3-Month Treasury Bill Secondary Market Rate (TB3MS)
- Rows
- 1,114
- Columns
- 2
- Data range
- 1934-01-01 to 2026-10-01
- Auto-refresh
- Re-checks its source every 14 days
- Last refreshed with new data
- 2026-11-06
Monthly 3-month Treasury bill rate from FRED.
AI analysis
Analysis: 3-Month Treasury Bill Secondary Market Rate (TB3MS)
1. Dataset Overview & Value for Correlation Studies This dataset captures the 3-Month Treasury Bill Secondary Market Rate sourced directly from the Federal Reserve Bank of St. Louis (FRED) via their public API at https://fred.stlouisfed.org/graph/fredgraph.csv?id=TB3MS. FRED is one of the most authoritative macroeconomic data publishers in the world, lending this dataset high credibility and institutional reliability. With monthly observations spanning January 1934 through May 2026 — nearly 92 years of data — this series offers exceptional longitudinal depth, capturing multiple full economic cycles including the Great Depression aftermath, post-WWII expansion, the stagflation era, the Volcker rate shock, the 2008 financial crisis, the near-zero rate environment of the 2010s, and the aggressive 2022–2023 rate hiking cycle. The freshness implied by the 2026 endpoint suggests this feed is actively maintained and near real-time, making it highly valuable for both historical backtesting and current macroeconomic correlation research.
2. Data Quality Observations Data quality here is essentially pristine. Total null cells across both columns register at exactly 0, meaning there are no missing observations across the full 1,109-row history — a rare and valuable characteristic for time-series analysis. The observationdate column shows 1,109 distinct values matching the 1,109 total rows, confirming zero duplicate dates and a perfectly regular monthly index with no repeated records. The TB3MS decimal column shows 556 distinct values across 1,109 rows, which is entirely expected for an interest rate series — rate levels repeat across time (e.g., multiple months at near-zero rates post-2008 and post-2020). No type mismatches are flagged. The 19 statistical outliers noted in the distribution are worth investigating but almost certainly correspond to the historically extreme Volcker-era rates of the early 1980s, rather than data errors. This dataset requires no imputation or deduplication and is ready for immediate analysis.
3. Key Column Distributions & Statistical Characteristics The TB3MS rate column tells a rich macroeconomic story through its statistics. The mean of 3.42% sits notably above the median of 2.96%, and with a positive skew of 1.06, the distribution is right-tailed — pulled upward by the extreme high-rate periods of the 1970s–80s. The maximum of 16.30% (almost certainly from 1981 during the Volcker disinflation campaign) and minimum of 0.01% (reflecting the post-2008 and post-2020 zero-bound periods) illustrate a massive operating range of over 1,600 basis points. The interquartile range — Q1 at 0.38% to Q3 at 5.21% — reveals that fully 50% of monthly observations fall in a relatively compressed low-rate band, meaning low-rate environments have been disproportionately common in this dataset's history. The standard deviation of 3.09% is large relative to the median, further confirming high volatility across regimes. These distributional properties make this series ideal for regime-based segmentation analysis.
4. Recommended Join Key Columns The observationdate column is the clear and sole join key, formatted as a standardized monthly timestamp (YYYY-MM-01T00:00:00). Its perfect distinctness (1,109 unique values = 1,109 rows) and zero nulls make it a reliable anchor for time-series joins with any other monthly macroeconomic dataset. When joining, analysts should be mindful of the series' start date (1934) — many modern economic datasets begin later (e.g., 1960s or 1980s), so inner joins will naturally truncate the usable history. For datasets with different frequencies (e.g., quarterly GDP), a date-granularity alignment step (e.g., mapping to quarter-end months) will be necessary before joining.
5. Suggested Datasets for Correlation Analysis This series pairs powerfully with a wide range of macroeconomic and financial datasets. Inflation data (CPI or PCE from FRED) is the most natural pairing — the relationship between short-term rates and inflation is the backbone of monetary policy analysis. Federal Funds Rate (FEDFUNDS) would show near-perfect correlation and help distinguish policy-driven versus market-driven rate movements. 10-Year Treasury Yield (DGS10) enables yield curve spread analysis (the 10Y–3M spread is a canonical recession predictor). GDP growth (A191RL1Q225SBEA) would allow study of rate-growth relationships across cycles. For equity markets, the S&P 500 index or earnings yield would support classic risk-free rate vs. equity premium analysis. Finally, unemployment rate (UNRATE) and mortgage rates (MORTGAGE30US) would allow exploration of the real-economy transmission channels of monetary policy — all available from the same FRED publisher, simplifying licensing and join alignment.
Columns
- observation_date (date)
- TB3MS (decimal)