How does the intensity of U.S. military interventions influence stock market reactions?
The Intuition We Wanted to Prove
It seems obvious: deadlier wars should hurt markets more. If an intervention involves more troops, more casualties and more escalation, the economic damage should be proportionally larger. We also expected that longer conflicts would compound these effects over time, creating a clear relationship between war intensity and market pain.
The reality is more nuanced. Our analysis reveals that intensity alone doesn't predict market outcomes. We observe a correlation between duration and negative returns, but here's the catch: 20-year conflicts like Afghanistan inevitably overlap with other major economic events like the 2008 financial crisis, Iraq War and the dot-com bubble. Correlation doesn't prove causation and we cannot isolate duration's true causal effect from these confounding factors.
Not all military interventions are equal in scale or severity. Some are brief and symbolic, while others are prolonged, escalatory and carry substantial human and geopolitical costs.
To analyze how markets respond to these differences, we construct a single intensity index that captures multiple dimensions of intervention severity.
But how do you combine lethality, escalation and duration into a single number without making arbitrary choices? This is where PCA comes in.
Understanding the Intensity Index
How we quantify severity without guessing
The Problem: Avoiding Arbitrary Choices
We could have simply decided that fatalities matter twice as much as duration or that troop deployment is the most critical factor. But that would be subjective, we would be imposing our human bias on the historical data.
The Solution: Letting the Data Speak
Instead, we used Principal Component Analysis (PCA). This technique allows the data itself to determine the mathematical "weight" of each factor based on how they correlate naturally.
How it works: If high fatalities and long durations tend to happen together, PCA detects this pattern and combines them into a single signal. The result is a purely data-driven Intensity Score from 0 to 1 that captures the true "magnitude" of an intervention.
What Goes Into the Intensity Score?
Lethality (Weight: 0.65)
The human cost. Based on fatality counts (log-transformed to handle the huge range from minor skirmishes to major wars). This is the strongest contributor and deadlier conflicts score higher on intensity.
Escalation (Weight: 0.58)
How aggressive is the posture? Combines 9 variables including hostility levels, whether it's declared a "war," and military activity indicators. Higher escalation means more aggressive military engagement.
Duration (Weight: 0.49)
How long does it last? Measured in days (log-transformed). Longer interventions indicate sustained military commitment. Interestingly, this has the lowest weight, suggesting lethality and escalation matter more for "intensity."
Why Does This Matter?
The first principal component explains 55% of the total variance across these three dimensions. That's a moderate correlation, meaning lethality, escalation and duration tend to move together, but not perfectly. Some interventions are deadly but brief like Desert Storm, while others are prolonged but low-intensity like peacekeeping missions. This nuance is exactly what we need to test whether intensity predicts market reactions.
Now comes the crucial test of our Intensity-Impact hypothesis (H5): do high-intensity interventions actually cause bigger market crashes?
Each point in the visualization below represents one of our 29 interventions, positioned by its intensity score and the market's cumulative response.
If our hypothesis is correct, we should see a clear downward slope where more intensity equals worse returns. But look closely at the outliers. They tell a different story.
Intensity vs Market Reaction
Each point represents an intervention. Toggle between return and volatility metrics.
The Tale of Two Wars: Afghanistan vs. Desert Storm
Afghanistan: Persistent Negative Returns
The War in Afghanistan stands out with a -28% cumulative abnormal return, the sum of daily market underperformance over the entire conflict. At intensity 0.86 and lasting 7,632 days, it's both severe and prolonged. This cumulative drag represents the sustained economic cost of open-ended military commitments compounded over 20 years.
Desert Storm: High Intensity, Positive Returns
Operation Desert Storm (intensity: 0.59) achieved +7.46% CAR despite being a major military operation. The difference? It lasted just 348 days with a clear objective and decisive outcome. Markets don't fear intensity; they fear uncertainty. A swift, successful operation actually boosted confidence by demonstrating U.S. military capability and removing the uncertainty of "what happens next."
The Duration Confound (H6)
This is where our Duration Matters hypothesis (H6) becomes complex. Vietnam (3,092 days, intensity 0.91) and Afghanistan (7,632 days, intensity 0.86) are the two longest conflicts and they're also the two worst performers. However, correlation doesn't prove causation. Multi-decade conflicts inevitably overlap with other major economic events like the 2008 financial crisis, concurrent Iraq War, and the dot-com bubble. We cannot isolate duration's true causal effect from these confounding factors.
Key Insight
Intensity alone doesn't predict market outcomes. While longer conflicts correlate with worse returns, we must be cautious about causation. Multi-decade wars overlap with other major economic events, making it impossible to isolate duration's true effect. Quick, high-intensity operations with clear objectives can actually boost markets by reducing uncertainty. Markets price in geopolitical risk gradually and don't panic proportionally to how many troops are deployed.
The scatter plot tells a story, but can we quantify it? Let's look at the statistical relationship between intensity and market reactions across all 29 interventions.
We group interventions into four intensity quartiles to see if the pattern holds when aggregating the data.
Correlation by Intensity Quartile
Average market reaction grouped by intervention intensity level.
Error bars show standard deviation.
What the Statistics Actually Tell Us
Return Correlation
Loading correlation statistics...
Volatility Correlation
Loading correlation statistics...
Translating the Numbers to Words
What Does r = -0.36, p = 0.057 Mean?
The Pearson correlation of -0.36 suggests a moderate negative relationship: as intensity goes up, returns tend to go down. But here's the catch: the p-value of 0.057 is just above the traditional 0.05 threshold. In plain terms, there's about a 6% chance this pattern is just random noise. We can't confidently claim intensity predicts returns.
Why Pearson and Spearman Disagree
Pearson measures linear relationships while Spearman measures rank-based relationships. Pearson says -0.36 (moderate negative), but Spearman says +0.06 (essentially zero). This discrepancy is a red flag: a few extreme outliers like Afghanistan and Vietnam are pulling the Pearson correlation down, but when we just look at rankings, there's no trend.
The Q4 Collapse: Duration in Disguise
The quartile chart shows cumulative returns (summed over each intervention's duration). The highest intensity quartile (Q4: 0.49-0.91) averages -264.73% cumulative return compared to +39.43% for the other three quartiles combined. But this extreme gap is misleading because Q4 contains Vietnam (3,092 days) and Afghanistan (7,632 days), the two longest conflicts. When you sum daily returns over 20 years, even small daily losses become enormous. Duration, not intensity, drives these cumulative numbers.
The Statistical Verdict
The borderline p-value (~0.06) means we cannot confidently claim that intensity alone predicts market returns. The relationship exists but is confounded by duration. High-intensity interventions tend to last longer and it's the prolonged exposure to uncertainty that hurts markets, not the intensity per se. This is why our Intensity-Impact hypothesis (H5) is only partially confirmed: intensity matters, but only when combined with duration.
Aggregate relationships between intensity and returns reveal broad trends, but individual interventions often deviate from the average pattern.
This comparison tool allows interventions to be examined side by side, highlighting how differences in intensity translate into sector-level outcomes.
By directly contrasting two cases, the tool helps clarify when intensity amplifies risk and when other factors dominate market behavior.
Compare Interventions
Select two interventions to compare their sector impacts side by side
Comparison Insights
Across our 29 interventions, 58.6% achieved positive cumulative abnormal returns, challenging the assumption that military action always hurts markets. The key differentiator isn't intensity alone, but how intensity interacts with duration and strategic clarity.
The Intensity Paradox
High-intensity interventions (>0.5) average -3.33% CAR, but this masks huge variation. The Lebanese Civil War (intensity: 0.73) achieved +4.33% CAR while Afghanistan (intensity: 0.86) suffered -28.08% CAR. The difference? Lebanon lasted 581 days with clear objectives; Afghanistan dragged on for 7,632 days.
The Regime Change Puzzle
Desert Storm and Just Cause had nearly identical intensity (~0.58), yet Desert Storm returned +7.46% while Just Cause lost -1.72%. The sector breakdown reveals why: Desert Storm boosted Consumer Discretionary by +0.73% daily while Just Cause dragged it down -0.67%.
Sector Resilience
Financial Services and Consumer Defensive are the most stable sectors during interventions (std dev ~0.26%). Meanwhile, Telecommunications shows the highest volatility (std dev 1.33%), making it a high-risk, high-reward sector during military conflicts.
Energy Sector Surprise
Counter-intuitively, Energy doesn't always benefit from military interventions. The sector averaged -0.23% daily returns across all interventions. The best Energy performance came from Gulf of Sidra (+0.81% MAR), while Libya-Egypt tensions saw the worst (-1.84% MAR).