A Simple Guide to Using AI in Economic Analysis

The article “A Simple Guide to Using AI in Economic Analysis” was originally posted on Alpha Architect blog.

The guest article below was submitted via a friend (Sailesh Radha) who develops software that helps investors analyze data (Borealis). We have no affiliation/connection to the firm, but the founder’s son is currently serving in the military, and we like to support veterans and their families. What is interesting about this case study/tool is that it highlights—at a practical level—how AI can make analysis more efficient (but you still probably need a human to minimize hallucinations!).

1. Introduction: What is Sentiment Analysis?

Sentiment analysis is a common AI technique used in finance to automatically gauge the underlying “feeling” or “tone” of a piece of text. It is a foundational tool for understanding large volumes of information quickly.

In basic terms, sentiment analysis:

• Uses technologies like Natural Language Processing (NLP) and machine learning.
• Identifies subjective information within text-based data, like news articles or reports.
• Classifies the text’s tone into simple categories such as positive, negative, or neutral.

This technique can be surprisingly powerful. For instance, the AI firm Alexandria Technology created a sentiment index that tracked positive versus negative news coverage of the U.S. housing market. This index revealed a persistent decline in positive sentiment in the months leading up to the 2008 financial crisis, serving as a powerful early warning signal.

While useful for a quick read on market mood, this simple labeling has critical limitations when analyzing complex economic situations.

2. The Problem with Simple Labels: Why Context is King

Labeling complex economic news as just ‘positive’ or ‘negative’ can be misleading because it often ignores the objective context of the information. The true impact of an economic event depends entirely on who is being affected.

Let’s consider a simple statement from a financial report: “The price of oil is expected to rise.”

A traditional sentiment analysis tool might label this as “positive” because “rise” is regularly associated with growth. However, the real economic impact is completely different depending on a country’s situation.

As the table shows, a single event has opposite effects. Traditional sentiment analysis, by assigning a single label, completely misses the opposing macroeconomic implications for different countries. For serious financial and economic analysis, this lack of context is a major flaw.

To overcome this challenge, a more advanced method is needed—one that measures objective impact, not just subjective tone.

3. The Advanced Solution: Resonance Analysis^© by Macro Pulse AI

To address the shortcomings of simple sentiment analysis, Borealis Global Analytics developed an advanced method called Resonance Analysis^©, which powers their Macro Pulse AI platform.

Instead of looking for a positive or negative tone, this approach treats each piece of news as a “macroeconomic impulse”—essentially a shock to the economic system. It then measures the ripple effects of that shock.

Resonance Analysis:

• Measures the net spillover effects of an economic event on key sectors like consumption, investment, and trade.
• Calculates a proprietary metric called the Economic Ripple Score (see Figure below).
• Provides a much more nuanced measurement on a continuous scale from -1 to +1, quantifying the systemic economic impact.

Fig. 1: Economic Ripple Score

The difference between the two methodologies is stark (see Fig. 2).

This advanced approach provides a far more sophisticated and accurate picture of how economic events will truly play out across the globe.

Fig. 2: Resonance Analysis, A New Framework Beyond Sentiments

4. Key Takeaway: From Vague Feelings to Measurable Impact

The fundamental difference between these two AI methods is what they are designed to measure. Traditional sentiment analysis tells you about the feeling of a text, while the more advanced Resonance Analysis tells you about its predictable economic consequence.

For any student of finance or economics, understanding the benefits of the Macro Pulse AI approach is key to appreciating the future of data analysis:

1. Objective Measurement: It moves beyond subjective labels like “positive” or “negative” to quantify the real-world economic impact of an event.
2. Context-Aware: It understands that the same event (like rising oil prices) has fundamentally different effects on different economies, providing a country-specific analysis.
3. Actionable Insight: It provides a more dynamic and forward-looking measure, allowing for a more profound understanding of how economies will likely be affected by new information.

Illustrative Case: Canada’s Trade Environment (October 2024–April 2025)

A Six-Month Forward-Looking Analysis

To demonstrate how Macro Pulse AI transforms raw policy documents into structured, forward-looking intelligence, this case study examines three sequential Bank of Canada Monetary Policy Reports spanning October 2024 through April 2025—a period marked by escalating trade tensions and macroeconomic uncertainty.

The analysis illustrates how the platform’s Economic Ripple Score methodology captures not only current conditions but also anticipates downstream impacts before they materialize in hard economic data.

October 2024: Cautious Optimism with Emerging Risks

The October 2024 outlook reflects a balanced macroeconomic assessment, driven primarily by structural improvements in Canada’s energy export infrastructure. The platform identified three key positive drivers:

• Export Capacity Expansion: The completion of the Trans Mountain pipeline increased oil export capacity, reducing reliance on U.S. refinery purchases and diversifying market access.
• Fiscal Strengthening: Higher energy export revenues improved government fiscal positions, supporting public investment capacity.
• Labor Market Effects: Infrastructure development generated high-skill employment, bolstering household consumption through rising disposable income.

However, forward-looking risk signals tempered the outlook. With the U.S. presidential election weeks away, Macro Pulse AI detected nascent concerns regarding potential tariff implementation—a qualitative signal embedded in policy language that presaged subsequent disruptions. The resulting Economic Ripple Score of 0.282 reflected this mixed but cautiously positive environment.

Here’s the excerpt from our Macro Pulse AI platform

January 2025: Anticipating Policy Shocks

By January 2025, the macroeconomic narrative had shifted decisively. Although the Bank of Canada’s baseline scenario explicitly excluded tariff assumptions, Macro Pulse AI extracted forward-looking language indicating heightened probability of punitive trade measures—specifically, tariffs up to 25% and reciprocal Canadian retaliation.

Critically, the platform’s analysis occurred two months before “Liberation Day” tariff implementation, yet successfully captured the anticipated economic impact:

• Policy Tone Analysis: Detection of conditional risk language in central bank communications.
• Scenario Sensitivity: Quantification of downside GDP impacts under tariff scenarios (-2.5% GDP growth reduction in Year 1).
• Investment Sentiment: Identification of business investment contraction driven by policy uncertainty.

The Economic Ripple Score declined sharply to -0.36, reflecting the material deterioration in forward-looking expectations—despite tariffs not yet being enacted. This demonstrates the platform’s capacity to transform qualitative policy signals into quantified risk assessments ahead of realized outcomes.

Here’s the excerpt from our Macro Pulse AI platform

April 2025: Realized Disruption and Compounding Effects

The April 2025 outlook validated the January forward-looking assessment. “Liberation Day” tariffs were implemented, triggering the anticipated trade conflict and supply chain fragmentation. Macro Pulse AI’s analysis captured the cascading economic effects:

• Export Demand Collapse: 25% tariffs on steel, aluminum, and automotive goods sharply reduced industrial export volumes, with export growth contracting by 1.0% in 2025 and 2.4% in 2026.
• Investment Freeze: Trade policy uncertainty drove business investment contraction of approximately 2%, as firms delayed capital expenditures amid supply chain instability.
• Inflationary Pressure: Retaliatory Canadian tariffs and currency depreciation raised consumer prices, with goods price inflation accelerating to 1.3% by March 2025, pushing headline inflation above the 3% threshold.

The Economic Ripple Score further deteriorated to -0.55, reflecting the compounding nature of trade disruptions on consumption, investment, and net exports—core GDP components.

Here’s the excerpt from our Macro Pulse AI platform

Key Analytical Insights:

This six-month case study illustrates three critical capabilities of Macro Pulse AI:

1. Anticipatory Signal Extraction: The platform identified trade policy risks in October 2024—before the U.S. election—and quantified potential impacts in January 2025, two months before tariff implementation.
2. Dynamic Scoring Under Uncertainty: Economic Ripple Scores evolved systematically as forward-looking language shifted from cautious optimism (0.282) to anticipatory risk (-0.36) to realized disruption (-0.55), providing portfolio managers with a continuous risk barometer.
3. Multi-Channel Impact Assessment: The analysis mapped tariff impacts through GDP components (consumption via inflation, investment via uncertainty, net exports via demand shocks), demonstrating the platform’s integration of qualitative policy signals with quantitative macroeconomic modeling.

By automating the extraction, contextualization, and scoring of forward-looking insights from central bank communications, Macro Pulse AI enables asset managers to incorporate macro intelligence into portfolio construction with greater speed, consistency, and foresight than traditional manual research workflows.

Please feel free to get a deep dive on Resonance Analysis and Economic Ripple Score here.