Ladies and gentlemen, thank you for standing by, and welcome to the SBA Second Quarter Results Call. [Operator Instructions] As a reminder, this conference is being recorded. And I would now like to turn the conference over to our host, Mark DeRussy, Vice President of Finance. Please go ahead.

Thanks, Caroline. Good evening, everyone, and thank you for joining us for SBA's Second Quarter 2021 Earnings Conference Call. Here with me today are Jeff Stoops, President and Chief Executive Officer; and Brendan Cavanagh, our Chief Financial Officer. Some of the information we will discuss on this call is forward-looking, including, but not limited to, any guidance for 2021 and beyond. In today's press release and in our SEC filings, we detail material risks that may cause our future results to differ from our expectations. Our statements are as of today, August 2, and we have no obligation to update any forward-looking statements we may make. In addition, our comments will include non-GAAP financial measures and other key operating metrics. The reconciliation of and other information regarding these items can be found in our supplemental financial data package, which is located on the landing page of our Investor Relations website. And with that, I will now turn the call over to Brendan.

Thanks, Mark. Good evening. SBA had a tremendous quarter with results for the second quarter, ahead of our expectations in most key areas. Total GAAP site leasing revenues for the second quarter were $524.1 million, and cash site leasing revenues were $514.6 million. Foreign exchange rates were ahead of our previously forecasted FX rate estimates for the quarter, contributing $3.1 million of incremental site leasing revenue in the second quarter. They were also a tailwind in comparison to the second quarter of 2020, positively impacting revenues by $3.5 million on a year-over-year basis. Same-tower recurring cash leasing revenue growth for the second quarter, which is calculated on a constant currency basis, was 3.4% over the second quarter of 2020, including the impact of 2.5% of churn. On a gross basis, same-tower growth was 5.9%. Domestic same-tower recurring cash leasing revenue growth over the second quarter of last year was 5.5% on a gross basis and 3% on a net basis, including 2.5% of churn. Domestic operational leasing activity or bookings representing new revenue placed under contract during the second quarter was up significantly from the prior quarter and represented the highest quarterly level since 2014. Even with this high level of execution, our domestic new lease and new amendment application backlog finished the quarter at a multiyear high. These backlogs support our expectations for continued strong domestic operational leasing activity throughout the balance of this year. During the second quarter, amendment activity represented 34% of our domestic bookings with 66% coming from new leases, the first time in many years the bookings from new leases outpaced that from amendments. The big 4 carriers of AT&T, T-Mobile, Verizon and DISH represented 97% of total incremental domestic leasing revenue signed up during the quarter. Internationally, on a constant currency basis, same-tower cash…

Thanks, Brendan. We ended the quarter with $12 billion of total debt and $11.7 billion of net debt. Our net debt to annualized adjusted EBITDA leverage ratio was 7.3x, a return to our target range of 7 to 7.5x, faster than originally anticipated after the PG&E acquisition during the first quarter. Our second quarter net cash interest coverage ratio of adjusted EBITDA to net cash interest expense was 4.4x. On May 14, the company, through existing trust, issued $1.165 billion of secured tower revenue securities Series 21-1c, which has an anticipated repayment date of November 9, 2026, and a final maturity date of May 9, 2051. These securities have a fixed annual interest rate of 1.631%, payable monthly. The net proceeds from this offering were used to repay the entire aggregate principal amount of the 2017-1c tower securities and for general corporate purposes. Then on July 7, the company amended its existing revolving credit facility. Among other things, this amendment increased the total commitments under the facility from $1.25 billion to $1.5 billion; in addition, extended the maturity date of the facility to July 7, 2026; lowered the applicable interest rate margins and commitment fees under the facility; and incorporated sustainability-like targets into the facility, allowing for interest rate and commitment fee adjustments based on how we perform against those targets. We are pleased to be one of the first companies to include such sustainability-focused provisions in our credit facility. As of today, we have no amounts outstanding under our revolver and the weighted average interest rate of our outstanding debt is 2.9% with a weighted average maturity of approximately 4.5 years. During the second quarter, we did not purchase any share of our common stock. As of today, we have $475.1 million of repurchase authorization remaining under our $1 billion stock repurchase price. The company's shares outstanding at June 30, 2021, were 109.5 million compared to 111.9 million at June 30, 2020, a reduction of 2.1%. In addition, during the second quarter, we declared and paid a cash dividend of $63.5 million or $0.58 per share. And today, we announced that our Board of Directors declared a third quarter dividend of $0.15 per share, which is payable on September 23, 2021, to shareholders of record as of the close of business on August 26, 2021. Today's dividend announcement represents a payout ratio of 22% of second quarter AFFO per share, which leaves us ample room for material future dividend growth. And with that, I will now turn the call over to Jeff.

Thanks, Mark, and good evening, everyone. As you have heard, we had a very strong performance in the second quarter, one of our best in quite some time. Domestically, our customers were extremely active. This quarter, we signed up more new leasing revenue than we have in any quarter in 7 years. We believe the second quarter was the start of a multiyear increase in U.S. wireless carrier CapEx. Our backlogs of lease and amendment applications ended up near record highs at the end of the quarter, notwithstanding the high levels of bookings, which obviously bodes well for activity levels throughout the rest of the year and into next. Our customers are focused on building out their 5G networks with a particular focus on upgrading their macro networks. Leasing activity levels were driven by initial C-band initiatives, 2.5 gigahertz deployments, FirstNet amendments, general coverage expansion and the beginning of DISH's network build-out. These initiatives also drove our services volumes to record highs. Last quarter, I was pleased to discuss how our first quarter services results were the best in 7 years, but that didn't last long as we reported 18% more services revenue in the second quarter than we did in the first quarter, producing over $51 million of quarterly revenue, the highest in our company's history. We also finished the quarter with the highest services backlog in our history. These results allowed us to increase our full-year services outlook by $25 million. As you could tell, our domestic customers are very busy right now. That sets us up for a very good 2021 and 2022. Internationally, the second quarter was also very solid for us, with quarter-over-quarter sequential increases in new lease and amendment executions and increasing CPI-based escalators in a number of our markets. During the quarter, we…

[Operator Instructions] Our first question comes from the line of Ric Prentiss from Raymond James.

A couple of questions. First, obviously, PG&E was a big transaction. We saw a story come out in the last few weeks that they plan to bury like 10,000 miles of electric line. Is there going to be any changes to the sites you've got, where they might be taking electrical lines down, making it obviously easier to couple on them? And what kind of interest activity have you seen on those sites?

We've seen great activity. We're ahead of plan, and there's a lot of demand in terms of application backlogs. And in terms of that story, Ric, that relates primarily to the lower wood poles, the ones that are closer to the end user as opposed to the high-power transmission poles that are the ones that were really are part of our transaction. So we don't see a lot of issues related with that. But if, in fact, there are poles or assets that were part of our transaction, there's a fairly elaborate -- if anything should happen to those in terms of being no longer needed or used by PG&E, there's a fairly complex set of rules that deal with that. So something that we're covered on and feel very good about. But for the most part, it's going to be the shorter wood poles that they're talking about.

Makes sense. Okay. Also, the Tanzania joint venture, one of your peers recently, doing a big private equity transaction in Europe, has started giving consolidated AFFO per share as well as attributable to common AFFO per share. Have you guys considered, as now you get a little more joint ventures going, that you might move to, obviously, EPS stories. They are -- always take out minority interest. But any thought about moving towards a measure of a more proportionate AFFO to common?

Yes, Ric, to this point, obviously, nothing has been material enough to do that. We do always back out -- we've previously identified when we haven't consolidated something. Obviously, the JV piece of the AFFO per share. But it's something we'll look at. If it becomes material, we'll highlight it for you guys. We just haven't had any, really, to date.

Makes sense. Final one, obviously, we make that adjustment from AFFO to funds available for distribution. We've been watching, I think you guys call it, amortization of capital contributions, which is still very low. I think it was $13 million in the first half of this year, $33 million all of last year. One of your peers has almost 20x that level of noncash items that get put into AFFO. Any thoughts of any changes coming from you that would take it up higher? Or is it fairly low? Because it's obviously one of the big noncash items that kind of is different between the tower companies.

Yes. I mean I don't think so, Ric. The things that would drive it up is obviously more augmentation activity of the towers, which is certainly possible, given the increased amount of activity in general that we're seeing. But the offset to that is that as we have longer terms to our leases, particularly the Verizon MLA, which we talked about last quarter, and some of the new agreements that come with longer term, that tends to stretch out the period over which you're amortizing and actually reduces the amount. Plus just -- it's a factor of which towers are getting hit, whether work needs to be done at the site, how much cost is being reimbursed, all those kind of things. So I don't expect it to actually move up materially at any point in the next year or 2.

Yes. It's very different, Ric, when you're not doing small cell fiber work where it's more common that your customer contributes a large amount of the initial CapEx upfront, then you amortize it into that amount. I mean, that's never really been a big issue in the macro tower side of the business other than just the regular augmentation stuff, which is -- we don't foresee any material changes in that.

Makes sense. Obviously, I always like going back to cash and the ability to pay dividends. So I appreciate that, and you guys stay well.

Our next question comes from the line of Simon Flannery from Morgan Stanley.

I was wondering if you've had any more clarity from T-Mobile about their plans for the Sprint towers. Do you think the churn is going to be consistent with the current contracts? Or do you think there's the potential to either keep some of those towers or have them pushed out? And any updates on the Oi restructuring and how we should think about the impact maybe opening up the potential for more investment in Brazil once that goes through?

Yes. In terms of the T-Mobile-Sprint question, Simon, T-Mobile is very organized and very, I think, ahead of what people thought in terms of their network planning. So there's a lot of crisp communications going back and forth. But I don't know that it materially has changed our prior statements on what we expect in terms of decommissioning. And we don't have any additional news on the uptake of the potential Sprint sites. But I would just say that T-Mobile is very prepared in their work in this area. In terms of Oi, I believe the latest is that we're not expecting anything until at least the fourth quarter of this year, maybe into next. And our prior statements in terms of our exposures to both -- or to all 3 of TIM, Tigo and Claro, those have not changed.

Okay. Great. And do you think there's an opportunity for you to do more on build-to-suit? I think you did about 100 in the quarter. And then it sounds like it's part of the Tanzania story as well, but you haven't done a huge amount relative to what we've been hearing from some of your peers.

Yes. I think there's always an opportunity. The question is does it represent the terms and the financial returns that we're looking for. I think the universe of opportunities would allow us to do more.

Our next question comes from the line of Phil Cusick from JPMorgan.

It's Phil Cusick. Start with the Paradigm deal. Was this the team you worked with from South Africa? Or is this a different team?

No. We didn't work with any of the AMT guys in South Africa, Phil. This is Steve Marshall and Hal Hess and Steve Harris, guys that I've known forever. And worked very closely with Steve Marshall at WIA. So that's really just through the industry, this all came together.

Okay. What's your potential capital commitment over time for this?

Well, it's not much beyond the $175 million. I mean, Brendan what's...

Yes, that's the total purchase price, the $175 million plus a few incremental costs. So that represents everything. So it wouldn't be materially more than that. I mean down the road, we'll have to see how things...

Yes. I mean we've modeled in some upgrade CapEx and some refurbishment CapEx, but talking about material numbers there as well.

Okay. It just sounds like if you're going to do something like this in a whole bunch of new countries, it's -- I would think you have a sort of number that you'd be willing to invest alongside them to make these more interesting over time.

Yes. The JV, Phil, is specific to Tanzania. There's no obligation to do anything alongside them going forward. So there's no commitment as far as that goes.

If we could find more opportunities like Tanzania, though, we'd be excited.

Our next question comes from the line of Michael Rollins from Citi.

With respect to the domestic leasing activity that you were highlighting just earlier in the call, how does this compare with the internal expectations that you've had? And what does this mean to the levels of gross leasing growth that you can achieve in the domestic business over the next couple of years? And then just separately, I just wanted to follow up on another thing that was said earlier. I was just curious if you could unpack why the mix of leasing activity had a greater amount of new sites versus amendments in the domestic business?

Well, the last one is an easy one to answer, Mike, and it's DISH. Because DISH is coming online in the form of entirely new leases, and they're very busy. So that's the answer to that. In terms of the second quarter, we achieved results that were ahead of our expectations. We did have some fairly robust expectations, but even those were exceeded by the level of second quarter activity. And in terms of the growth rate going forward, I don't want to get too specific other than to say it's going up.

In the past, I think, and you can correct me if I don't have this right, were you talking, I think, high single digits of where you expect the gross domestic growth to get over the next few years? I'm just curious if you could just revisit that perspective.

Yes. Mike, from a growth standpoint, we certainly expect it to be mid- to high single digits in the future given what's going on now. So yes, when you're talking about going out a lot of years, obviously, we're getting bigger and bigger, so you have to add that many more dollars to make that happen. But given the activity levels today, I think as we get a year or so out, you'll start to see us return to levels that are similar to what we've reported in the past when we were busier.

Our next question comes from the line of Nick Del Deo from MoffettNathanson.

First, for the question that Mike just asked, it sounds like DISH was really an outsized contributor to leasing this quarter. Can you just -- if you back that out, can you talk about how activity from the big 3 carriers, in particular, has changed over recent quarters? Was it DISH in particular that really drove the outperformance in leasing versus your expectation? Or was that outperformance more broad-based?

I would say that the ones that had kind of step function changes in their levels of activity were DISH and Verizon. And the Verizon story is very understandable. It's coming off the C-band auctions, their public statements and our MLA with them.

Okay. Okay, that's helpful. And then one on Tanzania and the partnership with Paradigm, do you have an option to eventually buy out their stake or vice versa? Or is there no mechanism in place to do that?

No, there is a mechanism that gives us the right to buy them out.

Can you talk about the time frame or any other parameters? Or is that proprietary?

It's not proprietary. It's 5 years-ish. Might be a reason why it gets accelerated a year or extended a year, but that's generally the time frame that we're looking at.

Our next question comes from the line of Brett Feldman from Goldman Sachs.

Following up a little bit on C-band. Now that you're starting to see what some of the carriers are doing with C-band, how much visibility do you have and to the extent to which that's mostly going to be a significant amendment project versus any visibility into a new site project? Because for all of the operators that you already have as tenants are on 1 C-band. This would be a higher frequency than anything they've ever historically used on a macro site. So it would apply that they might need densification, but I don't know if you have that visibility at this point in time. And then just following up on the services backlog, I'm just curious, is all of that services work being done on your towers, meaning is it a fairly good leading indicator of what your own leasing is? Or are you actually winning a degree of the business across other portfolios as well?

The answer to your last question is yes. It is mostly all on our towers, which is why we have the confidence we have. And in terms of your first question, Brett, there will be some leases in there. But for T-Mobile, Verizon and AT&T, it's going to be mostly amendments. And for DISH, it's going to all be new leases.

Our next question comes from the line of Walter Piecyk from LightShed.

Jeff, I'll preface this question by noting that last quarter, your average -- price of your share repurchases, I think, $258, and the stock is now $340 or $342. But in this quarter, you didn't buy anything. So I'm just kind of curious, is this -- I mean, I know this acquisition was, whatever, $175 million, is kind of in the ballpark of what you spent last quarter. Is that the connection that we should draw? Or is it maybe where the stock price is in terms of the activity this quarter?

No. We wanted to come down off the 7.6%, and we came down quicker than we thought. And by the time we knew that, we were blacked out.

Got it. And then on the domestic ramp, when you look at your guidance for new lease activity, it's obviously very high, so there's going to be a ramp. Can you give us a sense of Q3 versus Q4? I mean obviously, we're going to have to see some type of jump in Q3, but is the bigger jump really going to occur in Q4 in terms of the new lease activity?

Yes. You're talking about the same-tower growth rate, Walt, alright? Yes.

Yes, right.

Yes. We expect it to step up sequentially each of the next 2 quarters, but certainly, the fourth quarter would be a bigger step-up is our expectation based on the timing of when this -- the stuff we're signing...

Orders and stuff came in, yes. That makes sense. And then, I mean if -- T-Mobile is got to be a component of that. I mean, doesn't this kind of speak to the timing of how their integration is going, if it's still like more of a Q4 event?

Yes. I mean they're busy. They're busy and...

But they're not getting them activated in Q3, right? It's still not happening until Q4 for them.

They're getting stuff activated in Q3 as well. So I mean, it's a mix question, Walt, but they're busy in activating stuff in both -- we expect in both quarters.

It's just interesting given that C-band is right around the corner for Verizon, and they're still kind of back-end loaded, but whatever. And then international, that's also a big number. So that's going to be a big ramp second half, then, for international as well, right? To hit that $13 million target?

It will be a slight ramp. I don't know if I would call it a big ramp. we can walk through that number...

Slight relative to the overall company for sure, but for international -- for what your levels have been in the first half of the year, it seems like it's kind of big now?

Yes. It's increasing. I mean, certainly, the lease-up has been a little bit better just recently, and we expect it to be better as we move through the year. I don't know that I would agree with your characterization, though, so we should probably talk about that separately.

Sure. Okay. One last one then, sorry. But T-Mobile's use of DISH's 600 megahertz spectrum, did that generate an amendment or a colocation or anything?

It will when it happens.

Our next question comes from the line of Batya Levi from UBS.

A couple of follow-ups. First, can you provide some color on the pacing of Sprint churn for this year? I think it was close to $2 million in the first quarter. How do we think about it for the rest of the year and maybe a color if that $13 billion still holds for next year? And another question on the services side. The margins are coming in a little bit lighter than historical levels. Is there any opportunity to improve the services margins? And how do we think about the '22 services revenues?

Batya, on the Sprint churn piece, it's expected -- we had, I think, what we reported last time was about $1.7 million or so in the first quarter. We expect the full year to be somewhere between $8 million, maybe $8 million to $9 million or so for the year, the fourth quarter impact. And this is, again, on a full-year, year-over-year basis. Will be a little bit bigger because, obviously, there are some of those leases that are going away as we move through the year. In terms of next year, right now, we still think maybe $30 million to $35 million or so is appropriate. I have to hedge a little bit only because the timing of when we get notice is, what they specifically decided to do, is a little bit of a fluid situation. But based on what we know today and lease expiration dates, that we think that's a reasonable estimate.

Yes. In terms of the services margins, I mean, our services business is comprised of 2 distinct offerings. One is site acquisition, zoning, kind of soft work; and then the second is construction. Construction typically produces about half the margin of site ac and zoning. So the mix between site ac, zoning and construction is what drives the margin. And the margins have actually been, based on the type of work being provided, historically strong compared to QR experience. In terms of next year, I mean, we'll give you next year's guidance when we get there, but there's no reason why the services business, because it is primarily on our own towers, should materially slow down. It's going to reflect leasing activity.

Our next question comes from the line of David Barden from Bank of America.

Just a follow-up on a couple of things. So I guess, first, Jeff, I think last quarter, when we were talking about the services business, you said you didn't see any reason for there to be a material change to Q-on-Q. Obviously, there was a change. You kind of listed off a number of things that were contributing factors, but was it really like how quickly the C-band deployment came out of the gates? Or was it really how quickly maybe T-Mobile came out of the gates? Or was it just everything just turned out to be better than expected?

Really more of the latter. I mean C-band came out quicker. DISH came out quicker. T-Mobile continues to be extremely strong.

And just with respect to the disappointment, maybe you can comment on this, maybe you can't. But is it -- they've obviously publicly announced what their plan is, to build out the Vegas market. Is this the kind of bump that you saw specifically related to that? Or is there something grander that's going on that maybe we haven't heard about yet?

Our work is pretty well spread out, David. I'm not sure exactly how they geographically define those comments you made, but we're seeing a fair spread to the geography.

Okay, interesting. So just a couple of housekeeping ones. The Tanzania contract, is that all denominated in shillings?

The purchase price? Are you asking about the purchase price or the leaseback?

The leasing, the leasing contract.

It's roughly -- well, this includes the other third parties, so I don't know the exact breakdown. But the total leasing revenue that's contracted is about 67% in U.S. dollars, when you exclude the pass-through stuff.

Most of the space is rented in U.S. dollars, the energy pass-through is all shillings.

Got it. Got it. Okay. Then my last one, if I could, sorry. The roughly $100 million, $95 million, that's not Tanzania that you guys invested. Where did you guys put that money this quarter?

I'm sorry, the what?

Sorry, the $95 million to $100 million that you guys -- that's not part of the Tanzania deal that got spent this quarter, where did that go?

The $95 million discretionary CapEx?

Correct.

Yes, yes. Went into various acquisitions, also some new builds. We mentioned, we built about 100 sites during the quarter, so the total CapEx was up for a number of deals. The deals -- if you're asking where the deals were located, it was a mix. They were mostly deals in the U.S. in this quarter.

Our next question comes from the line of Eric Luebchow from Wells Fargo.

I was wondering on AT&T, you didn't mention them as one of the customers that had a step-change function in activity. Have you seen any change from them since they announced the WarnerMedia spin-off around C-band urgency? And you mentioned FirstNet as well in your comments. We have presumed that was starting to wind down, but maybe you could comment how much longer you expect FirstNet amendment to persist.

Yes, AT&T is relatively steady. Did not have the same change in velocity that some of the others that we mentioned did. And the FirstNet stuff, I think we are past -- well past the halfway point, but not close to being finished.

Okay. Fair enough. And just wondering, too, is there any update or anything to point out on SBA Edge, the data center platform, in terms of deal activity? And any update on timing when that could potentially become a modestly more material part of the business?

Yes. I believe, and I just heard this today internally that we've now gotten our third contract for a true mobile edge computing site, so that would give us 2 data centers and 3 or 4 mobile edge facilities. So moving forward, but quite a bit of ways away for reaching that materiality point you referred to.

Our next question comes from the line of Colby Synesael from Cowen.

You mentioned the JV's $175 million. I'm just curious what's your ownership stake of that. And why the JV? You obviously can afford to do that straight up on your own. Is that meant to signal some type of shift in strategies or simply they're the ones that brought the deal to you and therefore, were allowed to take a piece of it? And then secondly, the new AT&T-DISH partnership. I'm curious if you view that as incrementally negative or positive to your business and I guess just the broader tower industry.

Yes. The Paradigm folks were responsible for bringing the deal to us, Colby, and had a long relationship with Airtel in Tanzania with their former employer. So that's why we're where we are. I don't think we're disclosing the exact split, other than we are clearly the majority owner in all facets. In terms of the AT&T-DISH deal, it doesn't really change DISH's requirements with the FCC, which is for their own fixed network. Arguably, it could impact their growth beyond that to just roll on AT&T's network. But we'll have to see. I mean, pruning never is the first economic choice. But I mean, it certainly hasn't in any way impacted DISH's case out of the blocks here in 2021.

I just had a quick follow-up. One of the parts of that agreement is the ability for DISH to effectively give some of its spectrum to AT&T to deploy in its behalf, to which then, DISH can use. Do you know if that would meet the threshold that the FCC has in terms of their 70% coverage requirement by 2023, i.e., they could effectively just go on AT&T's towers using their spectrum and that would meet the requirement wherever they may choose to do that?

I don't know the answer to that.

Neither do I, but thank you.

Our next question comes from the line of Matt Niknam from Deutsche Bank.

Just on capital allocation, now that you're back within your target leverage range, I'm wondering how you're thinking about that and how you're sort of prioritizing uses of excess cash post the dividend? And then specifically on the M&A front as well, if you can talk about the latest you're seeing, both in terms of valuations and number of opportunities, particularly internationally.

Well, the number of opportunities internationally is high. Prices are high. You need to be very selective. I don't know that we will be able to do a bunch of things like we're doing in Tanzania. I hope we can, but I don't know that, that will be the case. In terms of capital allocation, generally, we continue to believe our shareholders are best suited by a "lower dividend, higher dividend growth strategy". Gives us the most flexibility, produces very good growth numbers. So that's what we will continue to pursue there. And given our access to capital and interest rates, we will look to stay fully invested within our target leverage range, whether that be portfolio growth or stock repurchases.

Our next question comes from the line of Jeff Kvaal from Wolfe Research.

Yes. I guess, first, would you mind offering us an update on how close you are to bumping up against supply or labor challenges in this market?

So far, so good. We really haven't seen any issues on the labor side. I mean the fact that we maintain a sizable in-house crew gives us some flexibility there. And in terms of the equipment side, we've not heard any issues around our customers delivering the equipment to the sites for installation. Now we read about all the same semiconductor issues that you read about, so I guess that could change, but it has not manifested itself so far.

Okay. And then secondly, I think Brett asked earlier about potential for densification on the higher frequencies of C-band. What are you hearing from your carrier partners about their propensity to add some cells? Or do you think that they will figure it out with better antennas, et cetera, et cetera?

There'll be densification. But just like every generational upgrade, you go first to your existing sites because that's the quickest, the best bang for your buck. So this is happening exactly like it always does.

Our next question comes from the line of David Guarino from Green Street.

Can you remind us of the company's views on expanding into new markets and if there's any off limits? And then assuming you are open to expanding into new markets, how do you determine whether or not to utilize a JV partner?

The last part of your question is based on what they would bring to the table and mostly operational context as opposed to capital, at least so far. And in terms of -- there are certainly ones that are off-limits, either by choice or by legal necessity, places like Cuba and China. I mean, we're just not allowed to go there. And then there will be others where the analysis of the risks outweigh the rewards. In a country where you have very little rule of law and you have the potential for nationalization of assets, those are not going to be markets where we're most excited about.

That makes sense. And then how do you handicap the risk in those markets? And I guess, particularly with the Tanzania acquisition, is there a model you guys have internally that you use? Just with -- curious how you think about...

We look at all kinds of things. And when you get into markets in Africa, most of the work is not around the assets as much as it is around political, regulatory, tax, government. And we spend a lot of time and taking a lot of resources before we make those decisions.

Okay. That's helpful. And then maybe one last quick one. On the unoccupied towers, the 28,000 as part of the PG&E transaction, is any activity from those sites showing up in your service revenue yet? Or is it still too early for that?

It wouldn't show up in our services revenue. The way that, that would -- that would be leasing revenue. I'm just not -- I know we have some interest. I don't know if we've actually booked anything or not. Do you know, Brendan?

I don't believe so. Not yet.

And our last question comes from the line of Brandon Nispel from KeyBanc Capital Markets.

You mentioned backlog a couple of times, Jeff. Can you quantify the year-over-year change in backlog of signed but not commenced new leases this quarter? And then just more broadly on build-to-suit, one of your peers has large-scale build-to-suit program going on. Can you share with us how many towers you plan on building in Tanzania and maybe how you're thinking about build-to-suit more broadly over the next 2 to 3 years?

Brendan, do we have the answer to the last part?

I don't have the exact number for you, Brandon. It's materially higher, I can say, generally. But we can actually look for something to maybe put a percentage increase or something on that for you after the call.

Yes. And in terms of the new-builds in Tanzania, we're committed to build hundreds of towers over the next 5 years. The hope is we increase that. I would point to South Africa, which we -- which is now, I believe, over 1,500 towers, most of which were built. So there's a lot of opportunity there, and we will continue to press it in those areas where we think it's going to provide great returns.

And remind us what you are generally looking for in terms of returns from like an NOI yield perspective on build-to-suit sites.

Well, we're looking for towers that, ultimately, either with the first, although that's more unlikely, but with the second get to a 15% cash on investment yield or higher. I want to thank everyone for tuning in today. And stay tuned, and we look forward to the next time we're together with third quarter results. Thank you.

And ladies and gentlemen, that does conclude our conference for today. Thank you for your participation and for using AT&T conferencing services. You may now disconnect.