Seasonally Adjusted Money Supply: How to Read M2SL and M2NS
Seasonally adjusted money supply removes estimated patterns that repeat at similar times each year. This makes nearby periods easier to compare. It does not remove inflation or change the currency. FRED labels US monthly M2SL as adjusted and M2NS as unadjusted. Both describe money balances, using different methods.
What seasonal adjustment changes
Seasonal adjustment changes how a money series is shown. It leaves the deposits themselves untouched. The US Census Bureau explains that it estimates and removes patterns that repeat through the year. It may also account for trading days and holidays that move between months. The result still includes irregular changes; it is not a pure trend. Source: Census seasonal-adjustment questions and answers.
Think of this as two questions. The unadjusted figure asks what the measured balance was. The adjusted figure asks how it compares after we remove the estimated seasonal pattern. Neither question is always better. The Bureau of Labor Statistics explains why the choice should fit the task. Source: BLS seasonal adjustment.
Seasonal adjustment and inflation adjustment also answer different questions. An adjusted M2 series can still be nominal: it uses current currency amounts. To account for prices, you need another step using a price index. We show how in nominal versus real money supply.
Which series answers your question
For US monthly money charts, check the series code. M2SL at FRED is monthly, seasonally adjusted M2 in billions of dollars. M2NS at FRED uses the same monthly frequency and dollar unit, but no seasonal adjustment. We checked these labels on 23 October 2026.
| Your question | Starting point | What to check |
|---|---|---|
| How did M2 change from the preceding month? | Seasonally adjusted monthly series | Same series, frequency, and data vintage |
| What balance did the unadjusted release report? | Not seasonally adjusted series | Whether the observation is an average or an end-period value |
| How does this month compare with the same month last year? | A consistent adjusted or unadjusted series | State which treatment you used |
| How much purchasing power does M2 represent? | Money series plus a price index | Seasonal adjustment alone does not answer this |
Scroll to see all columns
US H.6 monthly figures are monthly averages. A monthly label does not always mean the balance on the month’s final day. The Federal Reserve lists monthly averages and weekly data as distinct series. Source: H.6 release description.
When two charts disagree, compare these labels before comparing their peaks. An adjusted monthly average differs from an unadjusted end-month balance, even if both titles say “money supply.” Our methodology explains which sources and time periods M2 Index uses.
A worked seasonal-adjustment example
An unadjusted balance can rise while the adjusted balance falls. Our hypothetical example shows why. We made up these values to teach the method; they are not country data. We use a simple multiplicative model. Divide the unadjusted balance by the seasonal factor to get the adjusted balance.
| Hypothetical period | Unadjusted balance, currency units | Assumed seasonal factor | Adjusted balance, currency units |
|---|---|---|---|
| Month A | 100.00 | 1.00 | 100.00 |
| Month B | 102.00 | 1.04 | 98.08 |
Scroll to see all columns
We calculate unadjusted growth as (102 / 100 − 1) × 100 = 2.00%. For adjusted growth, we use the value before rounding: ((102 / 1.04) / 100 − 1) × 100 = −1.92%.
The assumed seasonal factor would lift a fixed adjusted balance to 104 currency units in Month B. Our teaching value reaches only 102 units. Once we remove the assumed effect, the adjusted series falls. The formula has not made money vanish. It has changed the basis for the comparison.
This example shows the math, not the Federal Reserve’s full method. The Fed builds US M2 from parts that it treats in different ways. FRED’s M2SL notes say it adds adjusted components, then subtracts unadjusted IRA and Keogh balances. A single assumed factor cannot replace that process. Source: M2SL construction notes.
Why adjusted history can change
Seasonal factors are estimates, so a new estimate can change earlier adjusted values. The Federal Reserve’s 24 March 2026 announcement reported revised factors estimated with X-13ARIMA-SEATS, using data through December 2025. The release also revised seasonally adjusted M1 and M2 levels and growth rates. Source: Federal Reserve H.6 announcements.
If your old spreadsheet differs from a fresh download, check the release date first. The maths may be right in both files. An old data point can keep its month while its estimated value changes. Our guide to money-supply data revisions explains how to record both dates.
A year-over-year comparison does not remove this issue. It compares points twelve months apart, which can reduce the effect of a stable annual pattern. But either value in the ratio can still change. Use both values from the same download unless you intend to compare data vintages.
How to compare money charts consistently
Start with the same measure. Name the money aggregate, currency, frequency, and seasonal treatment. State whether each point is an average or an end-period balance. Then choose a level, a growth rate, or an index with a clear base period. Two lines with the same color can still measure different things.
To find monthly growth, divide this month’s level by last month’s level. Subtract one, then multiply by 100. For twelve-month growth, use the value twelve months earlier. A change from one quarter to the next is not a monthly rate. Do not fill missing data points with zero.
When using our money-supply growth ranking, read the dates and source notes alongside the rate. To build your own chart, start with the data downloads and keep the source labels. We do not judge whether a series is adjusted just by how smooth its line looks.
Frequently asked questions
Is seasonally adjusted money supply inflation-adjusted?
No. Seasonal adjustment removes estimated patterns within the year. Inflation adjustment uses a price index. US M2SL is seasonally adjusted and nominal. The separate FRED series M2REAL adjusts M2 for prices. Give each step its own label.
Can adjusted and unadjusted M2 move in opposite directions?
Yes. Our hypothetical table shows an unadjusted rise of 2.00% and an adjusted fall of 1.92%. The assumed seasonal rise is larger than the rise we see. This changes how we read the growth rate. It does not create or destroy deposits.
Should I always use seasonally adjusted M2?
No. Use adjusted data when seasonal patterns would mask the change you want to study. Use unadjusted data when you need the raw balance or want to study the seasonal pattern itself. Keep your choice the same across the chart, and save the series code.