United States Unit Labour Costs QoQ

United States · Quarterly series · EC-US-ULCQQ

Latest, Q3 2026

-0.7%

-2.55pp from Q2 2026

Open interactive chartUnited States Unit Labour Costs QoQ in the app

16.711.2-14.31Q4 1996Q3 2027
History through Q3 2026, the forecast path to Q3 2027 (dashed) and its 80% interval (shaded).

Forecast

Forecast for Q3 2027: 2.2%, with an 80% interval of -5.4% to 9.7%.

Forecast path. Intervals are shown where available.
PeriodForecast80% low80% high
Q4 20262.2%-2.2%6.5%
Q1 20272.2%-5.2%9.5%
Q2 20272.2%-2.9%7.2%
Q3 20272.2%-5.4%9.7%

Model: Exponential smoothing (statistical). Forecast generated 2026-11-15.

AI confidence

Needs refresh

The data refresh is due. Confidence will be reassessed against the refreshed history.

Model: Exponential smoothing (statistical)

Accuracy vs naive
n/aNot evaluated
Direction
n/aNot evaluated
Turning points
n/aNot evaluated
How this assessment is made

AI confidence describes historical evidence for this forecast, not a probability of being correct. Each backtest fits on data available before its test window. The baseline carries the last value forward. Supported evidence requires at least six complete, non-overlapping tests, an error ratio below 0.90 (rounded to two decimals), and wins in a majority of tests. Ratios from 0.90 to below 1.10 are on par; 1.10 or above is weak evidence. Inconsistent wins produce mixed evidence. For a selected model, only the recent tests excluded from model selection determine this label. These are historical simulations using the available data, not a live trading track record; revisions and foundation-model pretraining can limit how closely they reproduce a past decision. Past performance can change.

Model comparison

Previous comparisonThe stored comparison uses earlier data or a previous model roster. An eligible nightly batch can compare both candidates, subject to call budgets and retry limits.

Models tested

Exponential smoothing scored lowest on the 2 older tests (0.78) and was chosen. The 6 recent tests played no part in the choice and give its grade (0.56). That grade beats the naive forecast. Scores are MASE: a model's errors divided by the naive forecast's, so 1 is a tie.

4 model classes completed the tests: Statistical, Foundation model, Machine learning, Neural network. Previous model roster.

Every model tested on this series.
ModelClass and computeOlder testsRecent testsRole
Exponential smoothing (statistical)Statistical · CPU0.780.56Serving
State-space model, 167M (foundation model)Foundation model · CPU0.780.54Tested
Agentic Investor TSFM (foundation model)Foundation model · CPU0.790.58Tested
Gradient-boosted trees (machine learning)Machine learning · CPU0.890.81Tested
Hierarchical neural network (neural network)Neural network · CPU0.930.64Tested
Naive forecast (last value held)Benchmark · CPU1.001.00Benchmark

Uses ci-proto, released by Continuous Intelligence. Continuous Intelligence does not endorse or approve this product or service. (ci-proto is shown as State-space model, 167M.)

Reading

It is above 15% of the 39 earlier readings since Q4 2016.

Written from the figures by a fixed template; no AI model is involved.

Recent history (24 observations)
The 24 most recent observations, newest first.
PeriodValue
Q3 2026-0.7%
Q2 20261.9%
Q1 20263.7%
Q4 20254.7%
Q3 20250.1%
Q2 2025-1.1%
Q1 20255.2%
Q4 20241.8%
Q3 20241.0%
Q2 20240.8%
Q1 20244.3%
Q4 20230.8%
Q3 20231.4%
Q2 20233.2%
Q1 20232.3%
Q4 2022-1.4%
Q3 20227.2%
Q2 20223.2%
Q1 20227.8%
Q4 20213.0%
Q3 20218.2%
Q2 20215.6%
Q1 2021-3.2%
Q4 202014.2%

Source

Data: public market data feeds.

More from Agentic Investor

Also stored for United States