I’m sure you saw that the Wall Street Journal reported this week T-Mobile and Sprint are once again flirting with each other, just several months after their parental chaperones told them to stop talking and come back home. Not surprisingly, the Towers sold off on this news.
We’ve published quite a bit on the effects a deal would have American Tower, Crown Castle, and SBA. “U.S. Towers: Churn Down for What? Dimensioning TMUS/S Consolidation Risk and What’s in the Stocks,” November 14, 2016 being our most recent “real” report on the topic), and a number of clients have asked if the math we used back then is still appropriate. The short answer is yes, but we thought it might be helpful to quickly refresh the numbers (Chart 1) and very briefly review the ideas behind them. If the deal happens and there are no offsets, we’d estimate an $8 per share hit to American Tower, $9 to Crown Castle, and $14 to SBA.
We’ve always thought same-tower overlaps – the metrics reported by the towercos to dimension this risk – understated the number of sites that would be decommissioned in a deal. In practice, T-Mobile and Sprint would choose a “surviving” network, presumably T-Mobile’s, and migrate spectrum from the “legacy” network over. When the legacy network’s spectrum and subscribers had been migrated, the legacy network would be shut down. A chunk of legacy sites would be retained, and some number of sites from the surviving network decommissioned. It’s part art, part science, but we’ve assumed ~60 percent of Sprint sites and ~15 percent of T-Mobile’s would ultimately go away. As a partial offset, the companies would have to upgrade all the sites they retain to accommodate each other’s spectrum bands – Sprint’s 800 MHz and 2.5 GHz on T-Mobile’s sites, and T-Mobile’s 600 MHz, 700 MHz, and AWS on Sprint’s sites – which would entail amendment revenue. Leasing activity would likely initially increase as the networks were prepped for spectrum migrations, and then churn would hit as leases ran out, with an average term of five years or so (we would note that the MLAs Sprint recently signed with Crown Castle and SBA included term extensions on at least a portion of its sites, so that would shield them for a couple more years). The present value of what’s being lost, netted against the interim benefit of heightened leasing activity, yields the NPV hit.
There are other factors to consider as well. For example, if Sprint and T-Mobile were forced by regulators to divest spectrum to close a deal, AT&T and Verizon would likely have to incur amendment revenue to deploy it, acting as another offset. We dimension what this might look like in Chart 1. From a bigger picture perspective, what would going from four to three players mean for the odds of Dish Network entering as a new number four? That’s hard to handicap, but the odds would be higher, and the value of a full Dish build-out would likely more than offset the hit from T-Mobile/Sprint. It’s worth bearing other limbs on the decision tree in mind.
Finally, we need to consider the probability that an agreement is reached and, if an agreement is reached, what the odds of approval in DC look like. We’re not smart enough to know what those are, but the stock reactions suggest Mr. Market is baking in a fairly high probability that it happens, at least based on our math.
I hope this refresh was helpful. Feel free to reach out to me if you want to discuss further.