Unpacking the Sentiment Dynamics of the Trump Trade War: A Derivative-Based Analysis

The trade war initiated by former President Donald Trump, starting notably in 2018, marked a tumultuous chapter in international economic relations. Characterized by a barrage of tariffs, escalating rhetoric, and shifting alliances, the trade war fundamentally altered market sentiments — among investors, consumers, and global political actors alike. But how did sentiment evolve during this period? More intriguingly, can we mathematically analyze the rate of change of sentiment — its “derivatives” — to better understand public and market reaction?

In this blog, we delve deep into a novel approach: applying sentiment derivatives to examine the Trump trade war discourse. Leveraging natural language processing (NLP) techniques and derivative-based sentiment analysis, we aim to uncover the underlying emotional dynamics and momentum shifts that shaped perception throughout the trade war.

Context: The Trump Trade War in Brief

Before the analytical dive, a quick refresher. The Trump administration’s aggressive trade policies targeted primarily China, with tariffs imposed on billions of dollars of goods. The stated goal was to reduce trade deficits and protect American industries. Yet, this strategy disrupted global supply chains, triggered retaliatory tariffs, and stoked fears of economic deceleration.

Media coverage, social media discourse, and investor sentiment oscillated wildly, reflecting confusion, hope, anger, and uncertainty. Traditional sentiment analysis captures snapshots of these emotions, but lacks the nuance to interpret how quickly sentiment changed or whether those changes accelerated or slowed — which can be critical to anticipating market moves or policy shifts.

What Are Sentiment Derivatives?

In calculus, derivatives represent the rate of change of a function. Applied to sentiment analysis, the first derivative of sentiment over time tells us how fast the emotional tone is shifting — are people becoming rapidly more positive or negative? The second derivative tells us about the acceleration of this change — is sentiment change speeding up or slowing down?

This layered insight helps detect inflection points, market panic, or growing optimism — often before absolute sentiment scores reveal them.

Methodology: Data and Tools

We sourced a vast corpus of trade war-related text:

• News articles from 2017–2021 (Reuters, Bloomberg, CNN, Fox Business)
• Social media posts, primarily Twitter, using hashtags #TradeWar, #ChinaTariffs, and #TrumpTariffs
• Financial analyst reports and earnings call transcripts mentioning trade policy

Using advanced NLP models, including VADER sentiment scoring and BERT-based contextual sentiment classifiers, we extracted daily sentiment scores for each source category.

We then computed the first and second derivatives of sentiment time series, smoothing data with rolling averages to reduce noise. This process enabled us to identify momentum shifts and sentiment volatility across platforms.

Findings: Sentiment Trends and Their Derivatives

1. Initial Shock and Escalation Phase (Early 2018)

• Sentiment Score: Sharp decline across all media, reflecting widespread concern over tariffs and potential economic fallout.
• First Derivative: Negative spikes indicating rapidly worsening sentiment, coinciding with tariff announcements on steel and aluminum
• Second Derivative: Positive in short bursts, indicating moments when the rate of sentiment decline slowed, possibly due to political reassurances or trade negotiation optimism.

2. Mid-Trade War Uncertainty (Mid-2018 to Early 2019)

• Sentiment Score: Hovered near neutral with high variance — mixed positive and negative news (e.g., talks of progress vs. new tariff threats).
• First Derivative: Oscillating around zero, reflecting sentiment volatility and market confusion.
• Second Derivative: Several negative troughs signaling sentiment momentum was decelerating and sometimes reversing, aligned with actual trade negotiation breakdowns.

3. Phase of Partial De-escalation (Mid-2019 to Early 2020)

• Sentiment Score: Gradual improvement as “Phase One” trade deal talks progressed.
• First Derivative: Positive values indicate improving sentiment, albeit at a modest pace.
• Second Derivative: Mostly positive, suggesting growing confidence and accelerating positive sentiment. Yet spikes of negativity corresponded with tariff increases on tech products.

4. COVID-19 Era and Trade War Lingering Effects (2020–2021)

• Senment Score: Mixed, with pandemic overshadowing trade but uncertainty about future tariffs lingering.
• First Derivative: Fluctuating, with temporary positive rebounds as markets hoped for normalized relations.
• Second Derivative: Less volatile than earlier periods but still reflecting sensitivity to political rhetoric.

Insights: Why Sentiment Derivatives Matter

• Early Warning: The first derivative flags rapid sentiment deterioration early, a critical alert for investors and policymakers.
• Momentum Detection: The second derivative identifies whether sentiment shifts are accelerating or losing steam, providing foresight into market confidence or anxiety.
• Platform Variance: Social media sentiment derivatives tend to be more volatile than traditional news, highlighting the role of public emotion and viral discourse.
• Policy Feedback Loop: Sharp sentiment shifts influenced political decision-making — a dynamic interplay between public perception and policy moves.

Implications for Stakeholders

Investors

Monitoring sentiment derivatives could enhance trading strategies, allowing earlier reaction to policy developments or market sentiment changes before price moves occur.

Policymakers

Understanding public sentiment momentum helps tailor communication strategies, aiming to stabilize market reactions or build confidence during trade negotiations.

Media Analysts

Derivative sentiment analysis can guide editorial focus, identifying when discourse intensifies and where misinformation or panic may be brewing.

Challenges and Limitations

• Data Noise: Sentiment derivatives amplify noise; smoothing techniques must be carefully chosen.
• Context Sensitivity: NLP models can struggle with sarcasm or coded language typical in political discourse.
• Attribution: Isolating trade war impact from concurrent events (like COVID-19) complicates analysis.
• Cross-Platform Differences: Harmonizing sentiment across heterogeneous sources requires normalization.

Conclusion: A New Lens on Trade War Sentiment

The Trump trade war’s complex emotional landscape defies simple analysis. Sentiment derivatives offer a powerful tool to decode the velocity and acceleration of public and market emotions — revealing patterns and turning points hidden beneath surface-level sentiment.

As trade tensions continue to shape global economics, integrating derivative-based sentiment analysis could transform how stakeholders anticipate and respond to geopolitical risks.