nep-cmp New Economics Papers
on Computational Economics
Issue of 2017‒05‒21
twelve papers chosen by
Stan Miles
Thompson Rivers University

  1. Deep learning with long short-term memory networks for financial market predictions By Fischer, Thomas; Krauß, Christopher
  2. The biofuel-development nexus: A meta-analysis By Johanna Choumert; Pascale Combes Motel; Charlain Guegang Djimeli
  3. The future of Long Term Care in Europe. An investigation using a dynamic microsimulation model By Vincenzo Atella; Federico Belotti; Ludovico Carrino; Andrea Piano Mortari
  4. Discretizing Nonlinear, Non-Gaussian Markov Processes with Exact Conditional Moments By Farmer, Leland; Toda, Alexis Akira
  5. Trade patterns in the 2060 world economy By Jean Chateau; Lionel Fontagné; Jean Fouré; Åsa Johansson; Eduardo Olaberría
  6. Assessing the Impact of Renewable Energy Sources: Simulation analysis of the Japanese electricity market By YOSHIHARA Keisuke; OHASHI Hiroshi
  7. The impact of macroprudential housing finance tools in Canada By Jason Allen; Timothy Grieder; Tom Roberts; Brian Peterson
  8. Actuarial Inputs and the Valuation of Public Pension Liabilities and Contribution Requirements: A Simulation Approach By Gang Chen; David S. T. Matkin
  9. Implications of Brexit to the Asia-Pacific region: with a focus on least developed countries By Jacob, Arun; Graham, Louis; Moller, Anders K
  10. Bagged artificial neural networks in forecasting inflation: An extensive comparison with current modelling frameworks By Karol Szafranek
  11. Algorithmic trading in a microstructural limit order book model By Frédéric Abergel; Côme Huré; Huyên Pham
  12. Decentralized Pricing and the equivalence between Nash and Walrasian equilibrium By Antoine Mandel; Herbert Gintis

  1. By: Fischer, Thomas; Krauß, Christopher
    Abstract: Long short-term memory (LSTM) networks are a state-of-the-art technique for sequence learning. They are less commonly applied to financial time series predictions, yet inherently suitable for this domain. We deploy LSTM networks for predicting out-of-sample directional movements for the constituent stocks of the S&P 500 from 1992 until 2015. With daily returns of 0.46 percent and a Sharpe Ratio of 5.8 prior to transaction costs, we find LSTM networks to outperform memory-free classification methods, i.e., a random forest (RAF), a deep neural net (DNN), and a logistic regression classifier (LOG). We unveil sources of profitability, thereby shedding light into the black box of artificial neural networks. Specifically, we find one common pattern among the stocks selected for trading - they exhibit high volatility and a short-term reversal return profile. Leveraging these findings, we are able to formalize a rules-based short-term reversal strategy that is able to explain a portion of the returns of the LSTM.
    Keywords: finance,statistical arbitrage,LSTM,machine learning,deep learning
    Date: 2017
  2. By: Johanna Choumert (EDI - Economic Development Initiatives - Economic Development Initiatives); Pascale Combes Motel (CERDI - Centre d'Études et de Recherches sur le Développement International - UdA - Université d'Auvergne - Clermont-Ferrand I - CNRS - Centre National de la Recherche Scientifique); Charlain Guegang Djimeli (, DPPP/DGEPIP/MINEPAT - Direction générale de l'économie et de la programmation des investissements publics - Ministry of the Economy)
    Abstract: While the production of biofuels has expanded in recent years, findings in the literature on its impact on growth and development remain contradictory. This paper presents a meta-analysis of computable general equilibrium studies published between 2006 and 2014. Using 26 studies, we shed light on why their results differ. We investigate factors such as biofuel type, geographic area and the characteristics of models employed. Our results indicate that the outcomes of CGE simulations are sensitive to model parameters and also suggest heterogenous effects of biofuel expansion between developed / emerging countries and Sub-Saharan African countries. Our quantitative meta-analysis complements existing narrative surveys and confirms that results are sensitive to key hypotheses on essential parameters. Simulations on longer time periods and in multi-country studies lead to results that indicate higher impacts of biofuel expansion on growth and household income. Moreover, simulations with a shock in agricultural productivity indicate positive welfare gains, unlike simulations with a shock on land expansion. Lastly, we find that biodiesels lead to higher welfare gains than biofuels.
    Keywords: Development,Biofuel,Bioethanol,Biodiesel,Energy,Meta-regression,Computable General Equilibrium Model.
    Date: 2017–04–24
  3. By: Vincenzo Atella (DEF and CEIS,University of Rome "Tor Vergata"); Federico Belotti (DEF and CEIS,University of Rome "Tor Vergata"); Ludovico Carrino (King's College London); Andrea Piano Mortari (CEIS, University of Rome "Tor Vergata")
    Abstract: In this paper we investigate the evolution of public European LTC systems in the forthcoming decades, using the Europe Future Elderly Model (EuFEM), a dynamic microsimulation model which projects the health and socio-economic characteristics of the 50+ population of ten European countries, augmented with the explicit modelling of the eligibility rules of 5 countries. The use of SHARE data allows to have a better understanding of the trends in the demand for LTC differentiated by age groups, gender, and other demographic and social characteristics in order to better assess the distributional effects. We estimate the future potential coverage (or gap of coverage) of each national/regional public home-care system, and then disentangle the differences between countries in a population and a regulation effects. Our analysis offers new insights on how would the demand for LTC evolve over time, what would the distributional effects of different LTC policies be if no action is taken, and what could be potential impact of alternative care policies.
    Keywords: Dynamic Microsimulation, Long Term Care, Forecast, disability, Regulation
    JEL: I11 I18 C53 C63 J14 J11
    Date: 2017–05–08
  4. By: Farmer, Leland; Toda, Alexis Akira
    Abstract: Approximating stochastic processes by finite-state Markov chains is useful for reducing computational complexity when solving dynamic economic models. We provide a new method for accurately discretizing general Markov processes by matching low order moments of the conditional distributions using maximum entropy. In contrast to existing methods, our approach is not limited to linear Gaussian autoregressive processes. We apply our method to numerically solve asset pricing models with various underlying stochastic processes for the fundamentals, including a rare disasters model. Our method outperforms the solution accuracy of existing methods by orders of magnitude, while drastically simplifying the solution algorithm. The performance of our method is robust to parameters such as the number of grid points and the persistence of the process.
    Keywords: asset pricing models, duality, Kullback-Leibler information, numerical methods, solution accuracy
    JEL: C63 C68 G12
    Date: 2016–11–01
  5. By: Jean Chateau (OECD - OECD); Lionel Fontagné (PSE - Paris School of Economics, CES - Centre d'économie de la Sorbonne - UP1 - Université Panthéon-Sorbonne - CNRS - Centre National de la Recherche Scientifique, CEPII - Centre d'Etudes Prospectives et d'Informations Internationales - Centre d'analyse stratégique); Jean Fouré (CEPII - Centre d'Etudes Prospectives et d'Informations Internationales - Centre d'analyse stratégique); Åsa Johansson (OECD - OECD); Eduardo Olaberría (OECD - OECD)
    Abstract: This paper presents long-term trade scenarios for the world economy up to 2060 based on a modelling approach that combines aggregate growth projections for the world with a detailed computable general equilibrium sectoral trade model. The analysis suggests that over the next 50 years, the geographical centre of trade will continue to shift from OECD to non-OECD regions reflecting faster growth in non-OECD countries. The relative importance of different regions in specific export markets is set to change markedly over the next half century with emerging economies gaining export shares in manufacturing and services. Trade liberalisation, including gradual removal of tariffs, regulatory barriers in services and agricultural support, as well as a reduction in transaction costs on goods, could increase global trade and GDP over the next 50 years. Specific scenarios of regional liberalisation among a core group of OECD countries or partial multilateral liberalisation could, respectively, raise trade by 4% and 15% and GDP by 0.6% and 2.8% by 2060 relative to the status quo. Finally, the model highlights that investment in education has an influence on trade and high-skill specialisation patterns over the coming decades. Slower educational upgrading in key emerging economies than expected in the baseline scenario could reduce world exports by 2% by 2060. Lower up-skilling in emerging economies would also slow down the restructuring towards higher value-added activities in these emerging economies.
    Keywords: General equilibrium trade model,long-term trade and specialisation patterns,trade liberalisation
    Date: 2015–11
  6. By: YOSHIHARA Keisuke; OHASHI Hiroshi
    Abstract: This paper evaluates the impact of renewable energy (RE) sources on market outcomes in Japan. We develop a simulation model to compute the kWh-market equilibrium, and conduct simulation exercises for 2015 and 2030. Using scenarios proposed by the government, we find that the diffusion of RE sources would lower the kWh-market prices and greenhouse gases by reducing fossil fuel consumption in 2030. It would also mothball many of the thermal power plants, which were active and profitable in 2015.
    Date: 2017–04
  7. By: Jason Allen; Timothy Grieder; Tom Roberts; Brian Peterson
    Abstract: This paper combines loan-level administrative data with household-level survey data to analyze the impact of recent macroprudential policy changes in Canada using a microsimulation model of mortgage demand of first-time homebuyers. Policies targeting the loan-to-value ratio are found to have a larger impact on demand than policies targeting the debt-service ratio, such as amortization. In addition, we show that loan-to-value policies have a larger role to play in reducing default than income-based policies.
    Keywords: macroprudential policy, household finnance, microsimulation models
    Date: 2017–05
  8. By: Gang Chen; David S. T. Matkin
    Abstract: This paper uses a simulated public pension system to examine the sensitivity of actuarial input changes on funding ratios and contribution requirements. We examine instantaneous and lagged effects, marginal and interactive effects, and effects under different funding conditions and demographic profiles. The findings emphasize the difficulty of conducting cross-sectional analyses of public pension systems and point to several important considerations for future research.
    Date: 2017–05
  9. By: Jacob, Arun; Graham, Louis; Moller, Anders K
    Abstract: Brexit might affect exports of some countries in the Asia-Pacific region disproportionately more than others. Simulation results, under different Brexit scenarios, show that the potential reduction in trade faced by least developed countries (LDCs) of the region can range from 16% to 50% of their current export value to the UK in key sectors such as fish, clothes, textiles and footwear. Simulations also show that it is the larger developing countries from the region that would benefit from any trade diversion that ensues in these sectors. Countries with higher exposure to Brexit induced risks need to engage in deeper analyses of the extent of such impacts and brace themselves for proactive discussions with the UK in order to limit negative impacts.
    Keywords: Brexit, LDCs, trade diversion, trade policy
    JEL: F13 F17 O24 P45
    Date: 2017–03–03
  10. By: Karol Szafranek (Narodowy Bank Polski, Warsaw School of Economics)
    Abstract: Accurate inflation forecasts lie at the heart of effective monetary policy. By utilizing a thick modelling approach, this paper investigates the out-of-sample quality of the short-term Polish headline inflation forecasts generated by a combination of thousands of bagged single hidden-layer feed-forward artificial neural networks in the period of systematically falling and persistently low inflation. Results indicate that the forecasts from this model outperform a battery of popular approaches, especially at longer horizons. During the excessive disinflation it has more accurately accounted for the slowly evolving local mean of inflation and remained only mildly biased. Moreover, combining several linear and nonlinear approaches with diverse underlying model assumptions delivers further statistically significant gains in the predictive accuracy and statistically outperforms a panel of examined benchmarks at multiple horizons. The robustness analysis shows that resigning from data preprocessing and bootstrap aggregating severely compromises the forecasting ability of the model.
    Keywords: inflation forecasting, artificial neural networks, principal components, bootstrap aggregating, forecast combination
    JEL: C22 C38 C45 C53 C55
    Date: 2017
  11. By: Frédéric Abergel (MICS - Mathématiques et Informatique pour la Complexité et les Systèmes - CentraleSupélec); Côme Huré (LPMA - Laboratoire de Probabilités et Modèles Aléatoires - UPMC - Université Pierre et Marie Curie - Paris 6 - UPD7 - Université Paris Diderot - Paris 7 - CNRS - Centre National de la Recherche Scientifique); Huyên Pham (CREST - Centre de Recherche en Économie et Statistique - INSEE - ENSAE ParisTech - École Nationale de la Statistique et de l'Administration Économique, LPMA - Laboratoire de Probabilités et Modèles Aléatoires - UPMC - Université Pierre et Marie Curie - Paris 6 - UPD7 - Université Paris Diderot - Paris 7 - CNRS - Centre National de la Recherche Scientifique)
    Abstract: We propose a microstructural modeling framework for studying optimal market making policies in a FIFO (first in first out) limit order book (LOB). In this context, the limit orders, market orders, and cancel orders arrivals in the LOB are modeled as Cox point processes with intensities that only depend on the state of the LOB. These are high-dimensional models which are realistic from a micro-structure point of view and have been recently developed in the literature. In this context, we consider a market maker who stands ready to buy and sell stock on a regular and continuous basis at a publicly quoted price, and identifies the strategies that maximize her P&L penalized by her inventory. We apply the theory of Markov Decision Processes and dynamic programming method to characterize analytically the solutions to our optimal market making problem. The second part of the paper deals with the numerical aspect of the high-dimensional trading problem. We use a control randomization method combined with quantization method to compute the optimal strategies. Several computational tests are performed on simulated data to illustrate the efficiency of the computed optimal strategy. In particular, we simulated an order book with constant/ symmet-ric/ asymmetrical/ state dependent intensities, and compared the computed optimal strategy with naive strategies.
    Keywords: Markovian Quantization,Markov Decision Process,Limit Order Book,High-Frequency Trading,Queuing model,pure-jump controlled process,High-dimensional Stochastic Control
    Date: 2017–05–03
  12. By: Antoine Mandel (PSE - Paris School of Economics, CES - Centre d'économie de la Sorbonne - UP1 - Université Panthéon-Sorbonne - CNRS - Centre National de la Recherche Scientifique); Herbert Gintis (Santa Fe Institute)
    Abstract: We introduce, in the standard exchange economy model, market games in which agents use private prices as strategies. We give conditions on the game form that ensure that the only strict Nash equilibria of the game are the competitive equilibria of the underlying economy. This equivalence result has two main corollaries. First, it adds to the evidence that competitive equilibria can be strategically stable even in small economies. Second, it implies that competitive equilibria have good local stability properties under a large class of evolutionary learning dynamics.
    Keywords: General equilibrium, Market games, Stability, Computational economics, Evolutionary game theory
    Date: 2016–03–01

This nep-cmp issue is ©2017 by Stan Miles. It is provided as is without any express or implied warranty. It may be freely redistributed in whole or in part for any purpose. If distributed in part, please include this notice.
General information on the NEP project can be found at For comments please write to the director of NEP, Marco Novarese at <>. Put “NEP” in the subject, otherwise your mail may be rejected.
NEP’s infrastructure is sponsored by the School of Economics and Finance of Massey University in New Zealand.