|
on Computational Economics |
|
Issue of 2018‒01‒15
eleven papers chosen by |
| By: | O. B. Sezer; M. Ozbayoglu; E. Dogdu |
| Abstract: | In this paper, a neural network-based stock price prediction and trading system using technical analysis indicators is presented. The model developed first converts the financial time series data into a series of buy-sell-hold trigger signals using the most commonly preferred technical analysis indicators. Then, a Multilayer Perceptron (MLP) artificial neural network (ANN) model is trained in the learning stage on the daily stock prices between 1997 and 2007 for all of the Dow30 stocks. Apache Spark big data framework is used in the training stage. The trained model is then tested with data from 2007 to 2017. The results indicate that by choosing the most appropriate technical indicators, the neural network model can achieve comparable results against the Buy and Hold strategy in most of the cases. Furthermore, fine tuning the technical indicators and/or optimization strategy can enhance the overall trading performance. |
| Date: | 2017–12 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:1712.09592&r=cmp |
| By: | Torsten Trimborn; Philipp Otte; Simon Cramer; Max Beikirch; Emma Pabich; Martin Frank |
| Abstract: | We introduce the simulation tool SABCEMM (Simulate Agent-Based Computational Economic Market Models) for agent-based computational economic market (ABCEM) models. Our simulation tool is implemented in C++ and we can easily run ABCEM models with up to several million agents. Thanks to the object-oriented software design, this tool enables the user to design and compare multiple ABCEM models from a unified perspective. Thus, one can easily change the market mechanism or agent types. This makes it possible to quantitatively compare ABCEM models e.g. regarding the ability of each model to reproduce stylized facts. We present a qualitative study of three known ABCEM models and several variants of those. Furthermore, we discuss finite-size effects and time discretizations of ABCEM models. Finally, we show the great impact of different random number generators on the run time of ABCEM models and even on the qualitative output of the model. The code can be downloaded from GitHub https://github.com/SABCEMM/SABCEMM, such that all results can be reproduced by the reader |
| Date: | 2018–01 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:1801.01811&r=cmp |
| By: | Masaya Abe; Hideki Nakayama |
| Abstract: | Many studies have been undertaken by using machine learning techniques, including neural networks, to predict stock returns. Recently, a method known as deep learning, which achieves high performance mainly in image recognition and speech recognition, has attracted attention in the machine learning field. This paper implements deep learning to predict one-month-ahead stock returns in the cross-section in the Japanese stock market and investigates the performance of the method. Our results show that deep neural networks generally outperform shallow neural networks, and the best networks also outperform representative machine learning models. These results indicate that deep learning shows promise as a skillful machine learning method to predict stock returns in the cross-section. |
| Date: | 2018–01 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:1801.01777&r=cmp |
| By: | Timo Boppart; Per Krusell; Kurt Mitman |
| Abstract: | We propose a new method for computing equilibria in heterogeneous-agent models with aggregate uncertainty. The idea relies on an assumption that linearization offers a good approximation; we share this assumption with existing linearization methods. However, unlike those methods, the approach here does not rely on direct derivation of first-order Taylor terms. It also does not use recursive methods, whereby aggregates and prices would be expressed as linear functions of the state, usually a very high-dimensional object (such as the wealth distribution). Rather, we rely merely on solving nonlinearly for a deterministic transition path: we study the equilibrium response to a single, small "MIT shock'' carefully. We then regard this impulse response path as a numerical derivative in sequence space and hence provide our linearized solution directly using this path. The method can easily be extended to the case of many shocks and computation time rises linearly in the number of shocks. We also propose a set of checks on whether linearization is a good approximation. We assert that our method is the simplest and most transparent linearization technique among currently known methods. The key numerical tool required to implement it is value-function iteration, using a very limited set of state variables. |
| JEL: | C68 E1 |
| Date: | 2017–12 |
| URL: | http://d.repec.org/n?u=RePEc:nbr:nberwo:24138&r=cmp |
| By: | Alberto Cardaci (Universita Cattolica des Sacro Cuore, Via Lodovico Necchi, Milan, Italie); Francesco Saraceno (OFCE, Sciences Po Paris, France, LUISS-SEP, Rome, Italie) |
| Abstract: | Our paper investigates the impact of rising inequality in a two-country macroeconomic model with an agent-based household sector characterised by peer effects in consumption. In particular, the model highlights the role of inequality in determining diverging balance of payments dynamics within a currency union. Inequality may drive the two countries into different growth patterns: where peer effects in consumption interact with higher credit availability, rising income inequality leads to the emergence of a debt-led growth. Where social norms determine weaker emulation and credit availability is lower, an export-led regime arises. Eventually, a crisis emerges endogenously due to the sudden-stop of capital ows from the net lending country, triggered by the excessive risk associated to the dramatic amount of private debt accumulated by households in the borrowing country. Monte Carlo simulations for a wide range of calibrations confirm the robustness of our results. |
| Keywords: | Inequality, Current Account, Currency Union, Agent-based model |
| JEL: | C63 D31 E21 F32 F43 |
| Date: | 2017–12–12 |
| URL: | http://d.repec.org/n?u=RePEc:fce:doctra:1730&r=cmp |
| By: | Giovanni Dosi (Scuola Superiore Sant'Anna Pisa Italy); Mauro Napoletano (OFCE Sciences PO Paris Franc); Andrea Roventini (Scuola Superiore Sant'Anna Pisa Italy also OFCE Sciences Po Paris); Joseph Stiglitz (Columbia University, New York, USA); Tania Treibich (Maastricht University and Scuola Superiore Sant'Anna,Pisa Italy & OFCE Sciences Po Paris France) |
| Abstract: | We analyze the individual and macroeconomic impacts of heterogeneous expectations and action rules within an agent-based model populated by heterogeneous, interacting firms. Agents have to cope with a complex evolving economy characterized by deep uncertainty resulting from technical change, imperfect information and coordination hurdles. In these circumstances, we find that neither individual nor macroeconomic dynamics improve when agents replace myopic expectations with less naïve learning rules. In fact, more sophisticated, e.g. recursive least squares (RLS) expectations produce less accurate individual forecasts and also considerably worsen the performance of the economy. Finally, we experiment with agents that adjust simply to technological shocks, and we show that individual and aggregate performances dramatically degrade. Our results suggest that fast and frugal robust heuristics are not a second-best option: rather they are “rational” in macroeconomic environments with heterogeneous, interacting agents and changing “fundamentals”. |
| Keywords: | Complexity, expectations, heterogeneity, heuristics, learning, agent based model, computational economics |
| JEL: | C63 E32 E6 G01 G21 O4 |
| Date: | 2017–12–14 |
| URL: | http://d.repec.org/n?u=RePEc:fce:doctra:1732&r=cmp |
| By: | Giorgio Ffagiolo (Scuola Superiore Sant'Anna Pisa Italy); Daniele Giachini (Scuola Superiore Sant'Anna Pisa Italy); Andrea Roventini (Scuola Superiore Sant'Anna Pisa Italy also OFCE Sciences Po Paris) |
| Abstract: | This paper extends the endogenous-growth agent-based model in Fagiolo and Dosi (2003) to study the finance growthnexus. We explore industries where firms produce a homogeneous good using existing technologies, perform R&D activities to introduce new techniques, and imitate the most productive practices. Unlike the original model, we assume that both exploration and imitation require resources provided by banks, which pool agent savings and finance new projects via loans. We find that banking activity has a positive impact on growth. However,excessive financialization can hamper growth. In- deed, we find a significant and robust inverted-U shaped relation between financial depth and growth. Overall, our results stress the fundamental (and still poorly understood) role played by innovation in the finance-growth nexus. |
| Keywords: | Agent based models, Innovation, Exploration vs Exploitation, Endogenous Growth, Banking Sector, Finance Growth Nexus |
| JEL: | C63 G21 O30 O21 |
| Date: | 2017–11–27 |
| URL: | http://d.repec.org/n?u=RePEc:fce:doctra:1728&r=cmp |
| By: | F. M. Stefan; A. P. F. Atman |
| Abstract: | Behavioral Finance has become a challenge to the scientific community. Based on the assumption that behavioral aspects of investors may explain some features of the Stock Market, we propose an agent based model to study quantitatively this relationship. In order to approximate the simulated market to the complexity of real markets, we consider that the investors are connected among them through a small world network; each one has its own psychological profile (Imitation, Anti-Imitation, Random); two different strategies for decision making: one of them is based on the trust neighborhood of the investor and the other one considers a technical analysis, the momentum of the market index technique. We analyze the market index fluctuations, the wealth distribution of the investors according to their psychological profiles and the rate of return distribution. Moreover, we analyze the influence of changing the psychological profile of the hub of the network and report interesting results which show how and when anti-imitation becomes the most profitable strategy for investment. Besides this, an intriguing asymmetry of the return rate distribution is explained considering the behavioral aspect of the investors. This asymmetry is quite robust being observed even when a completely different algorithm to calculate the decision making of the investors was applied to it, a remarkable result which, up to our knowledge, has never been reported before. |
| Date: | 2017–11 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:1711.08282&r=cmp |
| By: | Shuheng Wang; Guohao Li; Yifan Bao |
| Abstract: | Application of fuzzy support vector machine in stock price forecast. Support vector machine is a new type of machine learning method proposed in 1990s. It can deal with classification and regression problems very successfully. Due to the excellent learning performance of support vector machine, the technology has become a hot research topic in the field of machine learning, and it has been successfully applied in many fields. However, as a new technology, there are many limitations to support vector machines. There is a large amount of fuzzy information in the objective world. If the training of support vector machine contains noise and fuzzy information, the performance of the support vector machine will become very weak and powerless. As the complexity of many factors influence the stock price prediction, the prediction results of traditional support vector machine cannot meet people with precision, this study improved the traditional support vector machine fuzzy prediction algorithm is proposed to improve the new model precision. NASDAQ Stock Market, Standard & Poor's (S&P) Stock market are considered. Novel advanced- fuzzy support vector machine (NA-FSVM) is the proposed methodology. |
| Date: | 2018–01 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:1801.00681&r=cmp |
| By: | Iacopo Savelli; Antonio Giannitrapani; Simone Paoletti; Antonio Vicino |
| Abstract: | The European market clearing problem is characterized by a set of heterogeneous orders and rules that force the implementation of heuristic and iterative solving methods. In particular, curtailable block orders and the uniform purchase price (UPP) pose serious difficulties. A block is an order that spans over multiple hours, and can be either fully accepted or fully rejected. The UPP prescribes that all consumers pay a common price, i.e., the UPP, in all the zones, while producers receive zonal prices, which can differ from one zone to another. The market clearing problem in the presence of both the UPP and block orders is a major open issue in the European context. The UPP scheme leads to a non-linear optimization problem involving both primal and dual variables, whereas block orders introduce multi-temporal constraints and binary variables into the problem. As a consequence, the market clearing problem in the presence of both blocks and the UPP can be regarded as a non-linear integer programming problem involving both primal and dual variables with complementary and multi-temporal constraints. The aim of this paper is to present a heuristic-free, exact and computationally tractable model, which solves the market clearing problem in the presence of both curtailable block orders and the UPP. By resorting to an equivalent UPP formulation, the proposed approach results in a mixed-integer linear program, which is built starting from a non-linear integer bilevel programming problem. Numerical results using real market data are reported to show the effectiveness of the proposed approach. |
| Date: | 2017–11 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:1711.07731&r=cmp |
| By: | Louise Columelli (EURECOM); Miguel Núñez del Prado (Universidad del Pacífico); Leoncio Zarate-Gamarra (Peru IDI) |
| Abstract: | In the last decades, mobile phones have become the major medium for communication between humans. The site effect is the loss of subscribers. Consequently, Telecoms operators invest in developing algorithms for quantifying the risk to churn and to influence other subscribers to churn. The objective is to prioritize the retention of subscribers in their network due to the cost of obtaining a new subscriber is four times more expensive than retaining subscribers. Hence, we use Extremely Random Forest to classify churners and non-churners obtaining a Lift value at 10% of 5.5. Then, we rely on graph-based measures such as Degree of Centrality and Page rank to measure emitted and received influence in the social network of the carrier. Our methodology allows summarising churn risk score, relying on a Fuzzy Logic system, combining the churn probability and the risk of the churner to leave the network with other subscribers. |
| Keywords: | Churn, Data mining, Classification and Fuzzy logic |
| Date: | 2016–12 |
| URL: | http://d.repec.org/n?u=RePEc:pai:wpaper:16-22&r=cmp |