US consumer price index, all items · United States · 1 month ahead
2026-08: the model assigns an 80 percent probability to 333 to 334.3 index 1982-1984=100
Point estimate 333.5 index 1982-1984=100 from Ridge regression on lags and drivers, against 332.8 index 1982-1984=100 observed in 2026-07. ridge beats last_value on out-of-sample MAE by 47.4 percent over 46 cutoffs (minimum 24, margin 5 percent).
Fan chart
- Quantiles
- p5 332.8 · p10 333 · p25 333.3 · p50 333.6 · p75 333.9 · p90 334.3 · p95 334.5
- Source
- empirical quantiles of 35 in-window 1-step residuals, in log space; no normality assumed
Immutable fields
- Key
- us-cpi-all-items:USA:1
- Id
- 2a386143-72a1-4d01-9a96-dfaab7b2174e
- Created
- 2026-09-07 22:16 UTC
- Data cutoff
- 2026-09-07 22:16 UTC
- Origin
- 2026-07 = 332.8 (obs 167372)
- Target period
- 2026-08 (31 days)
- Point
- 333.5 index 1982-1984=100
- Interval
- 333 to 334.3 at 80 percent
- Model
- ridge models_v1.targets_v1
- Baseline
- last_value
- Snapshot
- 4b5c36705b7715292aa6af9242bdfa07f4c56a454a828c2261a7372221d7351c
- Data mode
- ingested_at
- Status
- active
Statistical association, not causal
Ridge coefficient times the standardised input at the origin, in log index 1982-1984=100. These are associations learned in the training window, labelled as such; nothing here is a causal effect.
| Feature | Input | Coef | Contribution |
|---|---|---|---|
| us-cpi-energy[t-3] | 326 | -0.0001 | -0.0004 |
| us-cpi-energy[t-1] | 319.3 | -0.0001 | -0.0003 |
| us-ppi-all-commodities[t-3] | 282.8 | -0.0001 | -0.0002 |
| us-ppi-all-commodities[t-1] | 286.3 | -0.0000 | -0.0001 |
| y[t-0] | 5.81 | 0.0001 | +0.0001 |
| us-nonfarm-payrolls[t-1] | 158,892 | 0.0000 | +0.0001 |
| y[t-1] | 5.81 | 0.0000 | +0.0001 |
| us-nonfarm-payrolls[t-3] | 158,798 | 0.0000 | +0.0001 |
| y[t-2] | 5.81 | 0.0000 | +0.0001 |
| y[t-11] | 5.78 | 0.0000 | +0.0000 |
Not yet realised
The observation for 2026-08 has not been ingested. The weekly forecast.evaluate job attaches it when it arrives; the record above does not change.
What the forecast takes for granted
- No structural break in us-cpi between the data cutoff and the target period.
- The source keeps publishing the series on the same definition; a rebasing or redefinition invalidates the comparison.
- The lagged statistical associations estimated in the training window hold over the horizon.
- Driver series used, lagged: us-cpi-energy, us-ppi-all-commodities, us-nonfarm-payrolls. Their own future values are not forecast; only values known at the cutoff enter.
What would show this forecast wrong
- If the first published value of us-cpi (USA) for 2026-08 is below 333 index 1982-1984=100 or above 334.3 index 1982-1984=100, the 80 percent interval is falsified.
- The model assigns a 10 percent probability to a value below 333 index 1982-1984=100 and a 10 percent probability to a value above 334.3 index 1982-1984=100; interval misses should occur about one time in five over many forecasts, and a run of misses well above that rate falsifies the calibration.
- The point forecast is above the last observed value (332.8 index 1982-1984=100); a realised value at or below that level falsifies the direction call.
Scorecard for this target and horizon
| Model | Status | Cutoffs | MAE | RMSE | Direction | 80% coverage | Skill vs last value |
|---|---|---|---|---|---|---|---|
| Ridge regression on lags and drivers models_v1.targets_v1 | champion | 46 | 0.4594 | 0.6532 | 96% | 85% | 47.4% |
| Drift models_v1.targets_v1 | challenger | 46 | 0.4707 | 0.652 | 96% | 83% | 46.1% |
| Holt (exponential smoothing with additive trend) models_v1.targets_v1 | challenger | 46 | 0.516 | 0.7626 | 96% | 91% | 40.9% |
| Last value models_v1.targets_v1 | baseline | 46 | 0.8732 | 1.03 | 0% | 89% | 0.0% |
| Linear trend models_v1.targets_v1 | challenger | 46 | 3.32 | 3.97 | 80% | 67% | -280.5% |
| Seasonal naive models_v1.targets_v1 | challenger | 46 | 11.18 | 11.91 | 4% | 96% | -1180.9% |
Ridge regression on lags and drivers
Ridge regression on lags and drivers
Purpose. Point and 80 percent interval forecasts of us-cpi (USA) 1, 3, 6, 12 months ahead. Owner: Forecast Lab (HV 3.0).
Method
A direct multi-horizon ridge regression: the change from the origin to the target period is regressed on lagged target values (lags [1,2,3,12]) and lagged driver series (lags [1,3]), features standardised, penalty lambda 1. Drivers: us-cpi-energy, us-ppi-all-commodities, us-nonfarm-payrolls. Every driver enters lagged, so no feature uses a period after the origin. The series is modelled in natural logarithms and transformed back, so intervals are asymmetric in level terms.
Uncertainty
Intervals are empirical: the model is refitted at earlier origins inside the training window, the realised errors at the same horizon are collected, and the interval is the 10th to 90th percentile of those errors around the point forecast. No normality is assumed.
Training and evaluation
Training window: 954 periods, 1947-01-01 to 2026-07-01. Evaluation: rolling-origin backtest over 46 cutoffs using pseudo real-time data (the earliest ingested value per period; no vintage archive exists for this series, so revisions between the cutoff and today are not reproduced).
Known limitations
- Structural breaks after the cutoff are not modelled.
- Driver contributions are statistical associations, not causal effects.
- Promotion to champion requires beating the last-value baseline on out-of-sample MAE by at least 5 percent over the minimum number of cutoffs; otherwise the baseline stays champion.
- Retraining: every scheduled forecast.run refits on the data then available; backtests re-run from /admin/forecasts.
Supersedes
This is the first record for its key.