Screen 3

Mirror Reconciliation

Indonesia's reported exports of HS 4402 set against what each destination independently reports importing from Indonesia. Two reporters, two submissions, two extraction runs — reconciled here.

Read the weight column, not the value column

Exports are reported FOB. Imports are reported CIF — freight and insurance included. A raw value comparison therefore shows a gap of roughly 5–12% even when nothing at all is wrong. Net weight in kilograms carries no such artefact, so the weight gap is the defensible signal and the value gap is shown only as context. In 2019 this is visible directly: Germany, Japan, Singapore reconcile on weight to within 5% while their value gaps run +30%, +21%, +22%. That difference is freight, not a discrepancy.

Pairs reconciled

0 awaiting a destination filing

Unreconciled weight

17,684,899 kg, absolute, 2019

Share of reported exports

of 339.0 kt reconciled

Persistent findings

≥3 consecutive years, same direction

Persistent findings

A one-year gap is noise. These are not one-year gaps.

Netherlands — 7 consecutive years, Indonesia consistently reports sending more than the destination reports receiving, peak +76%.

India — 4 consecutive years only 3 on filed weight, destination consistently reports receiving more than Indonesia reports sending, peak +1536%.

1 of those years uses a net weight Comtrade derived rather than received — state this before anyone finds it.

Malaysia — 3 consecutive years, Indonesia consistently reports sending more than the destination reports receiving, peak +44%.

Türkiye — 3 consecutive years, Indonesia consistently reports sending more than the destination reports receiving, peak +43%.

Germany — 6 consecutive years only 2 on filed weight, destination consistently reports receiving more than Indonesia reports sending, peak +307%.

4 of those years use a net weight Comtrade derived rather than received — state this before anyone finds it.

Persistence is scored as consecutive annual periods where the weight gap exceeds 15% and the direction does not change. A missing year breaks a streak rather than bridging it. Ordered by years standing on filed weight, not by streak length — a long streak resting on estimates is the weaker claim, and sorting by length alone would put it at the top.

Reconciliation · 2019

Destination	Direction	Persistence
NetherlandsNLD	over-reported at origin
MalaysiaMYS	over-reported at origin	·
ChinaCHN	under-reported at origin	·
Saudi ArabiaSAU	under-reported at origin	·
GreeceGRC	over-reported at origin	·
GermanyDEU	under-reported at origin	·
JapanJPN	under-reported at origin	·
Rep. of KoreaKOR	under-reported at origin	·
SingaporeSGP	under-reported at origin	·
IndiaIND	over-reported at origin	·
United Arab EmiratesARE	under-reported at origin	·
TürkiyeTUR	over-reported at origin	·

Sorted by absolute weight gap. Expand any row (+) for the full period-by-period series with both reported lines overlaid. Every figure opens the provenance panel showing both contributing extraction runs — the origin filing and the destination filing are separate collections and are shown as such. Rows marked not reported are cases where one reporter has published and the other has not; an inner join would have hidden them.