Systematic bias in economic consensus

Is a given economic data series consistently overestimated or underestimated?

Macro, FX and crypto assets MAC-003 Bar Sample

Download CSV Sample JSON Specification JSON Sample as of 2026-10-16

0700.HK Tencent Holdings | Hang Seng Index | 2025-10-16 – 2026-10-16 | HKD

Highest item: Consistent underestimate: 0.6
Lowest item: Consistent overestimate: -0.48
Average: 0.23
Number of items: 5 items

Main chart: mean forecast error / mean absolute error (-1 to 1), by economic indicator on the same basis

Mean forecast error / mean absolute error (-1 to 1) (Bar)

Supporting chart: overestimate count, underestimate count and sample size

Overestimate count, underestimate count and sample size (Line)

Detail: data rows, observation dates and missing-value status for the systematic bias in economic consensus

Mean forecast error / mean absolute error (-1 to 1) and Overestimate count, underestimate count and sample size (Table)

Sample data rows

Economic indicators on the same basis	Mean forecast error / mean absolute error (-1 to 1)	Overestimate count, underestimate count and sample size	Data date (sample)
Consistent overestimate	-0.48	96.8	2026-10-16
Slight overestimate	0.35	98.2	2026-10-16
No clear bias	0.55	100	2026-10-16
Slight underestimate	0.11	79.9	2026-10-16
Consistent underestimate	0.6	88.4	2026-10-16

Source and limits

Research question

Is a given economic data series consistently overestimated or underestimated?

Axes and units

Horizontal axis: Economic indicators on the same basis
Vertical axis: Mean forecast error / mean absolute error (-1 to 1)
Metric: Mean forecast error / mean absolute error (-1 to 1)
Comparison metric: Overestimate count, underestimate count and sample size

Calculation method

For each series with the same country+type+comparison, calculate e=actual-estimate; quantify the direction of bias as mean(e)/mean(abs(e)), leaving blank where the denominator is zero.

Limitations

The indicator basis and forecast version must be kept consistent. A bias does not mean it can continue to be exploited in the future. Assets must be aligned by currency, trading day and time zone. Revised macro figures do not represent the data known at the time.

The plotted values are generated sample data and do not result from this calculation.

Data API requirements

ECON

Data class: Derived
Complexity: High
Update cadence: Follows the release frequency of the source data. Market series updated daily.
Needs user-supplied input: No
Needs extra data plan: No

API names are the data groups the original specification requires, listed as methodology metadata only. This page makes no live data connection and is not connected to the live market data shown elsewhere on the site.

Source documentation

eodhd.com/financial-apis/economic-events-data-api

Design priority

Catalogue priority 828 of 1,000 (stage P4), a planning estimate from the original workbook.