Key Takeaways
- Seasonal adjustment is a statistical process that removes recurring seasonal patterns from economic data so that the underlying trend is visible.
- Without seasonal adjustment, economic data would be dominated by predictable seasonal effects — higher retail sales in December, lower construction in winter, higher employment in summer.
- Most headline economic indicators (CPI, retail sales, NFP, GDP) are reported as seasonally adjusted figures.
- Seasonal adjustment factors are recalculated periodically, which can cause small revisions to previously released data.
- For forex traders, the key insight is that seasonally adjusted figures show the underlying trend, while non-seasonally adjusted figures show the actual level — and the market reacts to the seasonally adjusted figure.
- Understanding seasonal adjustment helps traders avoid misinterpreting normal seasonal patterns as economic trends.
What Is Seasonal Adjustment?
Seasonal adjustment is a statistical process that removes recurring seasonal patterns from economic data so that the underlying trend is visible. Many economic variables follow predictable seasonal patterns — retail sales spike in December for holiday shopping, construction activity falls in winter, employment rises in summer as students enter the workforce, energy prices rise in winter for heating. These patterns repeat every year and are not related to the underlying economic trend.
Without seasonal adjustment, economic data would be dominated by these seasonal effects. A 15% jump in retail sales from November to December would look like an economic boom, but it is simply the normal holiday shopping pattern. Seasonal adjustment removes these predictable patterns, so that the remaining change reflects genuine economic activity rather than seasonal noise. See Macroeconomic Data for Traders.
Why Economic Data Is Seasonally Adjusted
Economic data is seasonally adjusted for three key reasons. First, comparability: seasonal adjustment allows traders and analysts to compare month-to-month changes without being misled by seasonal patterns. Without adjustment, you could not tell whether a 0.3% MoM CPI change was high or low for that month, because the seasonal pattern varies by month. With adjustment, the 0.3% figure is directly comparable to other months.
Second, trend identification: seasonal adjustment makes the underlying trend visible. If you look at non-seasonally adjusted retail sales, the December spike and January plunge would obscure the underlying trend. Seasonal adjustment smooths out these patterns, revealing whether retail sales are genuinely growing or shrinking.
Third, policy relevance: central banks need to see the underlying trend, not seasonal noise, to make policy decisions. The Federal Reserve does not want to raise rates because of a normal December retail sales spike; it wants to know whether the underlying trend in consumer spending is accelerating. This is why most headline economic indicators are reported as seasonally adjusted figures.
How Seasonal Adjustment Works
Seasonal adjustment is performed by statistical agencies using established methods, most commonly the US Census Bureau's X-13ARIMA-SEATS method or similar approaches. The process involves estimating the seasonal pattern from historical data and then removing it from the current observation. The result is a seasonally adjusted series where the seasonal pattern has been removed, leaving the trend-cycle and irregular components.
The key points for traders to understand are:
- Seasonal factors are estimated from historical data: The agency looks at several years of past data to identify the recurring seasonal pattern for each month or quarter.
- Factors are updated periodically: Seasonal patterns can change over time (e.g., online shopping has shifted the December retail pattern). Agencies recalculate seasonal factors annually, which can cause small revisions to previously released data. See Economic Data Revisions.
- Adjustment is not perfect: Unusual events (e.g., a pandemic, a major storm) can disrupt the normal seasonal pattern, making the seasonal adjustment less accurate for that period.
- Different agencies may use different methods: The BLS, BEA, ONS, Eurostat and other agencies all use seasonal adjustment, but the specific methods may differ slightly, which can affect comparability across countries.
Seasonally Adjusted vs Non-Seasonally Adjusted
Some indicators are reported in both seasonally adjusted (SA) and non-seasonally adjusted (NSA) forms. The key differences:
| Characteristic | Seasonally Adjusted (SA) | Non-Seasonally Adjusted (NSA) |
|---|---|---|
| Seasonal pattern | Removed | Present |
| Month-to-month comparability | Yes — figures are directly comparable | No — seasonal patterns dominate |
| Shows actual level | No — adjusted to remove seasonal effects | Yes — shows the actual reported value |
| Market focus | Markets react to SA figures | NSA figures rarely move markets |
| Year-on-year comparison | Less affected by seasonality | YoY naturally removes seasonality (same month, same season) |
For most forex traders, the seasonally adjusted figure is the one that matters — it is what the market reacts to and what central banks use for policy. But understanding that the NSA figure exists and what it shows can help you interpret data more fully. For example, if you want to know the actual level of retail sales (not the trend), you would look at the NSA figure.
Which Indicators Are Seasonally Adjusted
Most major economic indicators are reported as seasonally adjusted figures, but some are not:
- CPI: Reported as both SA and NSA. The headline MoM figure is typically SA; the YoY figure is typically NSA (because YoY naturally removes seasonality). See CPI and Inflation.
- Non-Farm Payrolls: The headline employment change is SA. The unemployment rate is also SA. See Non-Farm Payrolls Explained.
- Retail sales: Reported as both SA and NSA. The headline MoM figure is SA. See Retail Sales.
- GDP: Reported as SA (quarterly growth rates are SA and annualised). See GDP Explained.
- Industrial production: Reported as SA.
- Trade balance: Reported as both SA and NSA; the headline is typically NSA.
- Jobless claims: The headline is typically SA, but NSA figures are also published and can be more volatile due to seasonal effects.
When Seasonal Adjustment Can Fail
Seasonal adjustment is based on historical patterns, so it can be less accurate when those patterns are disrupted. The most notable example was the COVID-19 pandemic, which disrupted normal seasonal patterns in employment, retail sales, and many other indicators. Seasonal adjustment factors estimated from pre-pandemic data did not accurately capture the unusual patterns of 2020-2021, leading to distorted seasonally adjusted figures and larger-than-normal revisions.
Other events that can disrupt seasonal patterns include major weather events (hurricanes, severe winters), government shutdowns (which delay or distort data collection), and structural changes in the economy (e.g., the shift from brick-and-mortar to online retail). When seasonal adjustment is less reliable, traders should be more cautious in interpreting the data and pay more attention to revisions.
Seasonal Adjustment and Revisions
Because seasonal adjustment factors are recalculated annually, previously released data can be revised when the factors are updated. This is one reason why economic data is revised — the seasonal adjustment changes, which changes the seasonally adjusted figure even if the underlying NSA data has not changed. See Economic Data Revisions.
This is particularly relevant at the beginning of a new year, when many agencies update their seasonal adjustment factors. Traders should be aware that January data releases may include revisions to the previous year's figures due to seasonal factor updates.
Seasonal Adjustment and Market Expectations
Seasonal adjustment interacts with market expectations in an important way. The consensus forecast is typically expressed in seasonally adjusted terms — economists predict the seasonally adjusted figure, not the raw figure. The market prices in the seasonally adjusted consensus, and the surprise is calculated using the seasonally adjusted actual. This means the market reaction depends on the quality of the seasonal adjustment, not just the underlying data.
If the seasonal adjustment factors are wrong — for example, if a weather shock distorts the seasonal pattern — the seasonally adjusted figure may give a misleading signal. The market may react to a seasonally adjusted surprise that is actually an artifact of bad seasonal adjustment, not a genuine economic shift. This is why experienced traders check whether unusual weather, holidays, or one-off events may have distorted the seasonal adjustment before interpreting a surprise. See Why Markets Trade Expectations, Not Just Data and Actual vs Forecast vs Previous.
What is already priced in also matters. If the market expects a strong seasonally adjusted figure (because economists have raised their forecasts), a beat may produce a smaller reaction. If the market expects a weak figure, a beat may produce a larger reaction. The repricing runs through the same expectations channel as any other data release — but the added uncertainty of seasonal adjustment means the market may be more cautious in its repricing. See Consensus Expectations.
From Seasonally Adjusted Data to Currency Moves
The market reacts to the seasonally adjusted figure, not the NSA figure. The transmission runs:
Seasonally adjusted data release → change in growth/inflation expectations → change in rate expectations → change in bond yields → currency repricing
A higher-than-expected SA CPI MoM may strengthen the currency by raising inflation expectations and central-bank tightening probability. The NSA figure is rarely market-moving because it is dominated by seasonal effects and does not reveal the underlying trend. See Interest Rates and Forex Markets.
Regime Dependency: When Seasonal Adjustment Matters Most
Seasonal adjustment matters most when the seasonal pattern is strong and the underlying trend is subtle. For indicators with large seasonal swings (retail sales, construction), seasonal adjustment is essential for identifying the trend. For indicators with small seasonal swings (CPI, unemployment rate), seasonal adjustment is less critical but still standard practice. During periods of economic disruption (pandemic, major weather events), seasonal adjustment is less reliable and traders should be more cautious. See How Macro Regimes Change Forex Relationships.
Relative FX Analysis: Both Sides of the Pair
FX is relative. A higher US SA CPI does not determine EUR/USD solely from the dollar side. The correct analysis compares the US inflation data against what is happening with euro-area inflation. If US SA CPI beats but euro-area SA CPI also beats, EUR/USD may not move much. If US CPI beats while euro-area CPI misses, EUR/USD is likely to fall. Always compare like with like — SA vs SA. See Economic Growth Differentials.
Common Mistakes
- Comparing NSA figures month-to-month: NSA figures are dominated by seasonal effects and are not comparable across months.
- Ignoring that seasonal factors are updated: Annual updates can revise previously released data.
- Assuming seasonal adjustment is perfect: Unusual events can disrupt seasonal patterns, making adjustment less accurate.
- Confusing SA and NSA figures: Always check which figure you are looking at — the market reacts to SA.
- Forgetting that YoY naturally removes seasonality: YoY figures are less affected by seasonality because the same month a year ago had the same seasonal pattern.
- Ignoring the other currency's data: A USD SA beat does not guarantee EUR/USD falls if EUR-side SA data is also strong.
Practical Framework
- Confirm the figure is seasonally adjusted: Most headline figures are SA, but always check.
- Compare to consensus: Is the SA figure a beat or a miss? See Consensus Expectations.
- Check for seasonal factor updates: At the start of a year, be aware that seasonal factor updates may revise previous data.
- Be cautious during disruptions: During pandemics, major weather events, or structural shifts, seasonal adjustment may be less reliable.
- Assess central-bank implications: Does the SA data change rate expectations?
- Check the other side of the pair: What is happening with the counter-currency's SA data?
- Consider the macro regime: Is the market particularly sensitive to this indicator right now?
- Identify what would invalidate the interpretation: What subsequent data or revision would change the read?
Understanding seasonal adjustment is essential for correctly interpreting economic data. MacroDrivers® evaluates currencies using relative macro conditions across eight major currencies, ensuring seasonally adjusted data is always interpreted in the context of both sides of the pair.