Why I am not buying IPv4 addresses as an investment

samuelshadrach

2026-01-17

Disclaimer

Quick Note
I did the actual research back in 2024-10. I polished the notes and published them online in 2026-01. It's possible the situation has changed in this one year.

Summary

I noticed IPv4 address prices were going up. I considered purchasing IPv4 addresses in bulk as a investment, in order to make money. But I decided not to buy them.
Many ISPs are adopting dual stack (IPv4 and IPv6) and CGNAT, in order to server their users.
- Dual stack means they serve users over IPv6 (typically as first preference), but also support IPv4 (typically as second preference).
- CGNAT means they can serve multiple users over a single IP addresses, by assigning different ports to them. Most users are fine with CGNAT.
- ISPs cannot unilaterally shift to IPv6-only, because users don't want this, because users want to access all websites, and some websites are still hosted as IPv4-only.
Some websites owners (especially large enterprises) to switch their websites to either dual stack or IPv6-only, for technical reasons. They would rather pay the increasing prices of IPv4 addresses.
Maybe CGNAT means IPv4 address space won't exhaust.
- There are ~4.3 trillion IPv4 addresses. There are atleast ~20 billion website URLs. There are ~8 billion people on Earth.
- If ISPs all quickly shift to CGNAT, it is possible the current IPv4 address space is enough to support all website owners by a comfortable margin.
- Since I have high uncertainty on a) whether the current IPv4 address space is sufficient for most website owners, and b) whether most ISPs will find it easier to adopt CGNAT than force IPv6-only, I have high uncertainty on whether IPv4 address prices will go up or not.

Main

Why might you expect IPv4 addresses prices to go?

ipv4 prices have historically gone up due to ipv4 address exhaustion.
There are people are hoarding and leasing existing IPv4 addresses for this very reason.
All the internet standards bodies are demanding the world switch to IPv6. But adoption is lagging. This means there are some bottlenecks preventing people from switching immediately.

Compatibility between IPv4 and IPv6

both websites and ISPs (residential and mobile) need to enable ipv6 for the transition to work
IPv4 and IPv6 are not interoperable, an IPv4 cannot communicate with an IPv6
IPv4 requires NAT, which is not the best way of doing security. With IPv4 and NAT, often a static IP list is provided to routers, so the server that manages the firewall can go offline.
With IPv6, it is common (or maybe required?) to keep the server that manages the firewall online, so it provides the IP list in realtime to the routers. Usually they provide a deny-all rule but allow some services as needed.

Market players

End users - want to browse websites and get access to various services.
ISPs - want to sell whatever services the end user wants, as their preferred mode of accessing internet. Manages network infra (fiber optic cables, routers, etc)
Website owners - primarily large enterprises and small companies, who have financial reason to want to maintain websites. Also some individuals who have arbitrary motivations for maintaining websites. Most of them rent servers from cloud providers. A few large enterprises manage their own servers.
Cloud providers - Primarily an oligopoly of AWS, GCP, Azure, who want to sell whatever services the website owners want. Manages compute and storage infra, and network infra. Also has transit agreements with various ISPs.

Current adoption

End users
- Almost all new consumer devices (mobiles and PCs) support accessing both IPv4 and IPv6. No manual switching required.
- Some very old consumer devices may not support IPv6.
ISPs
- ipv4-only ISPs - constant adoption
- dual stack ISPs - increasing adoption
- ipv6-only ISPs - almost non-existent
Cloud providers
- Big 3 cloud providers provide all three options - ipv4-only, dual stack, ipv6-only
- Cloud providers are basically an oligopoly.
- Some smaller cloud providers provide IPv4-only (if they're old) or IPv6-only and dual stack (if they're new).
Website owners
- ipv4-only websites - constant adoption
- dual stack websites - increasing adoption
- ipv6-only websites - non-existent

More details on website owners using dual stack

Many sysadmins and network security people at enterprises dont know how to do IPv6 config yet. Hence they are still hosting IPv4-only websites, either on cloud providers' servers or on the enterprise's own servers.

Multiple ways ISPs deal with increasing number of users.

They can purchase IPv4 addresses (increasing their price), and pass the cost on to the user.
- Some ISPs are doing this.
They can adopt CGNAT, and serve more than one user using a single IPv4 address.
- Some ISPs are doing this.
They can provide IPv6-only to users.
- In practice, this is not an option as many websites still don't provide IPv6. Users don't want to purchase an internet connection that lets them only access some websites but not others.

More details on ISPs using CGNAT or dual stack

Some ISPs now use carrier-grade NAT, where they can assign different port numbers on the same IP to different customers. This allows them to serve more users on IPv4, without having to purchase more IPv4 addresses.
Most users don't care if they're being served over CGNAT. Most users just need a few ports open, they don't need the full range of ports associated with an IP address.
A few power users care, for example those hosting web servers on residential IP addresses. Webhosting requires unique IP, can't use CGNAT (but it can use its own NAT). Some devs are using VPNs and tunnels to host their websites on residential ISPs despite their residential ISP using CGNAT. reddit comment on this
Indian residential ISPs seem mostly to use CGNAT.
India is leading on IPv6 (dual stack) adoption. Indian mobile ISPs are increasingly transitioning from IPv4-only to dual stack. Most users are are mobile-first and new to getting internet access. Some Indian ISPs themselves provide mobile services like streaming platforms, and these are increasingly served over IPv6.

My question: Who is buying more IPv4 addresses (and driving the price increase)? Is it website owners or ISPs?

My first guess is ISPs, as consumer devices (mobile in particular) are more in number than websites. ISPs also heavily relying on CGNAT
Assuming CGNAT is sufficient to host 8B users on mobile, next question is whether residential ISP connections also use CGNAT
Assuming residential ISP connections also use CGNAT, it is possible we never (?) need ipv6 as all webservers and all clients can be serviced on 4B * 1000 IP addresses (??)
Conclusion
- not obvious whether CGNAT will be sufficient to bridge demand for ISPs
- not obvious why ISPs prefer to use CGNAT over just use IPv6. Both are software changes (not obvious there is any hardware change)
- not obvious whether IP adddress prices will go further up or not

Maybe CGNAT means IPv4 address space won't exhaust?

There are 2^32 = ~4.3 trillion IPv4 addresses. There are ~8 billion people on Earth. There are atleast ~20 billion unique website URLs as per Commoncrawl.
Most of the IPv4 addresses are currently used by end users, not by website owners.
If ISPs all quickly shift to CGNAT, it is possible the current IPv4 address space is enough to support all website owners by a comfortable margin.
Since I have high uncertainty on a) whether the current IPv4 address space is sufficient for most website owners, and b) whether most ISPs will find it easier to adopt CGNAT than force IPv6-only, I have high uncertainty on whether IPv4 address prices will go up or not.

Side Note: Leasing IPs

IP addresses are purchased from RIR in blocks. Individual static IPs can be leased from anyone including ISPs and cloud server hosts
Leased IP addresses often have a poor IP reputation on various IP blacklists, which can be scrubbed over time.
This means most leased IP addresses are not a liquid fungible asset.
AWS currently leases static IPs at $0.005 / hour = ~$40 / year. Since AWS is reputed, their leased IPs are presumed less likely to end up on IP blacklists, because they can proactively shut down people using their servers for activites likely to get blacklists.
Leased IP addresses are currently not a significant portion of the market.

Misc

openwrt is good router software supporting ipv6 (as per some hackernews comment)
IPv6 uses SLAAC to assign addresses without requiring router to do it. SLAAC tutorial?

[-]npostavs2mo43

There are ~4.3 trillion IPv4 addresses

[...]

Assuming residential ISP connections also use CGNAT, it is possible we never (?) need ipv6 as all webservers and all clients can be serviced on 4B * 1000 IP addresses (??)

There are 2^32 = ~4.3 trillion IPv4 addresses.

2^32 = ~4.3 billion, not trillion.

[-]Dagon2mo20

It's a bad investment for the same reason a lot of small-scale investing in commodities which don't have strong market infrastructure is - the overhead of transactions strongly outweighs the risk/reward. Anyone who's going to pay you (as opposed to paying their ISP or aggregator) for a small block of IP addresses will instead buy/hoard their own.

It's also subject to a fair bit of elasticity - when it gets expensive, there are technological options which are less ideal, but well worth it when the cost savings justifies it. cf. fracking for oil. IPv6 and CGNAT are two options you mention, and others could be invented if needed.

I'd argue that IPv6 is already well-established enough in the roaming/mobile/cell world that it'll put a cap on IPv4 prices. The expense/hassle of switching to IPv6 is real, but not worth $hundreds to most people, or $thousands to most small businesses.