Machine Learning Approaches for Asset Pricing Model and Relative Valuation Model: Focusing on the Korean Market

Abstract

I study on the application of Machine Learning in finance. To this end, we deal with two studies on the asset pricing model, which is a research field in traditional financial finance, and the relative valuation model, which is widely used in the industry. By applying machine learning methods to these two fields, it shows that the use of artificial intelligence methods in the financial field can be significant. The first study applied an autoencoder to an asset pricing model. This study analyzes the explanatory power of the latent factor conditional asset pricing model for the Korean stock market using an autoencoder. The autoencoder is a type of neural network in machine learning that can extract latent factors. Specifically, we apply the conditional autoencoder (CA) model that estimates factor exposure as a flexible nonlinear function of covariates. Our main findings are as follows. The CA model showed excellent explanatory power not only in the entire sample but also in several subsamples in the Korean market. Also, because of this explanatory power, it can better explain market anomalies compared to the traditional asset pricing models. As a result of examining investment strategies using pricing error, the CA model measures the expected return of stocks better than the traditional asset pricing model. In addition, the CA model indicates that the firm characteristic variables are important in asset pricing conditional on macro-financial states, such as the global financial crisis and the coronavirus disease 2019 pandemic. The result shows that the major variables considered in the explanation of stock returns through the CA model may vary depending on the time. This is expected to provide a broader perspective on asset pricing through the CA model in the future. The second study applies various machine learning methods to the valuation model. This study proposes a relative evaluation method using machine learning, which is widely used in the industry. Out of sample shows that the machine learning-based model examined in this study is superior to the existing model. In addition, it provides excellent explanatory power in other samples when learning from companies with a high market capitalization or a older firm. This shows that it can be useful for evaluating peer group selection or companies in the early stages of listing. If the valuation metrics predicted by machine learning are a good proxy for actual company intrinsic value, overvalued stocks will go down in price and undervalued stocks will go up in price over the next month. As a result of evaluating strategies using this, a statistically significant rate of return can be obtained. This indicates that stock valuation methods through machine learning can also be useful for real-world investments. This research highlights the potential benefits of using machine learning methods in the financial industry. By applying machine learning to traditional finance research fields such as asset pricing models and relative valuation models, we found that machine learning methods can significantly enhance the accuracy and efficiency of financial analysis. Overall, the results suggest that machine learning has the potential to revolutionize the financial industry and provide new opportunities for real-world investments. It is important for financial institutions to continue investing in and supporting the development of machine learning in finance to fully realize these benefits.