Corporate Funding Strategies with AI
Authors: Eliasy, A. and Przychodzen, J.
Journal: Array, Volume 6
Publication Date: Jul 2020
The stock prices that are predicted for the upcoming year against the actual values. The plot shows the ability of AI to predict stock prices quite accurately.
Summary:
In our research, we explored the potential of Artificial Intelligence (AI) to enhance the accuracy of the Capital Asset Pricing Model (CAPM) – a widely used method for estimating the cost of capital and expected returns. By predicting stock prices with AI, we aimed to optimise corporate funding strategies and maximise the value of corporations.
We studied the adjusted closing stock prices of 10 high-tech public companies from January 2013 to January 2019. Using a Recurrent Neural Network (RNN) with Long Short Term Memory (LSTM) and dropout layers, we predicted stock prices for the upcoming year. We then compared the results with traditional CAPM calculations, revealing that AI improved the accuracy of cost of equity estimations by over 60%.
Our findings challenge the Efficient Market Hypothesis (EMH) and Random Walk Theory, which claim that it is impossible to predict market prices. We demonstrated that up to one year of stock prices could be predicted with the least amount of information available, suggesting that AI can be used to replace traditional asset pricing models in the future.
However, as AI gains the ability to predict stock prices with greater accuracy, the estimation of risk, or beta, in its current form may no longer be relevant. Instead, we propose two methods for redefining risk: 1) introducing a new algorithm that incorporates uncertainty as a multiplier to the predicted returns on security, and 2) evaluating the reliability of predictions on historical data and incorporating the standard deviation as the uncertainty associated with predictions.
Our research has significant implications for corporate funding strategies. By accurately predicting stock prices and estimating the cost of capital, companies can better plan and optimise their capital structure. This new approach could be a valuable tool for financial decision-making by Chief Financial Officers (CFOs).
However, the use of AI in cost of equity estimation also presents challenges. As the cost of equity is significantly underestimated using traditional CAPM calculations, a new balance in the optimal capital structure may be needed to maximise the value of the firm for shareholders. Additionally, the availability of this information to the public may lead to further challenges.
Our study paves the way for future research to explore the impact of AI on financial decision-making, as well as the development of new parameters to evaluate the uncertainties associated with AI-predicted stock prices.