Time Serial-Driven Risk Assessment in Trade Finance: Leveraging Stock Market Trends with Machine Learning Models

Abstract

There is a lot of power in the Stock Market over both national and global markets. What affects stock prices? The performance of the industry, the news and success of the company, the confidence of investors, and both small and large-scale economic factors such as wage rates and employment rates. Trends in stock prices can be figured out by looking at the things that cause them and how the stock has done in the past. This research suggests using Long Short-Term Memory (LSTM)-based deep learning to evaluate risk in trade finance. It does this by using Yahoo Finance data from the stock market. The suggested LSTM model takes advantage of temporal correlations in sequential financial data to get an F1-score of 96.32%, an accuracy of 96.17%, a precision of 96.89%, a recall of 95.76%, and an accuracy of 96.17%. Compared to other machine learning models, the suggested model works better because it gets 91.2% accuracy for Support Vector Machine (SVM) and 88.72% accuracy for Random Forest (RF). The suggested model shows that it is strong and dependable by looking at its accuracy and loss curves along with its confusion matrix results. Improving the way trade finance evaluates stock price risk by using an LSTM model that is better at finding complicated patterns and long-lasting connections in market data makes the process more efficient.

Keywords: Microsoft, Support Escalation Engineer. 3 Microsoft, Support Escalation Engineer Stock Market Trends, Trade Finance, Risk Assessment, Financial Technology (FinTech), Stock Market Prediction, Machine Learning, Financial Forecasting, Stock Data