Summary

  • RIPE NCC reported 1,253,632 IPv4 addresses transferred in August. That exactly matches POLICY-coded rows in four public tables; another 285,184 addresses appear under MERGER_OR_ACQUISITION.
  • August’s four-table total was 408 rows and 1,538,816 addresses, versus 405 rows and 5,984,896 in July. Rows are records, not transactions.
  • The four largest exact public from/to label pairs fell from 89.3% to 38.7% of POLICY volume. This is record composition, not market concentration.

Analysis

RIPE NCC’s September Member Update reports 1,253,632 addresses transferred in August, 4,681,984 fewer than in July, but does not call the number POLICY. It exactly matches the POLICY rows in a 9 October 16:00 UTC snapshot of the four public IPv4 tables: 335 POLICY rows and 73 MERGER_OR_ACQUISITION rows. Together, the tables contain 408 rows and 1,538,816 addresses; July’s total was 405 rows and 5,984,896 addresses. The gap is first a classification question, not an economic explanation.

Rows rose by three while all-type address volume fell 4,446,080, or 74.3%. POLICY volume fell 4,681,984; the M&A-labelled subtotal rose 235,904 from a much smaller July base. The four largest exact combinations of public from/to labels accounted for 89.3% of July POLICY addresses and 38.7% of August’s. A tie at fourth place does not change the August sum. These repeated strings are not verified counterparties, owners or market participants.

A row need not equal one economic transaction. Address volume does not measure price, demand, liquidity or concentration. The reviewed sources give no cause for the monthly change and establish no shift in routing, utilisation or customer movement. One month is not a trend. The Member Update offers a concise statistic; the tables expose records across allocations, assignments and inter-RIR flows. Showing the all-table total beside both classification subtotals would make the denominator visible. The tables can be revised, so any comparison needs its retrieval time.

Sources