📊What the Pros Keep Quiet About Forecasting
📊What the Pros Keep Quiet About Forecasting
Insider Info
Wall Street Insider Report
Expert Guidance in Wall Street. Join +1.5k Wall Street Insiders across 30 US states and 51 countries who are scaling 10x their investments Mastering the Market Cycles.
In a session some years ago with Robert Dittmar from the University of Michigan, he uncovered the use of machine learning in financial forecasting. His focus was on how certain firm-specific metrics, like historical returns. He spotlighted how blending signals could significantly refine investment strategies.
Robert walked me through setting up a neural network tailored for this purpose. He explained the structure, starting from the input layer, where data such as market values ratios are introduced, progressing through a hidden layer that processes the data, and culminating with an output layer that forecasts returns.
This approach not only demonstrated a practical application of neural networks but also highlighted their potential to enhance the precision and efficiency of investment strategies.
How does Granger causality testing between your portfolio returns and economic indicators influence your investment timing?
Everyday our Premium Subscribers send questions through our Wall Street Direct Line, and I recently received this question:
In our ongoing exploration of advanced investment strategies, such as those discussed with Dittmar, we've learned the importance of using sophisticated analytical tools like neural networks to enhance forecasting and decision-making processes. And we can go even deeper using Granger Causality Testing.
Granger Causality Testing.
Granger causality testing is a statistical method used to determine if one time series can predict another.
Essentially, it breaks down into a few steps:
You take two sets of data over time, such as economic indicators and stock market returns.
You then use past values of one dataset (say, economic indicators) to see if they can consistently forecast future values of the other dataset (like stock returns). If the past values of the economic indicators provide statistically significant information that improves the prediction of stock returns, then economic indicators are said to "Granger-cause" stock returns.
This doesn’t imply true causality, but it suggests a predictive relationship that can be incredibly useful for forecasting and making informed decisions based on the observed patterns.
Granger causality testing is one such method that plays a crucial role in the strategic operations at Zurique Capital Research.
By employing Granger causality, we assess the predictive relationships between economic indicators and our clients’ portfolio returns, which allows us to understand which economic variables might influence future returns.
This understanding significantly impacts our investment timing. If an economic indicator Granger-causes portfolio returns, this implies that changes in these indicators can predict future performance changes.
As a result, this analysis becomes instrumental in adjusting our clients’ asset allocation and rebalancing their portfolio.
For instance, if an increase in interest rates is found to Granger-cause a decrease in stock returns, we may adjust our asset allocations before the effects materialize in the broader market. This proactive approach ensures we're well-positioned to act swiftly and effectively when clear market signals are identified, thus optimizing our performance against market cycles and volatility.
Mastering the Market Cycle
Over the last decade, we have helped our clients scale their investment portfolios by tenfold, mastering the Market Cycles and I sharing the blueprint here, in the Wall Street Insider Report.
If you want expert guidance to elevate the investment journey with personalized support and follow my Portfolio, in Real Time from a Real-life brokerage account, subscribe today and join +1.5k Wall Street Insiders across 30 US states and 51 countries.