Data Note
A higher semiconductor share in export revenue may reflect relative prices rather than greater physical specialization. A two-product calculation separates value concentration, quantity exposure and the interpretation of a technology cycle.
The semiconductor share of Korean export revenue can rise without an equivalent increase in physical production or a reduction in other export activity. Relative prices change the numerator and denominator of a value-based share. Such a share is useful for assessing revenue exposure, but it does not by itself measure changes in industrial specialization at constant prices.
The Bank of Korea's analysis of a semiconductor-led terms-of-trade improvement explicitly distinguishes export-price effects and their income implications from production growth. This note isolates a narrower measurement issue: how a price cycle changes a product's share of gross export value. It does not reproduce the Bank's forecasts or estimate a current Korean concentration effect.
The higher current-value share in the second row is not meaningless. A larger proportion of export receipts is now sensitive to semiconductor prices and demand. That is relevant to revenue and external-income analysis. It is different from saying that an additional third of factories, workers or physical capacity has moved into semiconductor production.
A constant-price decomposition can help isolate quantity developments, but physical units cannot be added indiscriminately across dissimilar goods. A semiconductor device, a vehicle and a ship are not interchangeable units of production. An index with explicit price weights is needed if a common volume measure is constructed.
Unit values also require care. Average revenue per exported item can rise because the exported mix shifts toward more advanced or higher-priced products. That is not necessarily a price increase for an unchanged product. A clean price-volume decomposition needs sufficiently detailed product matching or appropriate quality adjustment.
A single semiconductor category can contain products serving different customers and production processes. Splitting that category changes a statistical concentration index even when the underlying transactions remain the same. Conversely, combining many unrelated non-semiconductor goods into one residual category conceals diversity within that group.
Comparisons across time or economies should therefore preserve the product classification and reporting currency, identify price effects, and distinguish gross export value from domestic value added. A rise in shipment value can affect local income differently when imported inputs and the ownership of production differ.
For analysis of Korea's semiconductor cycle, a defensible presentation places the current-value export share beside matched price and volume evidence. The share describes where receipts are concentrated at prevailing prices. Claims about physical specialization or a lasting structural shift require additional evidence about quantities, capacity and the composition of production.