Deflation Probability (STLPPMDEF) | FRED | St. Louis Fed
- Rows
- 438
- Columns
- 2
- Data range
- 1990-01-01 to 2026-06-01
Graph and download economic data for Deflation Probability (STLPPMDEF) from Jan 1990 to Jun 2026 about inflation and USA.
AI analysis
Dataset Analysis: Deflation Probability (STLPPMDEF)
1. What this dataset represents
This series, sourced directly from the St. Louis Fed's FRED API (fredgraph.csv?id=STLPPMDEF), tracks the model-implied probability of deflation over time — a derived/smoothed metric rather than a raw price index. Given the publisher (FRED, a highly authoritative and continuously updated macroeconomic data hub), this series is refreshed regularly and benefits from rigorous methodology. Notably, the time coverage extends to June 2026, indicating this includes forward-looking/forecast values rather than purely historical observations — an important distinction for any correlation study, since projected values carry different reliability than realized ones. As a probability metric tied to inflation dynamics, it has strong potential value for studies linking monetary policy expectations, healthcare pricing trends (per the stated topic), and broader deflationary/inflationary cycles.
2. Data quality observations
The dataset is clean on the surface: 0 null cells across both columns, and both observationdate and STLPPMDEF report zero missing values across all 438 rows. However, the duplicate-row check is explicitly flagged as pending ("Phase D — recomputed against Parquet"), meaning duplication has not yet been verified — this should be treated as an open risk rather than confirmed cleanliness. Additionally, observationdate has 438 distinct values matching row count (good — no key collisions), while STLPPMDEF has only 421 distinct values out of 438 rows, implying some repeated probability readings (roughly 17 rows share values with others), which is plausible for a bounded probability metric but worth noting as reduced cardinality rather than a defect.
3. Key columns and distributions
The STLPPMDEF column is heavily right-skewed (skewness = 5.3), with a mean (0.035) far exceeding the median (0.00202) — indicating that in most periods deflation probability is very low, punctuated by occasional sharp spikes (max = 0.8896). The tight interquartile range (Q1 = 0.000068, Q3 = 0.0168) confirms that the bulk of observations cluster near zero, while 81 outliers (~18.5% of rows) represent crisis-driven spikes (e.g., 2008-09, 2020 COVID shock, and potentially forecast uncertainty near 2026). This distribution shape is typical of tail-risk/probability series and suggests that raw correlation analysis should consider log-transformation or rank-based (Spearman) methods rather than Pearson correlation, given the extreme skew.
4. Recommended join key
observationdate is the clear and only viable join key — it's a clean, fully-populated, monthly-frequency date field with no nulls and no duplicates. For cross-dataset correlation, align other series to month-start dates (2024-01-01 format) and consider whether target datasets are monthly, quarterly, or daily to determine necessary resampling/interpolation.
5. Suggested pairings for correlation discovery
Strong candidates include: (a) core CPI/PCE inflation series (e.g., CPIAUCSL, PCEPI) to contrast realized inflation against deflation probability; (b) Federal Funds Rate (FEDFUNDS) and Treasury yield curves to explore monetary policy responses; (c) healthcare-specific price indices (e.g., CPI Medical Care component) given the dataset's stated topic tag; (d) unemployment rate and industrial production for business-cycle correlation; and (e) breakeven inflation rates (T5YIE, T10YIE) as a market-based comparison to this model-based deflation probability. Given the forecast horizon extending to 2026, pairing with other forward-looking/expectation-based series (survey data, forward rates) would also be particularly insightful.
Columns
- observation_date (date)
- STLPPMDEF (decimal)