A utilização de Redes Neurais Artificiais para previsibilidade de retornos: uma comparação com os modelos lineares CAPM e Três Fatores.

Claudio Pilar Silva, Márcio André Machado

Resumo


OBJETIVO
O presente artigo teve por objetivo aplicar e comparar o modelo estatístico de redes neurais com os modelos lineares CAPM e três fatores de Fama e French para previsão de retornos no mercado acionário brasileiro e também verificar se esse modelo permite a obtenção de melhores previsões para os investidores, possibilitando o aumento de suas riquezas.

METODOLOGIA
Para o desenvolvimento do estudo, optou-se pela formação de 18 portfólios, com base na metodologia de Fama e French. Para testar as hipóteses de pesquisa, foram utilizados quatro modelos econométricos. Especificamente, construíram-se dois modelos lineares (CAPM e Três Fatores) e dois modelos não lineares (RNA). Por fim, para a comparação dos modelos, foram utilizadas três medidas de erro para se mensurar a acurácia na previsibilidade dos modelos.

RESULTADOS E CONCLUSÕES
Inicialmente, observou-se que não há diferença de previsibilidade para os modelos lineares, indo de acordo com os achados de Cao et al. (2011), onde uma possível explicação para o caso brasileiro seria o fato da não observância dos prêmios para tamanho e B/M no período analisado (Machado & Medeiros, 2011). Além do mais, verificou-se que os modelos de três fatores de Fama e French (1993) e RNAM apresentaram boa previsibilidade dos retornos na maioria das carteiras, não se podendo chegar a um consenso do melhor modelo para o período em estudo, pois a previsibilidade média de ambos os modelos é bastante próxima, hipótese também confirmada pela não rejeição do teste para comparação de médias de Wilcoxon.

IMPLICAÇÕES PRÁTICAS
As redes neurais testadas tiveram desempenho semelhante ao dos métodos tradicionais. Esse resultado sugere a preferência pelo método mais prático.

PALAVRAS-CHAVE
Modelos Econométricos, Redes Neurais Artificiais, CAPM.



THE USE OF ARTIFICIAL NEURAL NETWORKS IN THE RETURN PREDICTABILITY: A COMPARISON WITH LINEAR MODELS CAPM AND THREE FACTORS.

OBJECTIVE
This article aims to apply and compare the statistical model of neu-ral networks with linear models CAPM and three factors of Fama and French (1993) to forecast returns in the Brazilian stock market and also verify that this model makes it possible to obtain better forecasts for investors, enabling the increase of their wealth.

METHODOLOGY
To develop the study, we opted for the formation of 18 portfolios, based on the methodology of Fama and French (1993). To test the research hypotheses, we used four econometric models. Specifically, it built two linear models (CAPM and three factors) and two non-linear models (ANN). Finally, to compare the models, we used three error measures to measure the accuracy in the predictability of the models.

RESULTS AND CONCLUSIONS
Initially, there was observed that no predictability difference for line-ar models, going according to the findings of Cao et al. (2011), where a possible explanation for the Brazilian case is the fact of non-compliance with awards for size and B / M in the analyzed period (Machado & Medeiros, 2011). Moreover, it was found that the models of three factors of Fama and French (1993) and mANN showed good predictability of returns in most portfolios, not being able to reach a consensus of the best model for the study period, as the average predictability of both models is very close, a hypothesis also supported by not rejecting the test for comparing the means of Wilcoxon.

PRACTICAL IMPLICATIONS
The tested neural networks had similar performance to traditional methods. This result suggests the preference for the more practical method.

KEYWORDS
Econometrics Models, Artificial Neural Networks, CAPM.

Referências


Adebiyi, A. A., Ayo, C. K., Adebiyi, M. O., & Otokiti, S. (2012). Stock Price Prediction using Neural Network with Hybridized Market Indicators. Journal of Emerg-ing Trends in Computing and Information Sciences Vol 3 No 1.

Avci, E. (2007). Forecasting daily and sessional returns of the ISE-100 index with neural network models. Doğuş Üniversitesi Dergisi 8 (2).

Ball, R. (1978). Anomalies in Relationships between Securi-ties Yields and yield-surrogates. Journal of Financial Economics Vol 6.

Banz, R. W. (1981). The relationship between return and market value of common stocks. Journal of Financial Economics Vol 9 No 1.

Basu, S. (1977). Investment performance of common stocks in relation to their price-earnings ratios: a test of the efficient market hypothesis. Journal of Finance Vol 32 No 3.

Basu, S. (1983). The relationship between earnings yield, market value, and return for NYSE common stocks: further evidence. Journal of Financial Economics Vol 12.

Bekaert, G., Harvey, C. R., & Lundblad, C. (2007). Liquidity and Expected Returns: Lessons from Emerging Mar-kets. Review of Financial Studies Vol 20 (6).

Bhattacharya, U., Daouk, H, Jorgenson, B., & Kehr, C. (2000). When an event is not an event: the curious case of an emerging market. Journal of Financial Economics Vol 55 Issue 1.

Black, F. (1972). Capital market equilibrium with restricted borrowing. Journal of Business Vol 45.

Black, F., Jensen, M. C., & Scholes, M. (1972). The capital asset pricing model: some empirical tests. Studies in the theory of capital markets, ed. Michael Jensen. New York: Praeger.

Blume, M., & Friend, I. (1973). A new look at the capital as-set pricing model. Journal of Finance. v. 28, n. 1, pp. 19-33, 1973.

Carhart, M. M. (1997). On persistence in mutual fund per-formance. Journal of Finance Vol 52 No 1.

Cavalheiro, E. A., Ceretta, P. S., Tavares, C. E. M., & Trin-dade, L. L. (2010). Previsibilidade de mercados: um es-tudo comparativo entre Bovespa e S&P500. Sociais e Humanas, Santa Maria, Vol 23 No 01.

Cavalheiro, E. A., Vieira, K. M., & Ceretta, P. S. (2011). A-plicação de Redes Neurais Polinomiais GMDH na Previ-são do Índice Ibovespa. Revista CAP Vol 05 No 5.

Chan, L. K. C., Hamao, Y., & Lakonishok, J. (1991). Fun-damentals and stock returns in Japan. Journal of Fi-nance Vol 46 No 5.

Cao, Q, Leggio, K. B., & Schniederjans, M. J. (2005). A comparison between Fama and French’s model and ar-tificial neural networks in predicting the Chinese stock market. Computers & Operations Research Vol 32.

Cao, Q, Parry, M. E., & Leggio, K. B. (2011). The three-factor model and artificial neural networks: predicting stock price movement in China. Ann Oper Res Vol 185.

Chen, A.S., Leung, M.T., & Daouk, H. (2003). Application of neural networks to an emerging financial market: fore-casting and trading the Taiwan Stock Index. Comput-ers & Operations Research 30(6).

Copeland, T. E., Weston, J. F., & Shastri, K. (2005) Finan-cial theory and corporate policy. Pearson Addison Wes-ley.

Darbellay, G. A., & Slama, M. (2000). Forecasting the short-term demand for electricity: Do neural networks stand a better chance? International Journal of Forecasting Vol 16.

Dase, R. K., Pawar, D. D., & Daspute, D. S. (2011). Method-ologies for Prediction of Stock Market: an Artificial Neural Network. International Journal of Statistika and Mathematika Vol. 1.

Egeli, B., Ozturan, M., & Badur, B. (2003). Stock market prediction using artificial neural networks. In Proceed-ings of the third Hawaii international conference on business, Honolulu, Hawai.

Fama, E. F. (1970). Efficient capital markets: a review of theory and empirical work. Journal of Finance Vol 25 No 2.

Fama, E. F., & French, K. R. (1992). The cross-section of expected stock returns. Journal of Finance Vol 47.

Fama, E. F., & French, K. R. (1993). Common risk factors in the returns on bonds and stocks. Journal of Financial Economics Vol 33.

Ferson, W., & Harvey, C. R. (1993). The Risk and Predicta-bility of International Equity Returns. Review of Finan-cial Studies 6.

Gencay, R. (1996). Non-linear prediction of security returns with moving average rules. Journal of Forecasting 15.

Harvey C. R. (1995). Predictable risk and returns in emerg-ing markets. The Review of Financial Studies Vol 8.

Kara, Y., Boyacioglu, M. A., & Baykan, O. K. (2011). Predict-ing direction of stock price index movement using arti-ficial neural networks and support vector machines: The sample of the Istanbul Stock Exchange. Expert Systems with Applications 38.

Karsoliya, S. (2012). Approximating Number of Hidden layer neurons in Multiple Hidden Layer BPNN Architecture. International Journal of Engineering Trends and Tech-nology Vol 3.

Keene, M. A., & Peterson, D. R. (2007). The importance of liquidity as a factor in asset pricing. The Journal of Fi-nancial Research Vol 30 No 1.

Lintner, J. (1965). The valuation of risk assets and the se-lection of risky investments in stock portfolios and cap-ital budgets. Review of Economics and Statistics Vol 47 No 1.

Maassoumi, E., & Racine, J. (2002). Entropy and predicability of stock market returns. Journal of Eco-nometrics 107.

Machado, M. A. V., & Medeiros, O. R. (2011). Modelos de precificação de ativos e o efeito liquidez: evidências empíricas no mercado acionário brasileiro. Revista Brasileira de Finanças Vol 9.

Maciel, L. S., & Ballini, R. (2009). Design a neural network for time series financial forecasting: accuracy and ro-bustness analysis. In: Encontro Brasileiro de Finanças.

Mirzamohammadi, S., & Mansouri, M. (2013). Application of Regression and Neural Networks for Prediction of the Profit Rise on Assets Log in Tehran Stock Exchange. Journal of Basic and Applied Scientific Research Vol 3 (5).

Olson, D., & Mossman, C. (2003). Neural network forecasts of Canadian stock returns using accounting ratios. In-ternational Journal of Forecasting Vol 19.

Pérez-Rodríguez, J. V., Torra, S., & Andrada-Félix, J. (2005). STAR and ANN models: forecasting perfor-mance on the Spanish “Ibex-35” stock index. Journal of Empirical Finance 12.

Portugal, M. S. (1995). Neural Networks versus time series methods: a forecasting exercise. Revista Brasileira de Economia Vol 49 No 4.

Portugal, M. S., & Fernandes, L. G. L. (1996). Redes neurais artificiais e previsão de séries econômicas: uma intro-dução. Nova Economia Vol 6 No 1.

Rambalducci, M. J. G., Candido, M. R. L., & Dalmas, J. C. (2003). Avaliação da previsibilidade de retornos de a-ções emergentes negociadas no Brasil pelo coeficiente beta. Rev. Ciên. Empresariais da UNIPAR Vol 4.

Refenes, A.P. (1995). Testing Strategies and Metrics. In: Neural Networks in the Capital Markets. John Wiley & Sons, Inc.

Rosenberg, B., Reid, K., & Lanstein, R. (1985). Persuasive Evidence of Market Inefficiency. Journal of Portfolio Management Vol 11.

Rostagno, L. M., Kloeckner, G. O., & Becker, J. L. (2004). Previsibilidade de Retorno das Ações na Bovespa: Um Teste Envolvendo o Modelo de Fator de Retorno Espe-rado. Revista Brasileira de Finanças Vol 2 No 2.

Sharpe, W. F. (1964). Capital asset prices: a theory of mar-ket equilibrium under conditions of risk. Journal of Fi-nance Vol 19 No 3.

Stattman, D. (1980). Book Values and Stock Returns. The Chicago MBA: Journal of Selected Papers Vol 4.

Wallstrom, P., & Segerstedt, A. (2010). Evaluation of fore-casting error measurements and techniques for inter-mittent demand. Int. J. Production Economics Vol 128.

White, H. (1988). Economic prediction using neural net-works: The case of IBM daily stock returns. In Proceed-ings of the IEEE International Conference on Neural Networks.


Texto completo: PDF

Apontamentos

  • Não há apontamentos.




Licença Creative Commons
Este trabalho está licenciado sob uma Licença Creative Commons Attribution 3.0 .

_______________

Revista de Finanças Aplicadas

ISSN: 2176-8854

Contador Estatístico

Indexadores e diretórios

DOI Logotipo Gale Cengage Learning DOAJ

Adicionais
Gale Cengage Learning
Latindex - Sistema Regional de Información en Línea para Revistas Científicas de América Latina, el Caribe, España y Portugal
Livre - Revistas de Livre Acesso
Sumários - Sumários de Revistas Brasileiras
PKP - Public Knowledge Project