Inside the due diligence process from both the operator and towerco perspective

Neil Taylor joined Helios Towers Africa in May 2011. Neil has a unique perspective on leading due diligence processes for infrastructure sharing transactions as he’s sat on both the buyers and seller’s side of the negotiating table; Neil was previously General Counsel at Millicom and part of the core team that structured joint venture towercos with Helios in Ghana, DRC and Tanzania. He has been practicing for over 20 years, starting out as a commercial litigator before increasingly specializing in M&A transactions, including over 20 deals with leading electronic component distributor Avnet before moving to Millicom in 2008.

TowerXchange: Once an operator has decided to investigate selling their towers, how do they invite bidders? Is it simply a case of calling the investment bankers with experience in this market? What information is needed to determine the market for your towers?

Investment bankers would once have been the first calls to make, and I’ve got tremendous respect for them, but now everybody knows that Helios, ATC, IHS and Eaton Towers are the main players to talk to.

In terms of the information needed, ultimately it’s a real estate business, so it’s about location, location, location. Where are your towers? Are they secure? Are they in locations desirable to other prospective tenants? If your rollout plan was to put another tower next to competitors’ sites, then you may have a less valuable portfolio as overlap obviously affects desirability and value.

You need to know the condition of your towers and take a realistic view of their state of maintenance and capacity to add other tenants. And of course a lot depends on the demographics of the market – five competing operators means there are more prospective tenants than in markets with one or two license holders.

But the most important thing is to figure out your organisational goals from infrastructure sharing, as the sale value of your towers depends heavily on the lease rate you’re willing to pay. This is why in the same market you might see similar towers changing hands for $230k per site in one transaction, yet $70k per site in another – it’s a function of differing goals; balancing maximizing cash, minimizing opex, or finding a sweet spot that protects and stabilizes opex while releasing some cash.

TowerXchange: What were Millicom’s objectives in the deals with Helios?

Millicom saw the value of getting into tower game, perhaps because Millicom shareholders Kinnevik from Sweden had in the past “sold the paper mills but kept the forest” – they understood that it’s okay to not own a piece of the value chain.

I was part of the team that did Africa’s first substantial infrastructure sharing transaction between Millicom and Helios, and we realised that because towercos typically trade at higher multiples than Mobile Network Operators, if we moved assets from one balance sheet to the other, it meant benefitting from that higher multiple. Of course, if operators want to retain a stake, it’s important that the terms of the transaction aren’t so harsh as to kill the value for the towerco.

Millicom knew our objectives for each deal. With so many different moving parts, the deal team could have easily got confused, so it was important to decide what we cared about in advance.

The Ghana deal, and the Tanzania and DRC deals were definitely separate. We were confident enough to tether the latter two together as we knew what we were doing by then. So while we signed the Tanzania and DRC deals within three weeks of each other, we certainly didn’t wrap one up then start the next and finish it that quickly – much of the work was done concurrently.

They turned out to be good deals for both parties, hence I’m able to sit on both sides of the transaction. In fact there’s a bit of a running joke about it; when Helios come across any challenging points in the Master Lease Agreement and ask why they were included, I usually say “because I made you sign it!” It helped that Millicom took a stake and had an interest in the towerco being a success, and both sides are pretty happy.

Most M&A transactions are a one-off; you’re most likely to not see the buyer or seller again after the transfer, whereas tower deals have same intensity of negotiation, but you’re married at end of it, so you need to enter into it with the right spirit.

TowerXchange: One towerco, that shall remain nameless, complained “data on the physical network is terrible! Bid documents are not great, and we like to check everything.” What data does the towerco need to conduct it’s due diligence and, remembering back to your time on the operator side, why is assembling complete data such a challenge for the operator?

Assembling complete data is a challenge because when Africa’s networks were being built, operators weren’t measuring compliance against objectives for a future asset sale, they were measuring against fast rollout objectives. In some African countries it’s even written into the law that if you apply but don’t receive a building permit by a certain time, go ahead and build anyway.

What a towerco wants is a ground lease that’s transferable, and terms that that don’t allow the landlord to terminate or renegotiate every time a different operator’s installation or maintenance team shows up at the site. Towercos will expect the required aviation, building, and environmental permits to be available. It’s also good to have a sense of what the operator thinks is on each tower, available space, hub sites and MSC. Other nice to know information might be the age and model of the generator, anticipated decommissioning date, and an accurate maintenance record that reflects whether scheduled maintenance actually was undertaken.

The reality is that sometimes operators only discover what data is required when they’re putting together the due diligence package. I have brought in third parties to help put the due diligence package together, but in my experience the learning curve for those outside resources is too steep – it’s better for operators’ Group headquarters to send people.

While data doesn’t need to be complete for operators to start exploring infrastructure sharing opportunities, in the end somebody has to get all these pieces of paper, and towercos are usually happy to second people to help operators get everything together, although operators are sometimes reluctant to accept our help.

No towerco will walk away from a deal because documentation isn’t perfect, but on the other hand towercos aren’t going to pay until the documentation has gone from “not perfect” to “acceptable”. Ultimately we’ve got to make it clear to our investors that the portfolio has the value we ascribe to it.

TowerXchange: The most common complaint about infrasharing transactions doesn’t seem to be that deals don’t close, but that deals take so long to close – how can operators and towercos accelerate transactions without compromising the quality of due diligence?

Operators should spend a month or two, even if it’s at expense of first mover advantage, getting database together and knowing what they do and don’t have.

But data isn’t the only reason for delays. Adjustments based on what comes out on due diligence inevitably take time. For example if an operator claims their opex is $2000 per site, yet we see there’s been no maintenance on several sites because they fired the contractor several months ago, then the real opex might be $2400. The cost of maintenance is never zero, towercos will figure this out and adjust accordingly.

Another cause of delay can be the time taken to secure infrastructure provider licenses. The Tanzania transaction took longer to close for no other reason than that the infrastructure license took eight months rather than the four months it should have taken to issue.

It doesn’t have to be a long drawn out process. But it does require focus on details on a tower by tower basis. It’s not launching the towerco that is time consuming, in fact our CEO Chuck Green developed a hundred day launch plan to start a towerco from which we have implemented three times!

TowerXchange: Much of the work post-deal consists of the novation and transfer of hundreds of real estate contracts – tell us how the legal teams on both sides get to grips with all these leases, assessing what they’ve got, the right to extend and the terms to extend?

The MNO and towerco do this together as the MNO has established relationships with their landlords. It’s a big task, but it’s not complicated and it can be completed surprisingly quickly. There are simple, easy to explain things: we need to be able to add additional tenants without landlord consent, we need to be able to borrow against the tower, and we need to avoid terminations.

The towerco and the MNO will establish the extent to which payments under the ground lease may increase as a result of the process. The conversation starts with “we’d like to make changes and we’re willing to pay you something to make those changes,” so it’s not the worst conversation the landlords have ever had! The only complications might be if you can’t find a landlord, or a government entity might be the landlord, and if they’re not responsive to change.

TowerXchange: In the case of privatized national operators, where the government might retain a stake or at least have a political interest in the ownership of infrastructure assets, is the range of options in terms of deal structure limited – could sale and leaseback be too complex, forcing a preference for operational leases?

In many cases there are ways round that obstacle. The challenge facing Vodafone Ghana was a one-off. Vodafone had just completed the acquisition from the State and everyone knew what they’d paid for it. There were all these discussions going on in the press, and the local opposition were suggesting that the crown jewels had been undersold. If Vodafone had sold all the towers so soon, you could forsee an argument that they were selling it for more than they paid for it! Government involvement isn’t necessarily a problem for every deal.

TowerXchange: What regulatory conditions make countries more favourable to telecoms infrastructure investments? Are there any red flags that make countries less attractive for infrastructure sharing investments?

Some regulators have more to say about and actively support infrastructure sharing than others. We’d be concerned about any regulations that effectively prohibit sensible business – but we don’t see that in Africa.

Part of my job is to understand the regulatory environment; the license regime, the costs of the license, what regulations have been written for infrastructure sharing or that apply to infrastructure sharing. In many cases we’re going to find ourselves driving regulations that are still being drafted, but we don’t shrink from engaging with regulators.

As far as I know, only one market in Africa regulates the pricing of co-locations, and we’d certainly want to be involved in the conversation if any other countries were thinking of following that lead.

TowerXchange: So is it preferable to let markets set pricing?

I think operators should be allowed to prioritise different things; their organizational goal might not prioritise cash release or maximizing equity stake over negotiating the lowest possible rental rate, or they might want to pay a premium for optimum uptime rather than building a low cost network.

TowerXchange: Why would either an operator or towerco want to pass through energy opex? I can understand passing through exposure to diesel price changes, but why disincentivise the towerco to maximize energy opex savings by passing them all through the the MNO?

Well nothing’s ever completely ‘fixed,’ there’s always going to be an escalator on CPI or on unit price. However, at Helios we believe in a model where we invest to reduce fuel opex, giving some certainty about costs for the operator. It’s probably the biggest difference compared with the Indian model.

On the other hand, if an MNO wants a power pass through clause because they don’t want us to make money on energy so be it, but under such circumstances we’re not incentivized to invest in saving power. There’s enough value to go around for everyone in these deals.

TowerXchange: What are the main risks that operators can choose to pass on to towercos in tower transactions?

The risks that they can pass along with selling what they recognize to be non-core assets includes uptime, of which a subset of that is maintenance, power provision and security; foreign exchange; and not having to put capital into passive infrastructure.

TowerXchange: How does attractiveness of a deal alter in countries with higher country risk?

We’re in the DRC, so Helios certainly isn’t adverse to operating in countries that are perceived to be higher risk. In some ways, country risk opens up more opportunities to the towerco in that we’re not as visible as the MNO brand. MNOs are on every billboard reminding you that they’re there and doing a lot of business. In comparison we’re very dispersed. If rebels wanted to take over a towerco, they’ve got 700 sites to take over, and then they’ve got to run them! We’re recognized as doing something you want to have done, so whether you like the government or not, towercos are not an obvious target.

Of course we look at political risk, and ethics and compliance programmes need to be robust, but we wouldn’t say we don’t want to do business in countries with higher political risk, as long as our employees are safe.

TowerXchange: In this second edition of TowerXchange, Laurent Viviez from AT Kearney challenges whether the tenancy ratios targeted in the first round of deals in Africa are achievable – what do you think?

Helios are on track to do what we set out to do in Africa, but the honest answer is it’s too early to tell. Tower leasing is an extremely “lumpy business” – it’s never a straight line to achieving your objectives, some times you’re sitting “fat, dumb and happy” above your target tenancy ratio, while at other times you find yourself in a trough below target.

Factors that can push a towerco into one of those troughs below target tenancy ratios include variations in Group capex budgets, unanticipated consolidation, different bidders winning a new license than expected, or a new operator entering markets where there’s a supposed moratorium on new builds, yet still building hundreds of new towers!

Ultimately it comes down to location – if a tower is in the location you need, if space is available, and if there are enough subscribers around, then more than one tenant is going to want space on that tower. They’re unlikely to build.

I don’t think any of the towercos see a big gap between targeted tenancy ratios and what is proving to be achievable, but we’re not jumping for joy at realizing cheap acquisitions either – operators are getting good value for African towers.

TowerXchange: Could an operator-led towerco model work in Africa?

Every operator seems to have been been in conversation with one of Indus’ partners, yet it never seems to come together. Why? It may be because Africa is not one country, it’s 53. Maybe an operator-led model will happen in a big country such as South Africa or Nigeria, but there are established independent towercos in both those countries, so what would be the driver?

Operators worried that they’re giving away too much value should be worried about missing the boat in the value creation opportunity – and consider retaining a stake.