Skip to main content

How do dual Y-axis charts create false correlations?

A single chart with two datasets, each on its own vertical axis (a dual Y-axis chart), is one of the most deceptive visualizations in financial journalism. The trick is simple: if you give each dataset its own scale, you can make almost any two time series look perfectly synchronized even if they're weakly or inversely correlated. A news outlet can overlay Fed interest rates and stock prices on separate Y-axes and make them look like they move in lockstep, when their actual correlation is loose. This creates the illusion of causation and misleads readers about what's driving markets.

This article teaches you how dual-axis charts work, when they're legitimate, and how to spot the manipulation.

Quick definition: A dual Y-axis chart plots two datasets on the same timeline but with different vertical scales (left and right Y-axes), allowing the chart maker to stretch or compress each line's visual movement independently and create the appearance of correlation where little or none exists.

Key takeaways​

  • Dual-axis charts allow independent scaling of two datasets, making any two time series appear correlated if both are scaled to the same visual range.
  • Financial news outlets use dual-axis charts to suggest causation (e.g., "Fed rates drive stock prices") even when correlation is weak or absent.
  • A visual sync on a dual-axis chart is not evidence of correlation; it's only evidence that the scales were chosen to align the movements.
  • Detecting manipulation requires comparing the datasets on the same scale or checking the actual correlation coefficient.
  • Legitimate uses of dual axes exist (comparing metrics with very different units or scales) but are rare in financial news.
  • Professional investors ignore dual-axis charts unless the scales are explicitly justified and the correlation is independently verified.

How dual-axis manipulation works​

A dual-axis chart has two vertical axes: one on the left (usually for the first dataset) and one on the right (usually for the second dataset). Here's where the manipulation lies: the chart maker chooses the scale of each axis independently.

Example: Fed interest rates and S&P 500 stock prices.

Honest approach (single-axis, normalized):

  • Plot both as percentage changes from a baseline.
  • Fed rates: from 0% to 5% (absolute).
  • S&P 500: from -20% to +40% (percentage change from baseline).
  • The two lines don't move in sync because the correlation is weak.

Manipulated dual-axis approach:

  • Left Y-axis for interest rates: 0% to 5%.
  • Right Y-axis for S&P 500: scaled so that the 0–2,000 point move visually matches the 0–5% rate move.
  • The lines now appear to move in perfect sync, even though the actual correlation hasn't changed.

The chart doesn't lie about the numbers. The Fed rates are correctly shown as 0–5%, and the S&P 500 is correctly shown as its actual range. But the visual sync is artificial. A reader seeing this chart for the first time will conclude that Fed rates and stock prices are tightly correlated and likely causally linked. A reader who checks the actual correlation will find it's much weaker.

The mathematics of scale manipulation​

Here's how the scales are chosen to force correlation:

Visual height of dataset 1 = height of chart / (max value 1 − min value 1)
Visual height of dataset 2 = height of chart / (max value 2 − min value 2)

If you want both datasets to span the same visual height (say, 80% of the chart), you can scale the right Y-axis to compress or stretch dataset 2 as needed.

Example: Fed rates moved from 0% to 4% (range of 4). S&P 500 moved from 3,800 to 4,800 (range of 1,000). To make them span the same visual height:

  • Left axis: 0% to 4% (rate range).
  • Right axis: 3,800 to 4,800 (price range).

Now both span the same visual distance even though 4 percentage points and 1,000 price points are measuring very different things. The chart has forced them to look synchronized.

Why financial journalists use dual axes​

Dual-axis charts are popular because they're easy to produce and visually compelling. Modern charting tools (Excel, Tableau, Python libraries) can generate dual-axis charts with a single command. The resulting visual is engaging and suggests a story (correlation) even if the story isn't supported by the data.

Financial outlets use dual-axis charts to support narratives:

  • "Fed policy drives stocks" → overlay Fed rates and stock prices on dual axes.
  • "Dollar strength hurts earnings" → overlay dollar index and company earnings on dual axes.
  • "Oil prices drive inflation" → overlay oil and CPI on dual axes.

Each narrative might be partially true, but a dual-axis chart is not evidence. It's a visual technique that can make any two loosely synchronized time series look perfectly correlated.

Real-world examples​

Fed rates and stock prices (2022–2024): Financial media extensively covered the relationship between Fed interest rate increases and stock market declines. Many articles included dual-axis charts showing Fed funds rate and S&P 500 on aligned scales. The visual impression is that rate increases caused stock declines (higher rates → lower valuations). This is partially true, but the actual relationship is complex. Rate increases also reflect inflation concerns, and inflation expectations matter more for stocks than absolute rate levels. A dual-axis chart that visually syncs Fed rates and stock prices masks this complexity and suggests a cleaner causation than exists.

Bitcoin and S&P 500 correlation (2023–2024): In 2023, Bitcoin rallied strongly while stocks were flat or declining. Financial commentators published dual-axis charts comparing Bitcoin and stocks, scaling the axes to show the recent divergence. But when the same outlets wanted to claim Bitcoin was a hedge against stocks (a bull narrative), they'd use different time windows and scales to make them appear negatively correlated. The same data was used to support opposite conclusions depending on axis choice.

U.S. dollar and emerging-market ETF returns (2022–2024): A strong dollar typically hurts emerging-market performance. News outlets covering EM weakness often used dual-axis charts showing dollar strength and EM returns moving in opposite directions (dollar up, EM down). The charts looked compelling and supported the narrative. But the actual correlation coefficient was only moderate, and other factors (China economic slowdown, rate differentials) mattered more. The dual-axis chart's simplicity masked the true complexity.

Treasury yields and stock valuations (2023–2024): Rising Treasury yields compress stock valuations because higher bond yields make bonds more attractive relative to stocks. News articles comparing yields and P/E ratios often used dual axes to show the synchronous movement. This narrative is directionally correct (higher yields do pressure valuations), but it oversimplifies. Earnings expectations matter too. A dual-axis chart that visually syncs yields and P/E ratios suggests a mechanical 1:1 relationship that isn't supported by actual data.

How to detect dual-axis manipulation​

Step 1: Check for two Y-axes. If the chart has both a left and right Y-axis label, it's a dual-axis chart. This is the obvious tell.

Step 2: Check the axis ranges. Are they proportional? If the left axis goes from 0 to 10 and the right axis goes from 100 to 900, the chart maker has deliberately chosen ranges to align the visuals. This is suspicious.

Step 3: Look for perfect or near-perfect sync. If the two lines move in lockstep—rising and falling at exactly the same times and magnitudes—without any divergence, the chart is almost certainly manipulated. Real correlations are messier.

Step 4: Check the source or article for a correlation coefficient. If the article claims "Fed rates and stocks are strongly correlated" but doesn't provide a correlation coefficient (usually ranging from -1 to +1), assume the visual chart is doing the work of the unstated correlation.

Step 5: Replot on a single scale. The cleanest check: find the raw data for both datasets, normalize them to percentage change from a baseline, and replot on a single Y-axis. If the visual sync disappears, the original dual-axis chart was manipulated.

Legitimate uses of dual-axis charts​

Dual axes are not always manipulative. They can be legitimate when:

Comparing metrics with fundamentally different units. For example, overlaying quarterly earnings (in millions of dollars, left axis) and earnings per share—EPS (in dollars per share, right axis). The units are different, so separate scales are necessary. But even here, a normalized comparison (both as percentage change from baseline) is cleaner.

Showing a component and its total. For example, revenue on the left axis and a revenue sub-component on the right axis. The scales reflect actual values, and the visual is just for convenience.

Displaying indices that are by definition on different scales. For example, the VIX (volatility index, ranging 10–80) and the S&P 500 (ranging 2,000–5,000). Using dual axes here is practical because the absolute numbers are so different. But the chart should be labeled as such, and the axes should accurately reflect the index ranges without manipulation.

The key difference: legitimate dual axes are clearly labeled and justified, with axis ranges that reflect the data's natural scales, not scales chosen to force visual sync.

Professional norms and practices​

Wall Street analysts and institutional research rarely use dual-axis charts and typically avoid them when possible. When dual axes are used, the correlation is explicitly stated (e.g., "correlation coefficient of 0.65") and justified.

Academic finance research avoids dual-axis charts because they're considered misleading. Peer review typically rejects papers with dual axes unless there's a compelling reason (different units with no easy normalization). When published, the axis ranges are clearly justified.

Business media frequently uses dual-axis charts because they're visually engaging and require less explanation than correlation statistics. The practice is widespread in CNBC, Bloomberg, and Reuters breakdowns of Fed policy, oil prices, and earnings-growth relationships.

Social media and financial Twitter/X almost exclusively uses dual-axis charts when comparing two datasets because the visual impact is immediate and shareable. Context and correlation coefficients are often omitted.

How to correct for dual-axis manipulation​

Once detected, correction requires deliberate investigation.

Method 1: Normalize and replot. Find the raw data and express both datasets as percentage change from a baseline date. Plot both on a single Y-axis from -50% to +50% (or wider as needed). The true visual relationship emerges. This usually takes 10–15 minutes with a spreadsheet.

Method 2: Calculate the correlation coefficient. Use a statistical tool (Excel's CORREL, Python's numpy.corrcoef, or a basic calculator) to compute the correlation between the two datasets. Correlation ranges from -1 (perfect inverse relationship) to +1 (perfect positive relationship). A correlation above 0.8 is strong; 0.5–0.8 is moderate; below 0.5 is weak. If a dual-axis chart visually shows perfect sync but the actual correlation is 0.4, the chart is manipulated.

Method 3: Check multiple time windows. Compute correlation over different time periods (last month, last year, last 5 years). If correlation varies wildly across windows, the relationship is unstable, and the dual-axis chart's implication of a stable link is false.

Method 4: Compare to controlled sources. Official sources like the Federal Reserve publish analysis of relationships (e.g., "Fed monetary policy transmissions to stock prices"). Compare the outlet's dual-axis narrative to published research. Discrepancies suggest manipulation.

Common mistakes​

Mistake 1: Assuming visual sync means causation. Two time series that move in sync could be causally related, coincidentally correlated, or linked through a third variable. The dual-axis chart shows only sync, not cause.

Mistake 2: Trusting a dual-axis chart without checking the correlation. Always ask: what's the actual correlation coefficient? If the article doesn't provide it, the visual is doing unsupported work.

Mistake 3: Using dual-axis charts for decision-making without verification. An investor who trades based on perceived Fed-stock correlation derived from a dual-axis chart is betting on visual manipulation, not reality.

Mistake 4: Forgetting that axis choice is editorial. Even if the numbers are correct, the axis ranges are choices. These choices are not neutral.

Mistake 5: Conflating short-term and long-term correlations. Fed rates and stock prices might move in sync over a 3-month window but have a weak long-term correlation. A dual-axis chart of the 3-month window is not evidence of a sustained relationship.

FAQ​

When is a dual Y-axis chart actually necessary?​

When the two datasets have fundamentally different units (dollars and percentages, or rates and prices) and normalizing them would obscure important information. But honestly, most dual-axis charts in financial news are more for visual impact than necessity.

Can I trust a dual Y-axis chart from a reputable outlet like Bloomberg or CNBC?​

Reputation doesn't guarantee accuracy. Reputable outlets do use dual-axis charts for legitimate purposes, but they also use them for manipulation. Always check the axes, look for correlation coefficients, and verify with single-axis data before trusting the narrative.

How do I know if axis ranges are chosen to manipulate or are just convenient?​

If the ranges force the two lines to move in perfect sync and a correlation coefficient or source explains why, it might be legitimate. If the ranges force visual sync but no explanation is provided and the actual correlation is weak, it's manipulation.

What's the correct way to compare two datasets with different scales?​

Normalize both to percentage change from a baseline date and plot on a single Y-axis. Or calculate correlation separately and state it in the text. Or use small multiples (side-by-side charts, each on its own scale, clearly separated). Dual-axis charts are rarely the best approach.

Can I use dual-axis charts in my own analysis?​

Use them with caution and only when the axes are clearly justified and the correlation is independently verified. Better alternatives: side-by-side charts, correlation statistics, or normalized single-axis plots.

Summary​

Dual Y-axis charts are a powerful visual trick that creates false correlations by allowing independent scaling of two datasets. When each dataset is scaled to the same visual span, any two loosely synchronized time series will appear perfectly correlated, even if their actual correlation is weak. Financial news outlets use dual-axis charts to support narratives about causation (Fed policy drives stocks, oil prices drive inflation, etc.) without providing correlation statistics. Detection requires checking for two Y-axes, assessing whether the visual sync is real or forced by axis choice, and calculating the actual correlation coefficient. Legitimate dual-axis uses exist but are rare and clearly justified. A professional financial reader either verifies the correlation independently or treats the dual-axis chart as visual marketing rather than evidence.

Next​

→ Misleading bar charts