Optimal Diversity: Increasing Returns versus Recombinant Innovation , 2008 .

[21]  Joseph B. Mazzola,et al.  The Stochastic Learning Curve: Optimal Production in the Presence of Learning-Curve Uncertainty , 1997, Oper. Res..

[22]  J. Trancik,et al.  Historical Costs of Coal-Fired Electricity and Implications for the Future , 2009, 1001.0605.

[23]  J. C. van den Bergh,et al.  Optimal Diversity in Investments with Recombinant Innovation , 2008 .

[24]  M. Deakin.  Catastrophe theory. , 1977, Science.

[25]  M. Spence.  The Learning Curve and Competition , 1981 .

[26]  Tan Wang,et al.  Keynes Meets Markowitz: The Tradeoff between Familiarity and Diversification , 2009, Manag. Sci..

[27]  A. Alchian.  Reliability of Progress Curves in Airframe Production , 1963 .

[28]  P. Dasgupta,et al.  Learning-by-doing, Market Structure and Industrial and Trade Policies , 1988 .

[29]  P. Newbold,et al.  How well are long-run commodity price series characterized by trend components? , 2005 .

[30]  A. Atkinson,et al.  A New View of Technological Change , 1969 .

[31]  H. Markowitz,et al.  Mean-Variance versus Direct Utility Maximization , 1984 .

[32]  Chris Hope,et al.  Climate modelling with endogenous technical change: Stochastic learning and optimal greenhouse gas abatement in the PAGE2002 model , 2007 .

[33]  L. Pulley.  A General Mean-Variance Approximation to Expected Utility for Short Holding Periods , 1981, Journal of Financial and Quantitative Analysis.

[34]  P. Kort,et al.  Optimal Investment in Learning-Curve Technologies , 2011 .

[35]  Laura L. Veldkamp,et al.  Information Acquisition and Under-Diversification , 2008 .

[36]  D. Gregori,et al.  Inside the Black Box: Technology and Economics , 1984 .

[37]  P. David.  Clio and the Economics of QWERTY , 1985 .

[38]  Andrii Gritsevskyi,et al.  Modeling uncertainty of induced technological change , 2000 .

[39]  N. C. P. Edirisinghe,et al.  Multi-period stochastic portfolio optimization: Block-separable decomposition , 2007, Ann. Oper. Res..

[40]  Garleanu Nicolae,et al.  Dynamic Trading with Predictable Returns and Transaction Costs , 2009 .

[41]  Alban Kitous,et al.  Mitigation strategies and energy technology learning: an assessment with the POLES model , 2015 .

[42]  J. Dutton,et al.  Treating Progress Functions as a Managerial Opportunity , 1984 .

[43]  T. P. Wright,et al.  Factors affecting the cost of airplanes , 1936 .