A Machine Learning Approach To Volatility Forecasting

Hi! Here's Iván with this week's exciting newsletter, brimming with insights and discoveries on building robust investment strategies and risk models using Machine Learning.

In this edition, I am presenting the following sections:

• 🕹️ AI-Finance Insights: I summarize three must-read academic papers that mix cutting-edge ML/DL with Asset Pricing & Quant Finance:

  • A Machine Learning Approach To Volatility Forecasting

  • Forecasting Cryptocurrency Prices Using Deep Learning: Integrating Financial, Blockchain, and Text Data

• 💊 AI Essentials: Today, I'm thrilled to share an outstanding video titled "PyTorch Python Tutorial | Deep Learning Using PyTorch | Image Classifier Using PyTorch | Edureka Rewind."

• 🥐 Asset Pricing Insights: In this edition, I recommend the paper "Improving Forecast Accuracy with Scaled Principal Component Analysis," which introduces a novel approach to dimension reduction for forecasting.

Today’s Sponsor

Don’t just Copy, Create your Strategy

Unlock hedge fund-level strategies with Surmount's platform. Craft, analyze, execute instantly. Full Code or No-code builders, advanced analytics elevate your portfolio precision. Experience the future: advanced tools, comprehensive data for market conquest.

Start elevating your trading today!

AI-Finance Insights

“A Machine Learning Approach To Volatility Forecasting”

The paper demonstrates the superior performance of Machine Learning (ML) techniques in one-day-ahead volatility forecasts for stocks in the Dow Jones Industrial Average (DJIA). By integrating various ML methods, the study significantly improves the accuracy of predicting realized variance. The main contributions and findings are as follows:

👉 ML models outperform the traditional Heterogeneous AutoRegressive (HAR) model, with elastic net reducing mean squared error (MSE) by 9.0% and random forest by 9.4%.

👉 Neural networks achieve the best overall performance, reducing MSE by 11.5%, with ensemble approaches further enhancing results.

👉 Including additional volatility predictors such as firm-specific characteristics and macroeconomic indicators significantly enhances ML model performance, with elastic net showing an 8.5% reduction in MSE.

👉 ML techniques automatically identify and leverage nonlinear relationships among variables, as demonstrated by Accumulated Local Effect (ALE) plots, which show the influence and ranking of dominant predictors.

👉 ML models maintain superior forecasting ability even when irrelevant, noisy variables are introduced, demonstrating resilience in complex, data-rich environments.

“Forecasting Cryptocurrency Prices Using Deep Learning: Integrating Financial, Blockchain, and Text Data”

The paper explores the application of Machine Learning (ML) and Natural Language Processing (NLP) techniques in forecasting cryptocurrency prices, specifically Bitcoin (BTC) and Ethereum (ETH). By incorporating financial, blockchain, and text data, the study aims to improve prediction accuracy through the integration of advanced deep learning methods.

The main contributions and findings are as follows:

👉 The study integrates NLP data from news and social media (Twitter and Reddit) into cryptocurrency price forecasting models. This novel approach captures public sentiment and its influence on cryptocurrency valuations, significantly enhancing the models' predictive power.

👉 Various pre-trained NLP models, such as Twitter-RoBERTa and BART MNLI, are employed to gauge market sentiment. These models, especially BART MNLI with its zero-shot classification capabilities, excel in extracting bullish and bearish signals from textual data.

👉 The research adopts both regression and classification approaches to price forecasting, including the prediction of price movements (up or down) and the identification of local extrema. The inclusion of NLP data substantially improves forecasting performance, with models consistently generating profits across different validation scenarios.

👉 Empirical results highlight the superior profitability and accuracy of models incorporating NLP data. These models exhibit higher AUC ROC (Area Under the Receiver Operating Characteristic Curve) and accuracy compared to those without NLP data. The binary price movement models demonstrate higher profitability, while local extrema models show superior precision.

Sponsor 👇

Learn AI in 5 Minutes a Day

AI Tool Report is one of the fastest-growing and most respected newsletters in the world, with over 550,000 readers from companies like OpenAI, Nvidia, Meta, Microsoft, and more.

Our research team spends hundreds of hours a week summarizing the latest news, and finding you the best opportunities to save time and earn more using AI.

Sign up with 1-Click

AI-Essentials

📹 Check out this video: "PyTorch Python Tutorial | Deep Learning Using PyTorch | Image Classifier Using PyTorch | Edureka Rewind." This Edureka PyTorch Tutorial video will help you understand the basics of PyTorch 👇

Asset Pricing Insights

“Improving Forecast Accuracy with Scaled Principal Component Analysis”

🔊 Scaled PCA -> A New Approach to Dimension Reduction. 👇

The paper introduces scaled principal component analysis (sPCA), a new method for improving forecasting. sPCA enhances traditional PCA by giving more weight to predictors with stronger forecasting power. The main contributions and findings are as follows:

👉 The study introduces sPCA, which scales predictors by their regression slopes on the target variable before applying PCA. This approach better captures the predictive information, significantly outperforming traditional PCA, especially in scenarios with weak factors or heterogeneous noise.

👉 Empirical results validate the superiority of sPCA in forecasting key U.S. macroeconomic variables such as inflation, industrial production, unemployment, and S&P 500 index volatility. The sPCA consistently delivers more accurate forecasts than PCA and other advanced techniques like LASSO and ridge regression, achieving higher Sharpe ratios and reduced maximum drawdowns.

👉 Theoretical analyses and extensive simulations support the robustness and effectiveness of sPCA, showing its capacity to outperform PCA under various conditions. sPCA's targeted dimension reduction offers significant improvements in forecast accuracy, making it a valuable tool in the era of big data.

If you're enjoying our newsletter and want to support us, please recommend it to anyone you know who's interested in AI and Finance. Your referrals are the biggest compliment and help us grow! 🌟🤖💼