Skip to main content

Why does seeing two rising lines in a chart make us assume one caused the other?

When you read financial news, charts with two upward-trending lines side by side feel like proof that one movement caused the other. A headline shows Bitcoin rising alongside the stock market, or inflation climbing while central bank rates stayed flat, and the visual pattern convinces you that you've understood the causal link. In reality, correlation—two things moving together—is not the same as causation, one thing causing the other. This distinction is one of the most misused concepts in financial reporting, and missing it will lead you to make confident but wrong predictions about markets and policy.

Quick definition: Correlation means two variables move together (up or down); causation means one variable directly causes the change in the other. Charts often show correlation but journalists and commentators interpret them as proof of causation, leading readers astray.

Key takeaways​

  • Two variables can rise or fall together for reasons that have nothing to do with each other, making spurious correlation dangerous in financial journalism.
  • A chart showing a strong visual correlation between two trends does not prove that one caused the other—there may be a third factor, coincidence, or reverse causation.
  • Confounding variables (a third factor affecting both) are especially common in financial and economic data and invisible in a simple two-line chart.
  • News stories built on "x caused y" claims require evidence beyond a matching chart: experiment, detailed mechanism, or longitudinal analysis.
  • Learning to ask "what else could explain this pattern?" is your first defense against misleading financial headlines.

The danger of seeing two lines move together​

When a financial news outlet publishes a chart with two lines—say, Fed interest rates on one axis and stock market returns on the other, both sloping downward—your brain instantly searches for the causal story. "The Fed raised rates, so stocks fell." That feels right, and the visual correlation reinforces it. But the chart alone does not answer whether the Fed caused the stock decline or whether both responded to a shared third cause (like inflation worries), or whether the stock market actually fell before the rate hike (making it a poor predictor, not a cause).

This trap is especially dangerous in finance because markets are real-time, reactive systems. Everything influences everything else. When you see two lines on a chart that trend together, your first instinct should not be "A caused B," but rather "A and B both moved—what is the actual causal mechanism, and what evidence would support it?" A chart is a summary of correlation, not a proof of causation.

Consider a real-world example: Over the past decade, cryptocurrency prices have risen sharply, and so have venture-capital funding amounts. A headline might read "VC money floods crypto as digital assets soar." Both lines go up. But correlation here is not causation. VC funding may be responding to adoption trends, regulatory clarity, or broader tech sentiment, not the crypto price itself. Or the causation might run the other way—VC funding for blockchain startups drives awareness and adoption, which then increases prices. Or both move in response to a third force like Federal Reserve policy loosening money supply. A chart showing them together tells you they are correlated; it does not tell you which one (if either) caused the other.

Confounding variables hide in every financial dataset​

The most common reason correlation fails as proof of causation is the presence of a confounding variable—a third factor that influences both of the variables you're looking at, creating the illusion of a direct link between them.

Imagine a chart showing unemployment declining alongside wage growth over the last five years. The headline reads "Lower joblessness pushes wages higher." But what if the real story is that a booming tech sector (the confounding variable) both reduced unemployment and drove wage growth among skilled workers? Or that low interest rates (another confounder) encouraged hiring and wage competition at the same time? The chart shows correlation but hides the real causal driver.

In financial markets, confounders are everywhere. When inflation falls and the stock market rises, it feels like "lower inflation caused stock gains." But consider what else was happening: central banks may have paused rate hikes (pushing valuations up), corporate earnings guidance may have improved (a separate positive), and consumer sentiment may have strengthened (a third reason stocks would rise). A chart showing inflation and stock price alone cannot separate these overlapping forces. You need data breaking down the sources of market movement—or you are guessing based on a visual pattern, not reasoning from evidence.

Reverse causation and timing matter more than the chart shows​

Another hidden assumption in correlated charts is the direction of causality. If chart shows A and B rising together over a ten-year period, human intuition often assumes A caused B. But what if B caused A, or both moved in response to a hidden C?

Real example: Financial news frequently publishes charts showing Fed interest rates rising alongside stock-market volatility. The story often reads "higher rates are making stocks more volatile." But timing matters. Did volatility rise after the Fed announcement, or before? If volatility spiked ahead of the rate hike, then the Fed was responding to market conditions, not causing them. Reverse causation flips the entire causal narrative, and a simple chart showing two correlated lines cannot reveal it.

This timing problem is rampant in macro news. When oil prices rise and the dollar weakens, journalists often write "stronger oil demand is weakening the dollar." But the causation might run the other way: a weaker dollar makes U.S. oil cheaper for foreign buyers, so they buy more, pushing oil prices up. Or both respond to investor risk appetite (strong demand for commodities, weak demand for safe-haven dollar). A two-line chart cannot show timing or reveal reverse causation.

The difference between correlation strength and causal proof​

Here's a subtle but critical point: even a very strong correlation (lines that move nearly in lockstep) is not proof of causation. This is counterintuitive. If two lines track each other perfectly over years, shouldn't one be causing the other?

No. The classic example is the correlation between shoe size and reading ability among elementary schoolchildren. They move together (older kids have bigger feet and read better), but shoe size does not cause reading ability. The confounding variable is age. Or consider Nicolas Cage film releases and swimming-pool drownings—they have been shown to correlate strongly by year, but one does not cause the other.

In financial journalism, a strong correlation often gives a false sense of certainty. "Tech stock earnings growth and AI-related news sentiment have a 0.92 correlation"—this sounds like a causal link, but it only means they move together. They could both be responding to the same underlying reality (growing AI adoption driving real earnings and hype simultaneously), or the correlation could be coincidence over a short time period.

The strength of the correlation tells you nothing about causation. A moderate correlation with a clear temporal sequence and mechanism (A happens first, then B, and here's why A would cause B) is far more convincing evidence than a perfect correlation with no plausible mechanism.

Common mistakes in interpreting correlated charts​

1. Assuming visual strength means causal strength. A chart where two lines move in perfect sync looks compelling, but visual power and causal certainty are unrelated. A chart showing two variables that are perfectly correlated by coincidence over a short period (five years) is visually strong but causally meaningless. A chart showing imperfect correlation (0.6 or 0.7) combined with clear timing, mechanism, and control for confounders is far more reliable.

2. Ignoring the time period. A correlation that holds over ten years may break down in the next five years because the underlying causal mechanism changed or confounding factors shifted. Financial headlines often highlight recent correlations (tech earnings up 30% and AI stocks up 40% in the past year) without noting that this same correlation did not hold in the previous decade. Zoom out, and the causal claim weakens.

3. Mistaking statistical significance for practical significance. A correlation of 0.4 between two variables might be "statistically significant" (meaning it's unlikely to occur by chance alone) but still explain only a small fraction of the movement in either variable. A journalist might write "study finds correlation between Fed rate moves and unemployment" without noting that the Fed's decisions explain only 15% of unemployment changes; 85% comes from other factors. The chart is technically accurate but misleading about causation.

4. Not asking what else could explain the pattern. When you see a financial headline built on a correlated chart, always ask: "What other factors could produce this same chart?" If you can think of plausible alternatives, the headline is not evidence of causation, only a story. For instance, "rising grocery prices and wage growth move together" could reflect (a) wage growth driving inflation, (b) inflation expectations rising before wages adjust, (c) labor shortages driving both wages and input costs higher, or (d) fiscal stimulus inflating both. The chart does not distinguish these stories.

Real-world examples​

Example 1: Bitcoin and inflation expectations. In 2020–2021, Bitcoin price and U.S. inflation expectations (measured by the 5-year inflation-expectation swap rate) both rose sharply. Financial headlines claimed "Bitcoin is an inflation hedge" and showed correlated charts. The causation story: investors fearful of inflation buying Bitcoin, pushing its price up. But the correlation was partly spurious. Both Bitcoin and inflation expectations were driven by a common shock: unprecedented government stimulus and central-bank money printing. Bitcoin rose because broader sentiment shifted toward risk assets and speculation, not because inflation fears directly caused Bitcoin buying. When inflation expectations fell in 2022 but Bitcoin kept falling further, the causation story fell apart. The chart had shown correlation, not causation.

Example 2: Fed rate hikes and market returns. In 2022, the Federal Reserve raised rates from near zero to 4.25–4.5% in twelve months. Stock markets fell roughly 18% (S&P 500). A straightforward headline: "Fed rate hikes trigger stock decline." The chart showed rising rates and falling stock price, moving in opposite directions with remarkable consistency. But the causation is muddied by timing and confounders. Rate hikes were a response to inflation that had already damaged earnings expectations. Earnings forecasts fell before the rate hike cycle ended. Corporate bond spreads (the extra yield investors demand for risk) widened, indicating rising default risk independent of the Fed's choice. Multiple factors drove stocks down; the rate hikes were both a cause and a response to the deteriorating environment. A simple chart could not separate these forces.

Example 3: Oil prices and airline stock prices. When oil prices rise, airline stock prices typically fall (airlines face higher fuel costs). This correlation is strong and holds across many time periods. Is it causal? Yes, in part—oil is a direct input cost, and higher fuel costs reduce airline profitability. But the correlation also reflects a confounding variable: both oil and airline stocks respond to economic growth expectations. If recession fears rise, oil prices fall (weak demand) and airline prices fall (weak travel demand). If growth expectations improve, both rise. The chart shows strong correlation, and causation is partly real (fuel costs do affect profits) but also partly spurious (both respond to shared economic conditions). A financial journalist writing "rising oil prices hurt airline stocks" is accurate but incomplete; they are also both responding to shifts in demand and growth.

FAQ​

A: Correlated variables are related—they move together. But being related is not the same as one causing the other. They are related because a causal mechanism connects them, or because a third variable influences both, or by coincidence. The chart shows the relationship; you must do additional work to determine the cause.

Q: How do I know if a causal claim in financial news is reliable?​

A: Look for evidence beyond the chart. Does the journalist explain the mechanism (why A would cause B)? Does the timing make sense (did A happen before B)? Are confounding variables considered or ruled out? Is there data from multiple time periods or even a controlled experiment? Strong causal claims rest on more than visual correlation.

Q: Can a journal article or academic study claiming causation still be wrong about financial charts?​

A: Yes. Even peer-reviewed studies sometimes claim causation based on correlation. The study may use statistical controls for some confounders but miss others, or rely on data from a specific period that does not generalize. Always ask: what was the method, could reverse causation explain the results, and do the authors acknowledge limitations?

Q: Does a high R-squared or correlation coefficient (like 0.9) mean one variable is causing the other?​

A: No. R-squared tells you how much of the variation in one variable is explained by another, but explanation is not causation. A correlation of 0.9 between ice cream sales and swimming-pool visits makes sense (both respond to summer heat), but neither causes the other. A high correlation without a causal mechanism is still just correlation.

Q: If the causation is partly real and partly spurious (like airline stocks and oil prices), how should I interpret financial news about it?​

A: Acknowledge both parts. Yes, oil prices affect airline profitability directly. Yes, both respond to economic conditions. A complete reading of the financial environment requires disentangling these forces, not accepting the correlated chart as a full explanation. The chart is a starting point, not a conclusion.

Q: Why do financial journalists use correlated charts if they can be so misleading?​

A: Correlated charts are easy to create and visually compelling. They tell a simple story that feels intuitive. Journalists are often under time pressure and may not have the statistical background to question the causal claims built on them. Your job as a reader is to maintain skepticism where the chart alone cannot provide evidence.

Summary​

When a financial news chart shows two variables moving together, correlation is visible but causation is not. The most dangerous assumption is that a strong visual correlation proves one variable caused the other. Confounding variables, reverse causation, and timing are invisible in a simple chart. By learning to ask "what else could explain this pattern" and demanding evidence beyond the visual correlation—clear mechanism, timing, control for confounders—you protect yourself from headlines that mistake correlation for causation. Charts summarize data; they do not prove causation. Causation requires reasoning, not just looking.

Next​

→ Base effects in charts