FinanceFinance in TelecomTelecom

Everyone sells towers to cut debt, then MTN paid $6.2bn to buy its own back

Telcos sold a record number of towers in 2026 to cut debt, while MTN agreed to pay $6.2bn to own its towers again. This playbook shows telecom CFOs how to price a leaseback, keep control of the network and decide when sharing beats selling.

October 2026 was a busy month for tower deals. BCE agreed to sell a 49.9 per cent stake in its wireless tower company to La Caisse for about $2.45 billion CAD, keeping majority control of more than 5,000 sites. The European Commission cleared the sale of VodafoneZiggo's passive tower assets to DigitalBridge, TD Greystone Infrastructure and L&G for €669 million. Ooredoo and Zain closed their first regional tower transaction, combining their tower assets in Qatar and Kuwait. Buyers still want more. In April, SBA Communications Corp. was exploring options including a potential sale after the tower owner and operator received preliminary takeover interest, and Vertical Bridge secured a $1.5 billion equity investment from KKR.

The most useful tower deal of the year went the other way. In February, MTN regained its old telecoms towers it initially sold to IHS Towers, at an enterprise value of about $6.2bn. That is the thesis of this piece. When you sell towers you are not selling steel. You are signing up to a long, inflation-indexed payment out of your own operating costs that is hard to get out of. It only creates value if the cash earns more than that payment costs and the contract still lets you reshape your network. Each step below follows from that.

How to monetize towers without hollowing out the operator

Step 1: Price the leaseback as a bond you are issuing

Under IFRS 16,the leaseback ends up on your balance sheet as a lease liability anyway, and rating agencies treat it that way too. MTN shows how big this gets: at the end of 2024, lease liability across MTN Group stood at $4.65 billion.

A hypothetical example, arithmetic only. You sell a portfolio for 1,000 and pay 60 in rent in year one. That is a 6% starting yield. With a 2.5% annual escalator, the rent in year 15 is about 84.8 (60 × 1.025^14). If your marginal cost of debt is 5%, you have swapped cheap debt for a dearer obligation that keeps growing, and you have given up the asset as well.

The buyer's target return gives you a second reference point. Writing about Bell and Telus, one analyst predicted the Caisse will earn a low double-digit return on its investments in the Bell and Telus infrastructure, with fee income from renting space on the towers rising over time in pace with inflation. Part of that return comes from future co-tenants, not from your rent. Even so, it tells you what this money really costs.

Step 2: Pay rent in the currency you bill in, with escalators you can pass on

MTN's U-turn came down to a currency mismatch. One fund manager put it plainly: "Tower lease costs also skyrocketed a few years ago, especially in Nigeria, following the aggressive currency devaluation." Analysts said buying back IHS allows the group to internalize lease payments and reduce exposure to dollar-linked rental contracts.

In Europe the risk is inflation. Cellnex says in its own investor presentation (the tower company selling its own story) that its contracts are c.65% CPI-linked and c.35% fixed escalators (1-2%). Its CEO has said that up to an inflation of 3.5%, we are very effective in passing the inflation to our clients. Above 3.5%, what normally enter are some caps. Set the cap at or below the price rises you can actually push through to your own subscribers. If you can't pass on a CPI increase to your customers, don't agree to pay it to your landlord.

Step 3: Sell a minority and keep what defines the network

The Canadian deals keep control with the operator. In Bell's deal, Bell will retain full ownership and control of its spectrum, all active network equipment and owned land, and TowerCo is expected to have no financial debt at closing and its financial results will be consolidated by Bell into its financial statements. Telus kept the lease term short: Terrion will lease tower capacity back to TELUS under an initial eight-year agreement with options to renew. Ooredoo and Zain each own 50 percent of the new platform and will retain their active telecommunications infrastructure.

Worked example with Bell's figures. Bell's valuation of C$4.91 billion for "more than 5,000" sites works out to under C$982,000 per site. For Telus, a Desjardins analyst put the deal at more than $800,000 per tower. At the end of June, Bell's net debt leverage was 3.7 and the company had $37.5-billion of debt. That implies EBITDA of roughly C$10.1bn. If all C$2.45bn went to repaying debt, debt would fall to C$35.05bn and leverage to about 3.46x, just under Bell's 3.5x target. Because Bell consolidates TowerCo, the rent stays inside the group and reported EBITDA doesn't change. The real cost shows up lower down: 49.9% of TowerCo's profits now go to a minority shareholder.

Step 4: Pay for flexibility before you need it

Network consolidation is where tower contracts hurt most. When MasOrange wanted out of sites after the Spanish merger, Cellnex says the new contract signed with MasOrange in Spain gives our client network flexibility in the short term in exchange for a single longer contract until 2048. In other words, the operator paid for flexibility with a longer commitment. Cellnex describes its contracts as having watertight all-or-nothing renewal mechanisms, CPI-linked escalators, and very limited churn concessions. Before you sign, negotiate a yearly allowance of sites you can decommission, a clause that applies on a merger, and the right to move equipment to another site. Once the deal is signed, you have nothing left to trade for them.

Step 5: Agree active sharing before you sell the steel

Where a single operator can't make a site pay, sharing beats selling. TIM and Fastweb+Vodafone set out a RAN-sharing plan in which "each operator will be responsible for network development in 10 regions leading to a network of approximately 15,500 sites per operator by the end of 2028", aimed at Italian municipalities with fewer than 35,000 residents. In New Zealand, One NZ and 2degrees would contribute their respective RAN assets into a jointly owned commercial entity. MTN and Airtel Africa have signed network-sharing agreements covering Nigeria and Uganda. On the size of the prize, according to ITU-T Recommendation D.264, passive infrastructure sharing can reduce capital and operating expenditure by up to 30%.

The order matters. A RAN-sharing deal that switches off overlapping sites is worth a lot to you, and much less once a tower company holds an all-or-nothing contract over those same sites. Work out the overlap with your partner first. The platform's lesson onthe per-tenant economics of a shared site explains why each extra tenant changes the numbers.

Step 6: Commit the proceeds in writing

Tell the board where the money goes before the deal closes. VodafoneZiggo did: proceeds will be used to reduce VodafoneZiggo's debt ahead of the planned listing of Ziggo Group in Amsterdam in mid-2027 through a spin-off. Telus put a number on it, saying proceeds would lower its net debt-to-EBITDA ratio by an estimated 0.17 times. Bell's plan is less clear. One analyst wrote: "We expect BCE to recycle the cash into growth areas such as AI-powered solutions, so we don't see the transaction as meaningfully deleveraging the balance sheet." If the money goes into new bets, the leaseback has to clear their hurdle rate, not just your cost of debt.

When does a tower sale destroy value for an operator?

A tower sale destroys value when the leaseback costs more than the cash earns, or when the contract locks in a network you will need to change. Here is the rule, short enough for a whiteboard:

Sell or share passive infrastructure only if all four hold:

  • Starting rent ÷ proceeds, plus the expected escalator, is below the return you will earn on the cash.
  • Rent is paid in the currency you bill subscribers in.
  • The escalator cap is no higher than the price rises you can pass on.
  • You keep the spectrum, active equipment, land and a written right to decommission sites.

MTN's South African deal failed condition 2 and its Nigerian leases failed condition 3. In 2022 MTN's South African unit sold off and leased back its entire portfolio of almost 6,000 towers to IHS for R6.2bn. Four years later it is paying to own them again.

Where tower deals and network-sharing pacts go wrong

Assuming you can walk away from a master lease

DISH showed how hard that is. When EchoStar left the network business, rent under Dish Wireless' tower leases in 2025 was about $567.8 million. Crown Castle went after in excess of $3.5 billion in remaining payments owed. The FCC required EchoStar to establish and fund a $2.4 billion trust intended to satisfy certain qualified claims, and DISH entities then filed for Chapter 11. The obligation followed DISH into bankruptcy. There is a knock-on risk for other tenants too. The Brattle Group, in work commissioned by the Wireless Infrastructure Association (which represents tower owners, so read it with that in mind), warned of rent increases of 5.7% to 10.7% to others who lease towers. One tenant's default can end up in your next renewal price.

Ignoring the regulator's veto on the way back in

Buying towers back is not purely a commercial decision. Nigeria approved MTN's deal on the condition that MTN must sell 30 percent of IHS Nigeria to local Nigerian investors. Across Africa, authorities are examining whether shared infrastructure should operate under published, cost-oriented wholesale rates. In Malaysia, when Telekom Malaysia tried to leave its wholesale deal, DNB formally rejected TM's termination notice, and DNB maintains that the original agreement remains valid and enforceable until October 2032.

Selling the co-tenancy upside too cheaply

Most of a tower's value comes from the second and third tenants, and the buyer gets that upside. Independent research from Delfineo estimates that a third tenant on the same mast can triple the original revenue with almost zero additional expense, pushing site-level margins into the 90% range. Desjardins thought Telus did well precisely because "we see no clear indication of an immediate agreement for additional tenants". If your sites have obvious co-tenants waiting, get that reflected in the price.

Signing minority deals with undisclosed governance

Look at what the Bell announcement leaves out. One pension-sector analysis noted that it does not disclose lease escalators, the length of Bell's commitments or the governance rights attached to the 49.9% stake. Those terms decide who controls where new sites go and who pays for densification.

What to do on monday as a telecom CFO

  • Pull every tower master lease agreement into one table covering escalator type, cap, currency, renewal mechanics and decommissioning rights, and flag any all-or-nothing renewal due before 2031.
  • Work out the implied leaseback yield on your portfolio at today's prices, using C$800,000 to C$1m per site as a Canadian benchmark, and set it against your marginal after-tax cost of debt.
  • Ask treasury to stress-test rent paid in dollars or euros against revenue in local currency under a sharp devaluation, the scenario that hit MTN in Nigeria.
  • Map your site overlap with your most likely RAN-sharing partner before you take a call from any tower company.
  • Draft the use-of-proceeds memo for the board now, stating the debt reduction and leverage change, in the format of Telus's 0.17x.
  • Check whether any tower company you rent from has a tenant in default, and prepare for its pricing at your next renewal.

Telus and Bell kept 50.1% of their towers, along with the spectrum, the active gear and the land. MTN sold its sites outright with dollar-linked rents and is now paying to own them again. Before you sign, settle three things: how rent rises, what currency it is paid in, and your right to switch sites off.

The full course on this sector:Finance in Telecom.

Frequently asked questions

Is a tower sale-and-leaseback really debt for a telecom operator?

In practice, yes. Under IFRS 16 the leaseback creates a lease liability on the operator's balance sheet, and the escalating rent works like a coupon on an indexed bond. MTN Group reported $4.65 billion of lease liabilities at the end of 2024, a big reason it agreed to buy IHS Towers back.

Why did MTN buy back IHS Towers after selling its towers?

MTN agreed to buy the IHS Towers stake it did not already own, at an enterprise value of about $6.2 billion, so it could bring lease payments in-house and cut its exposure to dollar-linked rents. Tower lease costs soared in Nigeria after the naira devaluation, and Nigerian regulators approved the deal only if MTN sells 30% of IHS Nigeria to local investors.

How much are telecom towers selling for per site in 2026?

In Canada, recent minority tower deals suggest just under C$1 million per site. Bell's TowerCo, valued at C$4.91 billion for more than 5,000 sites, comes out below C$982,000 each, and Desjardins put Telus's Terrion deal at more than C$800,000 per tower. Prices go up when sites have obvious co-tenants waiting.

When is active RAN sharing better than selling towers?

Active RAN sharing tends to win in low-density areas where one operator can't make a site pay on its own. TIM and Fastweb+Vodafone split Italian regions so that each builds about 15,500 sites by 2028, aimed at towns under 35,000 residents. Operators should agree sharing before any tower sale, because all-or-nothing tower contracts make it harder to switch off overlapping sites.

Go deeper

The lessons that take this article further, free to read.

  1. 1Sharing the burden: tower sales and network partnershipsFinance in telecom
  2. 2Consolidation and coalitions: mergers, tower sales and joint ventures as power movesTelecom: how the sector works
  3. 3Merger review and spectrum caps: how antitrust law kills or blesses dealsTelecom: how the sector works
  4. 4Financing the network: capex intensity and spectrum auctionsFinance in telecom
  5. 5Lease accounting (IFRS 16): the $3 trillion balance sheet shockReporting, accounting & technical finance

Sources

Finished reading?

Validate your read to earn XP and feed your radar.