Start free trial

By Yair Knijn · October 7, 2025

You spent six figures leasing IPv4 you already owned but couldn't find

The request lands on the infrastructure director's desk as a budget line, not a question. Two teams say they are out of public address space, the next project needs a /22, and the broker quote is sitting in the inbox. The director signs it, because the assumption underneath is that the company is full. Nobody checked whether that assumption is true before the money left.

It usually isn't true. The space is there. It is just invisible, because the record that should show it is wrong, and a wrong record reads exactly like scarcity.

The 2025-2026 IPv4 transfer market: what addresses now cost

Pricing depends heavily on block size. Through 2025, /24 transfers traded in the rough range of $35 to $45 per address, while larger blocks like a /16 came in lower per-IP because buyers pay a premium for the small, routable pieces they can actually use. A single /22 bought at the top of that band is well into six figures before anyone has routed a packet.

Leasing looks cheaper on the invoice and is where a lot of orgs quietly end up. Market rates settled around $0.30 to $0.50 per IP per month across most RIPE and ARIN blocks, with APNIC running higher on tighter supply. That is recurring, forever, on space you are renting because you could not see the space you own. Reported lease utilization stayed above 80%, so this is not a thin market you can wait out.

Phantom scarcity: idle space hidden in stale records

Phantom scarcity is the gap between what your IPAM spreadsheet claims is allocated and what is actually answering on the wire. A /24 got assigned to a project that shipped two years ago and was decommissioned without anyone updating the record. A range marked "reserved — do not touch" protects a service that no longer exists. Half a /23 was carved for a lab that three engineers used in 2022.

None of that shows up as free, so it never enters the math when someone asks "do we have room." The director funding the purchase is reading a document that overstates utilization, and the document has no idea it is lying. In allocations I have audited, the idle fraction routinely lands around a quarter to a third of the public space on the books.

Reclaim-before-buy: the audit that pays for itself

Before any transfer request gets approved, the org should be able to answer one question per range: is anything real using this, and who owns it. That means reconciling the claimed allocation against what DHCP leases, DNS records, and active ARP and connection data actually show. An address with no lease, no resolving name, and no traffic for 90 days is a reclaim candidate, not an asset you must rebuy.

The arithmetic is blunt. If an audit surfaces a /23 of genuinely idle public space, you have avoided a transfer worth tens of thousands at current rates, or killed a lease that bills every month until someone notices.

Utilization reporting that makes the business case visible

A director will not block a purchase on a hunch. They will block it on a report that shows true utilization sitting at 68% with a named list of reclaimable ranges and the dollar value of buying versus reclaiming. The job is to turn "we feel full" into a number the finance conversation can use, refreshed continuously rather than rebuilt by hand every budget cycle.

Spot IPAM reconciles each Environment's claimed allocations against what DHCP and DNS actually serve, surfaces idle and overlapping space with an owner attached, and reports real utilization you can put in front of the person signing the check. Reclaim before you buy, and see the features that make the idle space visible first.