WSTS World Semiconductor Trade Statistics — Historical Billings & Forecasts

World Semiconductor Trade Statistics (WSTS) is the semiconductor industry's own statistics program: an organization whose member companies report actual billings, which are compiled into the consensus numbers the whole industry quotes when it says how big the market is. The defining property is provenance - these are aggregated reported revenues, not modeled estimates, so one methodology stretches across four decades.

The flagship deliverable is the Historical Billings Report, described at the source as "four decades of WSTS statistics from the WSTS Blue Book" and spanning 1986 to date. Its current edition runs to June 2026, when worldwide monthly billings reached roughly $152 billion. Alongside the history sit two forward-looking releases a year - the Spring and Fall Semiconductor Market Forecasts, produced around WSTS's Forecast Conferences (the 73rd is set for November 16-19, 2026 in Kobe, Japan) - projecting billings several years ahead.

Where it sits in the catalog is mapped on our Semiconductor Materials & Equipment industry hub.

Illustrative rows in the delivered shape - one region, one period, one measure:

region         period         measure                     unit      value
Worldwide      June 2026      monthly billings            1000 US$  151934915
Asia Pacific   June 2026      monthly billings            1000 US$   84848976
Americas       June 2026      monthly billings            1000 US$   53250752
Worldwide      June 2026      three-month moving average  1000 US$  134447227
Worldwide      January 1986   monthly billings            1000 US$    1645914

Two things in that block carry the analytical weight. First, the fixed regional layout - Americas, Europe, Japan, Asia Pacific, Worldwide - repeats identically under every year header back to 1986, which is what makes the file a panel rather than forty separate tables. Second, the units are constant: everything is stated in thousands of US dollars, so a 1986 worldwide month (~$1.65 billion) and a June 2026 month (~$152 billion) read off the same scale without conversion. That roughly ninety-fold expansion is the industry's growth curve in two rows. Live values for your named windows arrive with the sample.

Ten verified fields form the core dictionary: the year-block structure, twelve month columns plus Q1-Q4 subtotals and a Total Year column, the five regional rows, and the smoothed companion series. All ten carry definitions confirmed against the published workbook layout rather than guessed from headers.

The wider program extends beyond these ten: product-category breakdowns - Discretes, Optoelectronics, Sensors, Analog, Micro, Logic and Memory - and end-use detail fold under additional fields on request, pinned down against a delivered extract before anything depends on them.

Geography - five fixed cuts that have held steady across the whole series: Americas, Europe, Japan, Asia Pacific, and a worldwide total. The stability matters more than the count - a regional ratio computed in 1990 computes identically today. Asia Pacific now carries the largest single share of the world market; in January 1986 it was the smallest of the four regions.

Temporal - 41 years deep, 1986 through the latest reported month (June 2026 in the current edition), at both monthly and quarterly resolution. Every series also ships as a three-month moving average, the form analysts quote when the question is direction rather than timing. Forecast releases extend several years beyond the present.

Granularity - regional totals, monthly to annual. What the public files do not carry is the product dimension: the device-family split lives in the members-only reports. For sizing the equipment and materials demand behind those billings, pair this with plant-level sources such as the Wikipedia list of semiconductor fabrication plants.

API, files, or your warehouse. Daily, weekly, or hourly.

You pick the channel and the cadence; extraction, normalization and schema stability are our problem. The conventions above - thousands-of-dollars units, the five-row regional layout, the 3MMA companion - travel fixed across all three channels, so a nightly warehouse load and an on-demand lookup agree row for row. Each new monthly edition rotates into place without reshaping your tables, and the twice-a-year forecast releases land as their own dated vintage rather than overwriting history. A sample scoped to your regions and date window ships first either way, sized to test in your own pipelines the same day.

• Market sizing and forecasting - the worldwide and regional totals are the denominator most semiconductor market models start from; the Spring and Fall projections give the forward half of the same baseline.
• Equipment and materials demand modeling - fab operators' revenue converts into demand for wafer-processing tools, consumables and process chemicals; analysts use regional billings as the leading indicator for equipment orders in each geography.
• Investor fundamentals and quants - the monthly cadence makes billings momentum a tradable alternative signal for the semiconductor complex, with the 3MMA smoothing out the noise that raw months carry.
• Competitive intelligence - tracking which region is expanding fastest tells supply-chain teams where capacity investment is actually landing; see competitive intel product teams use cases.
• Corporate strategy and site selection - multi-decade regional history supports long-horizon questions about where the industry concentrates; see investors quants use cases.

Within semiconductor market data, WSTS owns the consensus regional billing series and concedes on product depth. The SIA Semiconductor Market Data & Factbook resells the same underlying WSTS program through a subscription dashboard with 200+ reporting categories by product family and end use back to 1976 - more axes, gated delivery. The OEC HS 8542 trade profile counts integrated circuits crossing borders, which measures trade flows rather than producer billings. UCI SECOM serves an entirely different job - sensor readings from a manufacturing line for yield-modeling work.

The usual pattern is to stack them: WSTS sets the authoritative market-size baseline, product-level dashboards add the category split, and trade or fab-level datasets localize it. This record ranks third on our best semiconductor materials & equipment datasets shortlist.

Cards worth reading next: the rest of the Semiconductor Materials & Equipment shelf - patent corpora, fab-location lists and materials-discovery databases that put physical assets next to these dollar figures - plus the glossary entries that unpack the program's vocabulary (the Historical Billings Report itself, how the 3MMA works, what the forecasts promise), and the head-to-head against the fab list for teams deciding between market dollars and plant assets.