Summary

  • ARIN’s annual report lists 78,234 IPv4 /24s in its 2025 inter-RIR flow table. Of the 34,689 /24s listed as leaving the ARIN region, 27,804 went to LACNIC; of the 43,545 listed as entering, 39,587 came from RIPE NCC.
  • These are directional volumes, not a record of prices, motives, unmet demand or individual buyers and sellers. ARIN also distinguishes completed transfers, request counts and processed tickets; those units cannot be collapsed into one “market” number.

Analysis

A regional transfer total can suggest a one-way current. ARIN’s own 2025 table shows why the direction matters. It lists IPv4 space moving both out of and into the ARIN region, and the two sides have different leading counterparts. LACNIC is the main listed destination for outward volume. RIPE NCC is the main listed source for inward volume. The total does not describe one uniform flow.

The six published /24 values are:

Direction IPv4 space
ARIN region → LACNIC 27,804 /24s
LACNIC → ARIN region 584 /24s
ARIN region → RIPE NCC 4,352 /24s
RIPE NCC → ARIN region 39,587 /24s
ARIN region → APNIC 2,533 /24s
APNIC → ARIN region 3,374 /24s

Add the three outbound rows and the table gives 34,689 /24s. Add the three inbound rows and it gives 43,545. Together they reconcile exactly to the report’s 78,234 /24s. The directional difference is 8,856 /24s more listed as entering the ARIN region than leaving it during the report’s 2025 accounting period.

That difference is real arithmetic, but it is not a market balance sheet. It says nothing by itself about whether the resources were sold, leased, moved after a merger, or transferred for another reason. It does not identify the parties, the price, the negotiation, the operational purpose or whether the new registration was used in a live network. The report records registry movement; it does not expose the commercial transaction behind each record.

The pairwise split is more informative than the net. LACNIC accounts for about 80% of the listed outward volume: 27,804 of 34,689 /24s. RIPE NCC accounts for about 91% of listed inward volume: 39,587 of 43,545. APNIC appears on both sides at lower volumes, with 2,533 /24s moving from ARIN and 3,374 moving toward it. These are measured lanes in one ARIN report, not evidence of a preference, cause or durable regional pattern.

The units also need protection. A /24 is a size unit for IPv4 space, equal to 256 addresses; it is not a transfer request, an organization or a transaction. ARIN’s annual report separately lists 445 inter-RIR transfers. That count should not be divided into the 78,234 /24s to produce a typical transfer size. ARIN’s transfer-statistics guidance says a single request can involve multiple blocks, and the annual report’s 8.4 volume chart is specifically an IPv4 view. The count and the volume answer different questions.

ARIN’s own historical statistics distinguish another tempting category. A “processed” ticket is one that has been closed, regardless of the request’s final outcome. A completed transfer record, a processed ticket, a request count and an address-volume total are not interchangeable. A chart that merges these measures could make a queue’s activity look like completed movement, or make address volume look like the number of counterparties. Keeping the labels attached is basic evidence hygiene.

For operators, the table offers a bounded view of where registered IPv4 space crossed the ARIN boundary in 2025. For people assessing registry policy, it shows that directional data can be published without naming the parties or disclosing contractual terms. What remains absent is a transaction-level account of outcomes and reasons. The aggregate therefore supports a question about reporting design, not an answer about market demand: can the public record keep direction, resource size, transfer count and request status separate all the way from the annual summary to the underlying transfer log?

The most useful reading is neither “ARIN is a net importer” nor “space is leaving for one destination.” It is narrower. One year’s ARIN record shows a concentrated flow toward LACNIC and a separate, more concentrated flow from RIPE NCC, with a smaller two-way APNIC lane. The report does not explain those patterns, and no explanation should be supplied on its behalf.

Sources

ARIN 2025 Annual Report, page 8 · ARIN 2025 Statistics archive · ARIN specified transfer log and definitions