SBA Communications Corporation (NASDAQ: SBAC) — The Best Tower Operator, On Sale Because Its Growth Is in an Air-Pocket

Independent fundamental research. Report date: 2026-06-13. The analysis body carries no recommendation and no price target; the sole exception is the clearly-labeled Author’s Take block below.

⚡ Author’s Take

This block is the author’s own subjective opinion. It is general information, not investment advice. The analysis that follows takes no position and names no price target.

Verdict: BUY / accumulate — quality compounder at the cheapest it has been in a decade, with one honest catch. The constructive mirror of Crown Castle. SBA is the best-run of the three public US tower companies, and it is not close: the highest tower-cash-flow margins (~80% domestic), the highest ROIC (~12.4% vs ~7–8% for CCI/AMT), the lowest dividend payout (~41% of AFFO, leaving the most reinvestment optionality), the only shrinking share count in the group (buybacks, not dilution), the fastest-growing dividend (+13% to $5.00 in 2026), the cleanest capital-allocation record (no activist, no fiber disaster, disciplined M&A, and willing to prune — it sold Canada in 2025), and a best-in-class comp plan keyed to AFFO/share + ROIC. And it trades at ~17x forward AFFO / ~18.5–19x EV/EBITDA — the low end of its own 10-year range (which spent years at 25–37x), at the 7th percentile of its own valuation history.

The catch, and it is real: AFFO/share peaked in FY2024 (~$13.21) and is guided down ~9% to ~$12.06 in 2026. SBA is not compounding right now — it is in an air-pocket driven by Brazilian-real weakness, a 2026 peak in international churn, US Sprint/EchoStar churn, the Canada divestiture, and a refinancing wall that rolls 1.6–2.6% legacy securitized debt into ~5.25% paper through 2027. So you are buying the best operator in the group at a trough multiple while its per-share metric is declining — the cheapness is half-earned. The framing is “great operator, genuine air-pocket, priced for the air-pocket to be permanent.” The upside case is that 2026 is the trough management says it is: US organic re-accelerates to 4–5% in 2027–2029, international churn normalizes, the investment-grade transition lowers the cost of debt, and a low multiple on resuming mid-single-digit-plus AFFO growth — plus a ~2.4% fast-growing dividend and buybacks — compounds nicely from here. My fair-value accumulation zone is ~$205–230 (~17–19x a recovering ~$12.50–13 AFFO), with a genuinely attractive entry in the high-$160s–$180s (near the 52-week low / ~14–15x) where the risk/reward is asymmetric and the rumored ~$250 private-equity take-private interest provides a soft floor. Conviction: medium-high. What flips me more bullish: a 2027 US organic guide confirmed at 4–5%, a successful IG bond at a reasonable spread, or the Brazilian real stabilizing. What flips me bearish: leverage creeping above ~7x into the refi wall, the air-pocket extending (FX + churn) so AFFO/share keeps grinding lower, or an EM/sovereign shock in Brazil. Tag: the sector’s best operator, marked down for a storm management says is already clearing.

1. Executive Summary

SBA Communications is the third-largest of the three public US-listed tower companies and, by most measures of operating and capital-allocation quality, the best-run. It owns and operates 46,328 communications towers as of year-end 2025 — ~17,400 in the United States and ~28,900 internationally across Brazil, Central America, the rest of South America, and Africa — and leases vertical space on those towers to wireless carriers under long-term contracts with ~3% escalators. The model is the same site-monopoly annuity that defines the sector: ~98%+ retention, ~75% gross margins, ~65–68% EBITDA margins, and powerful co-location operating leverage.

What sets SBA apart from peers Crown Castle (CCI) and American Tower (AMT) is discipline and returns, not the asset. SBA earns the highest return on invested capital of the three (~12.4% in 2025, versus ~7% at CCI and ~7.8% at AMT), runs the leanest cost structure, distributes the lowest share of its cash flow (~41% AFFO payout, versus ~90% at CCI and ~93% at AMT), and recycles the retained cash into a combination of accretive build-to-suit towers, disciplined bolt-on M&A (the ~$975M Millicom/Tigo Central America portfolio in 2024), land buyouts, a fast-growing dividend (+13% to $5.00/share in 2026), and consistent buybacks that have shrunk the share count from ~112M (2018) to ~106M today. Its compensation plan is keyed to AFFO/share and ROIC. There is no activist, no accounting controversy, and an orderly 2024 CEO handoff from long-time chief Jeff Stoops to 28-year company veteran Brendan Cavanagh. This is, in short, the capital-allocation antithesis of Crown Castle’s decade of fiber value-destruction.

The honest complication is growth. AFFO/share peaked in FY2024 at ~$13.21 and is guided down to ~$11.84–12.29 (~$12.06 midpoint) for 2026 — a ~9% decline from the peak and a ~1.5%/yr five-year CAGR. The drivers are a confluence of mostly cyclical/transitory headwinds: a weak Brazilian real (guidance assumes ~5.20/USD), a 2026 peak in international lease churn (Oi’s collapse and carrier rationalization in Brazil), US Sprint churn (~$55M in 2026, though far less than CCI’s exposure) plus the EchoStar/DISH default (all its revenue stripped from guidance, now in litigation), the 2025 divestiture of sub-scale Canada, and a refinancing wall that rolls ~1.6–2.6% legacy securitized notes into ~5.25% paper in 2026–2027. Management frames 2026 as the trough and guides US organic growth to re-accelerate to 4–5% across 2027–2029.

Valuation reflects the air-pocket. At ~$205, SBA trades at ~17x forward AFFO and ~18.5–19x EV/EBITDA — roughly in line with AMT (~16x AFFO) and well below CCI (~21x), and at the low end (~7th percentile) of SBA’s own ten-year valuation history, which spent the 2018–2023 period at 22–37x EV/EBITDA. The negative book equity (−$4.8B) is a buyback artifact and renders P/B meaningless; GAAP EPS is flattered by a collapse in reported depreciation (a large tranche of towers became fully depreciated), so AFFO and EV/EBITDA — not P/E — are the right lenses.

The investment question is therefore clean: is 2026 the trough of a temporary air-pocket in an otherwise best-in-class compounder bought at a decade-cheap multiple — or is the AFFO/share decline the start of a structurally lower-growth, FX-and-rate-pressured plateau? The leverage (~6.6x net debt/EBITDA) and the refi wall mean the downside is real if the headwinds persist; the operating quality, the low payout, the buyback engine, and the discounted multiple mean the upside is real if they don’t.

2. Business Overview

What SBA is. SBA Communications is a REIT that owns, operates, and leases communications towers and related infrastructure. Founded in 1989 and based in Boca Raton, Florida, it generates revenue from two segments: site leasing (~98% of segment operating profit) and site development services (~2%, a lower-margin, lumpier, US-only business that does construction, zoning, and installation work for carriers). The investable core is site leasing — renting vertical space on ~46,000 towers to wireless carriers and other tenants under long-term leases with contractual ~3% annual escalators and multiple renewal options.

The domestic/international split — the defining structural feature. At year-end 2025 SBA operated 46,328 towers: ~17,394 in the United States (37.5% of the count) and ~28,934 internationally (62.5%). But the cash flow is inverted: the US generated ~72.6% of site-leasing revenue and the international portfolio only ~23% of segment operating profit. In other words, the majority of SBA’s towers are international, but the large majority of its profit is domestic — US towers earn far more per site (higher rents, more tenants, hard currency, lower churn). International is the growth and diversification leg; domestic is the profit leg. Brazil is by far the largest international market (~30% of total towers, ~12,000+ sites), followed by Guatemala (~10%); no other single market exceeds ~5%. In 2025 SBA exited the Philippines and Colombia entirely and substantially all of Canada, pruning sub-scale geographies — a discipline worth noting. Average tenancy is ~1.8 tenants per tower.

The tenant base. US site-leasing revenue is concentrated in the three national carriers — T-Mobile (~31.1% of total revenue), AT&T (~20.3%), and Verizon (~15.1%), together ~66.5%. That is meaningful concentration, but materially lower than Crown Castle’s ~90%, precisely because the international portfolio diversifies the tenant base (TIM, Vivo, Claro, Telefónica in Brazil; Tigo/Millicom and Claro in Central America). The lower concentration is a genuine, if modest, quality advantage versus the US-only CCI.

The lease contracts and their durability. US leases typically carry 5–10-year initial terms with multiple five-year renewal options at the tenant’s election, fixed ~3% annual escalators, and only narrow termination rights — the contractual scaffolding of a high-visibility annuity. International MLAs are often longer (the Millicom agreements run 15 years) and frequently CPI-linked. Tenant retention runs ~98%, and the practical “churn” the business experiences is rarely a tenant choosing a competitor’s tower (almost never economic) but rather a tenant decommissioning a site after a network merger — the Sprint and Oi cases — which is episodic and identifiable rather than chronic. The combination of long contracts, escalators, near-universal renewals, and the physical impracticality of relocation is what gives the tower model its bond-like cash-flow profile, and SBA’s contracted backlog and ~98% retention place it squarely in that profile.

How revenue grows. The same three levers as the rest of the sector: contractual escalators (~3% in the US, often CPI-linked abroad), amendments (carriers adding equipment to existing leases — high incremental margin), and new colocations, net of churn. Internationally, organic growth is structurally higher (less-mature 4G/5G networks, far lower tower density — Brazil has roughly 4 sites per 10,000 people versus ~16 in the US) but is partly eroded by FX translation and higher churn. SBA also builds towers — thousands per year, predominantly international, often under build-to-suit commitments tied to anchor tenants (e.g., the right to build up to 2,500 sites for Millicom in Central America) — and acquires portfolios opportunistically.

The build-to-suit machine. Unlike a pure landlord, SBA is also a tower developer, constructing thousands of new towers per year — predominantly international — typically under build-to-suit arrangements where an anchor carrier commits to the first lease before the steel goes up (de-risking the build) and SBA retains the colocation upside as additional tenants follow. The Millicom partnership exemplifies this: alongside the ~7,110 acquired towers, SBA holds seven-year exclusivity to build up to 2,500 new sites for Tigo across Central America. New builds are lower-return than mature US co-located towers initially but season over time as tenancy ratios climb — the developer’s version of the co-location flywheel. This is a genuine organic-growth channel that CCI (which builds far fewer towers) largely lacks.

The services segment. Site development services (~2% of segment operating profit) provides construction, zoning, permitting, and equipment-installation work for US carriers. It is low-margin, project-based, and cyclical with carrier deployment activity — useful as a customer-relationship and market-intelligence tool more than as a profit center. It should be valued as a small, volatile adjunct to the leasing annuity, not as a growth driver.

Corporate form and why the numbers read oddly. SBA elected REIT status in 2016. Two accounting features distort the GAAP optics. First, negative book equity (−$4.8B) — the cumulative result of years of buybacks and dividends exceeding retained earnings — which makes P/B and book-based return metrics meaningless (not a solvency signal). Second, reported depreciation has collapsed (from ~$716M in 2023 to ~$292M in 2025) because a large tranche of older towers became fully depreciated, which mechanically flatters GAAP EPS (~$9.80 in 2025). Both argue for valuing SBA on AFFO and EV/EBITDA, not GAAP earnings.

Verdict. A high-quality, recurring, high-margin tower annuity with a distinctive two-legged geography: a hard-currency, high-profit US base and a lower-margin but faster-growing (and FX-exposed) international portfolio that diversifies the tenant base but adds emerging-market risk. The asset is excellent; the geography is the swing variable that distinguishes SBA from its US-only peer.

3. Industry Dynamics

The US base — a three-player oligopoly, late in its leasing cycle. SBA’s domestic business sits in the same structurally attractive US macro-tower market as CCI and AMT: three dominant tower owners, investment-grade carrier customers, high barriers to entry (zoning, site scarcity, anchor-tenant economics), near-zero new-supply risk, ~98%+ retention, and ~3% contractual escalators. As with the rest of the sector, the US is now a stable three-network market (Verizon, AT&T, T-Mobile) after DISH/EchoStar’s greenfield build collapsed into a January 2026 default. Carrier capex has normalized off the 5G mid-band peak, so domestic organic leasing has decelerated toward the low-to-mid-single digits, with the next leg dependent on the C-band continuation, the planned upper-C-band auction (~mid-2027, deploying ~2029–30), and eventually 6G densification. This is the quiet part of the capital cycle — no overbuild to mean-revert, but the demand pulse that drove the last cycle has passed.

The international leg — higher growth, real risk. What separates SBA’s industry exposure from CCI’s is its large international footprint, concentrated in Brazil and Central America. The structural case is genuine: emerging markets have far lower tower density, less-mature 4G/5G coverage, and rising data consumption, so organic lease-up runs higher than in the saturated US. But the risks are equally genuine and distinct:

• Currency. A large share of international revenue is Brazilian-real-denominated; the BRL has been weak and volatile (guidance assumes ~5.20/USD), and FX translation has been a direct drag on reported AFFO/share. SBA’s Central American (Millicom) cash flows are largely USD, which mitigates somewhat, but Brazil is the swing factor.
• Carrier consolidation and churn. Emerging markets are prone to operator consolidation. In Brazil, the breakup and absorption of Oi by the big three (TIM/Vivo/Claro) has driven elevated lease churn, which management says peaks in 2026 and improves thereafter. Central American carrier rationalization (Millicom/Claro) adds churn too.
• Sovereign/inflation/political risk across the LatAm footprint, and the operational complexity of running ~12 markets.

Brazil — the swing market, examined. Brazil deserves specific attention because it is ~30% of SBA’s tower count and the single largest source of both its international growth and its current pain. The structural case is strong: Brazil has roughly a quarter of the US’s tower density per capita, a large and data-hungry population, and three well-capitalized carriers (TIM, Vivo/Telefônica, Claro/América Móvil) deploying 4G and 5G. The complications are equally specific. First, the Oi unwind: Oi, the long-troubled fourth carrier, was carved up and absorbed by the big three, and the resulting network rationalization has driven elevated lease churn that management says peaks in 2026. Second, currency: SBA’s Brazilian revenue is real-denominated, and the BRL’s multi-year weakness (guidance assumes ~5.20/USD) translates directly into lower reported dollar AFFO — a pure translation drag that has nothing to do with operational performance but hits the headline metric all the same. Third, inflation-linked escalators cut both ways — they protect real value but add volatility. The net is that Brazil is simultaneously SBA’s best long-term organic-growth opportunity and the proximate cause of its near-term AFFO/share air-pocket; the two are inseparable, and an investor in SBA is, to a meaningful degree, taking a view on Brazil.

Profit pools. The US profit pool is large, stable, and hard-currency; the international pool is smaller, faster-growing in local terms, but FX- and churn-eroded. SBA’s strategy is to use the cheap securitized US debt and the high-return US base to fund international growth and buybacks — a sensible structure, provided the international risks stay contained.

Why SBA’s mix is defensible despite the FX noise. A skeptic reasonably asks why a US investor should accept Brazil exposure at all when CCI offers pure US towers. The answer is threefold. First, diversification: the international leg lowers SBA’s single-customer concentration (~66.5% big-three versus CCI’s ~90%) and spreads exposure across more carriers and economies. Second, growth runway: US tower lease-up is structurally mature, whereas emerging markets are early in their densification, so the international leg extends SBA’s organic-growth horizon by years. Third, and most importantly, the international business is additive return on top of a hard-currency US core that already generates ~73% of leasing revenue — SBA is not a Brazil play with a US sideline; it is a US-anchored operator with an EM growth option. The FX drag is real and currently painful in the reported numbers, but it is a translation effect on a minority of cash flow, not an impairment of the underlying assets, and it reverses if the real stabilizes. The mix is a defensible strategic choice, not a liability — provided the EM risks stay contained, which is the live question.

Verdict: a structurally good industry, with SBA’s geography a net modest positive offset by a distinct, live risk vector. The diversification lowers customer concentration and adds a faster-growing (if lower-margin) leg, and emerging-market tower penetration is a real multi-decade tailwind. But Brazil FX and emerging-market churn are genuine swing factors that CCI simply does not carry, and they are the proximate reason SBA’s reported growth is currently depressed. The industry is good; the international exposure is the double-edged differentiator.

4. Competitive Position

The moat, named. SBA’s moat is identical in kind to the rest of the sector — local economies of scale plus customer captivity (switching costs) in Greenwald’s taxonomy. Each tower is a local quasi-monopoly: once a carrier integrates and propagation-tests its antennas at a site, relocating is slow, costly, and risks coverage gaps, and a suitable alternative structure often does not exist nearby. Long contracts with escalators and renewal options reinforce the captivity. The moat passes Greenwald’s diagnostics: market share among the operators is extremely stable (towers rarely change hands or tenants), and returns are persistently high.

Why SBA out-earns its peers — the real differentiator. Where SBA distinguishes itself is not the asset but the economics it extracts from the asset. On the key tests it leads the group: ROIC ~12.4% in 2025 (versus ~7% at CCI and ~7.8% at AMT), ~80% domestic tower-cash-flow margin, and a consolidated EBITDA margin of ~65.6% (down from ~68.6% in 2024, but the decline is international/Millicom mix dilution, not core deterioration). The reasons are structural and cultural:

1. The leanest cost base — SBA runs the lowest SG&A in the group, with a decentralized operating model and ~1,720 employees against ~46,000 towers.
2. More owned land and accretive land buyouts (e.g., buying the land under ~3,900 Guatemala sites at ~7x cash flow) — reducing the largest operating cost.
3. Disciplined site selection and willingness to prune — SBA exited the Philippines, Colombia, and Canada in 2025 rather than chase scale for its own sake.
4. Critically, no strategic detour. SBA never built or bought a fiber/small-cell business; it stuck to towers. That avoidance of the exact misadventure that destroyed ~$7–12B of Crown Castle’s capital is, in retrospect, the single largest source of SBA’s relative outperformance.

The co-location flywheel, worked through. The economics are worth making concrete because they explain SBA’s margin leadership. A tower carries largely fixed costs — ground rent (or, increasingly for SBA, owned land), maintenance, taxes, insurance. The first (anchor) tenant covers those costs and a modest return. The second tenant’s rent is almost pure margin — the ground lease and steel are already paid for — and the third more so. SBA’s domestic tower-cash-flow margin of ~80% and consolidated ~65% EBITDA margin are the financial fingerprint of this leverage, and SBA pushes it harder than peers through two levers: a higher domestic tenancy ratio cultivated by disciplined site selection, and aggressive land buyouts (buying the dirt under its towers — e.g. ~3,900 Guatemala sites at ~7x cash flow) that convert a variable ground-rent cost into owned land, permanently lifting site margins. The flip side, as everywhere in towers, is symmetric: churn comes off at near-100% margin, which is why Brazil/Oi and US Sprint churn bite EBITDA disproportionately — and why the 2026 trough is sharper in AFFO than in revenue.

The Greenwald/quality pressure-test — be honest. SBA’s “best operator” badge rests on capital discipline, the lowest payout, and a favorable US mix — not on a structurally superior asset versus AMT or CCI domestically. Its international towers are genuinely lower-quality than its US towers (FX, churn, lower revenue per site), and the consolidated margin is being diluted as lower-margin Central America scales in. The buyback-driven per-share model also runs on a heavily levered, negative-equity balance sheet. So the moat is real and SBA extracts more from it than peers — but the edge is operational and allocative, durable only as long as management stays disciplined and the international bets stay contained.

The Marathon capital-cycle read. Through the supply-side lens, towers are a rare industry where high returns have not drawn in destabilizing new supply — zoning, anchor-tenant economics, and site scarcity prevent it — so the usual mean-reversion mechanism is muted. Where capital did flood in was tower M&A multiples (which peaked alongside the 2020–21 low-rate era) and, in CCI’s case, the adjacent fiber business. SBA’s discipline shows in how it played that cycle: it kept buying at sensible multiples, declined the fiber temptation entirely, pruned sub-scale geographies when returns didn’t justify the complexity, and used the cheap-capital window to term out long-dated low-cost debt and buy back its own stock rather than overpay for trophy assets. The capital cycle is now in its quiet, higher-cost phase — which is precisely when the disciplined operator’s relative advantage compounds and when entry multiples for the patient buyer improve. SBA’s behavior across the cycle is the Marathon ideal; CCI’s fiber binge was the cautionary opposite.

Verdict: a real, second-to-none-in-execution position within a shared structural moat. SBA owns the same kind of local-monopoly assets as its peers but converts them into the best returns in the group through cost discipline, low payout, accretive capital recycling, and the avoidance of value-destructive diversification. It is the highest-quality operator of the three — with the caveat that the advantage is earned through behavior, not bestowed by a uniquely better asset, and that its international leg is structurally lower-quality than its domestic core.

5. Growth History and Forward Opportunities

The track record — strong, then an air-pocket. From 2016 through 2023, SBA was one of the best per-share compounders in the REIT universe, growing AFFO/share at a low-double-digit rate while shrinking the share count. That engine has stalled. AFFO/share ran ~$11.21 (2021) → ~$12.10 (2022) → ~$13.08 (2023) → ~$13.21 (2024, the peak) → guided ~$11.84–12.29 (~$12.06 midpoint) for 2026 — a ~9% decline from peak and a ~1.5%/yr five-year CAGR. This is the single most important fact about the stock today, and it must be confronted head-on: SBA is not currently compounding. It is in a multi-year plateau/decline.

What is driving the air-pocket — mostly cyclical/transitory, but real. Five overlapping headwinds: (1) Brazilian-real weakness and FX translation — the largest single drag, with guidance assuming ~5.20/USD; (2) a 2026 peak in international churn as Oi’s absorption and carrier rationalization in Brazil/Central America wash through; (3) US Sprint churn (~$55–56M in 2026, with <$20M remaining beyond 2026 — notably less than CCI’s exposure); (4) the EchoStar/DISH default, with all of its revenue stripped from 2026 guidance and the dispute in litigation; (5) the 2025 Canada divestiture removing a revenue slug; and (6) the refinancing wall rolling ~1.6–2.6% legacy securitized notes into ~5.25% paper, which raises interest expense and pressures AFFO/share through 2027. Several of these (FX, the refi step-up, Canada) are non-recurring or cyclical; the question is whether they clear on management’s timeline.

The organic detail — domestic versus international. The two legs grow differently. Domestically, organic growth is built from ~3% escalators plus 2–3% of new leasing/amendments minus ~1% normalized churn, netting to a ~4–5% steady-state — but 2026 dips below that range because of a specific, identifiable air-pocket: AT&T’s master-lease structure front-loaded its activity (less incremental in 2026), T-Mobile is in a cyclical post-integration lull, and EchoStar’s revenue was removed entirely; partly offsetting is a newly signed Verizon MLA. Internationally, local-currency organic growth runs higher (often high-single-digit, with CPI-linked escalators in Brazil) but is eroded by FX translation into dollars and by the 2026 churn peak, so reported international growth understates the underlying lease-up. The composite is a 2026 trough that management frames as mechanical and contracted (~80% of near-term organic is already under contract) rather than demand-driven — an important distinction, because a contracted trough is more likely to be temporary than a demand collapse.

The re-acceleration case. Management guides US organic growth to recover to 4–5% across 2027–2029 (~3% escalators + 2–3% lease-up − ~1% normalized churn), with international churn improving after the 2026 peak. Forward drivers: the C-band continuation and the ~mid-2027 upper-C-band auction (deploying ~2029–30), 6G’s uplink-heavy densification, fixed-wireless access (~15M+ subscribers and over half of network capacity, pressuring carriers to densify), Brazil’s 450/700MHz auctions (~2027), early-stage edge-compute/AI-inference colocation at tower sites (immaterial today), the Millicom build-to-suit pipeline (up to 2,500 sites), continued buyback accretion, and the investment-grade transition lowering the cost of debt. None individually is transformative, but in aggregate they support the trough-then-recover thesis.

The Millicom deal — a model of the growth machine. In 2024 SBA acquired ~7,110 towers (~7,266 total) across Guatemala, Honduras, Panama, El Salvador, and Nicaragua for ~$975M — nearly all USD cash flows, 15-year MLAs plus a 15-year extension on ~1,500 existing leases, 7-year build-to-suit exclusivity, and an earn-out. Management reports lease-up is exceeding initial projections. This is the SBA growth template: buy hard-currency-or-contracted cash flows at a reasonable multiple, layer organic lease-up, fund with cheap securitized debt.

Verdict: high-quality growth model, currently in a genuine air-pocket. Unlike CCI — whose near-term per-share progress is essentially financial engineering on a flat, churn-pressured base — SBA’s model (organic lease-up + accretive M&A/land + buybacks + a fast-growing dividend on a low payout) is high-quality and per-share-accretive when it is working. The honest caveat is that it is not working right now: AFFO/share has declined for two years and is guided lower again in 2026. The investment case rests substantially on management’s claim that 2026 is the trough and 4–5% US organic resumes thereafter — a claim that is plausible and partly contracted, but not yet proven.

6. Financial Quality

Read AFFO and EV/EBITDA, not GAAP. Two accounting quirks make GAAP misleading. First, reported depreciation collapsed from ~$716M (2023) to ~$292M (2025) as a large tranche of towers became fully depreciated (Schedule III accumulated depreciation ~$4.36B), mechanically inflating GAAP net income (~$1,054M, EPS ~$9.80 in 2025) and depressing the GAAP P/E to a misleadingly low level. Second, book equity is negative (−$4.8B, −$67.6/share), a pure artifact of cumulative buybacks and dividends, which makes P/B meaningless. AFFO/share and EV/EBITDA are the honest metrics.

The AFFO picture. FY2026 guidance (per the Q4-2025 release): AFFO $1,260–1,308M; AFFO/share $11.84–$12.29 (~$12.06 midpoint); Adjusted EBITDA $1,912–1,932M; net income $774.5–827.5M. At ~$205 that is ~17x forward AFFO — in line with AMT (~16x), well below CCI (~21x), and at the low end of SBA’s own multi-year range. As discussed, AFFO/share is down ~9% from the 2024 peak; the bridge from net income adds back real-estate D&A and amortization, non-cash items, and SBC, less sustaining capex and straight-line adjustments. The trajectory is the crux: AFFO/share ~$11.21 (2021) → ~$12.10 (2022) → ~$13.08 (2023) → ~$13.21 (2024 peak) → ~$12.06 (2026E). A buyer today is paying ~17x a 2026-trough number; if management’s 2027 inflection materializes, the forward multiple on a recovering ~$12.50–13.50 AFFO is closer to ~15–16x, and the “decade-cheap vs own history” framing becomes a genuine entry point rather than a value trap. The entire valuation case turns on which of those two readings is right.

Interest expense — the swing line. Net interest expense was ~$445M in 2025 against ~$1,847M of EBITDA (the ~3.87x coverage). Because SBA’s debt is ~$13B and its EBITDA grows only slowly near-term, the change in interest expense as the cheap legacy ABS refinances is a first-order driver of AFFO/share — arguably the largest single controllable variable. A ~$1.2B tranche moving from ~1.6% to ~5.25% is ~$44M of incremental annual interest, roughly ~$0.40/share of AFFO — material against a ~$12 base. This is why the IG transition and the blended go-forward cost of debt matter so much, and why management’s guidance explicitly embeds the refi assumptions.

Margins — best in class but compressing. Adjusted EBITDA margin was ~65.6% in 2025, down from ~68.6% (2024) and ~67.5% (2023) — still the highest of the three peers, with the decline attributable to the lower-margin Central American (Millicom) mix scaling in, plus EchoStar bad-debt, rather than core erosion. Domestic tower cash-flow margin remains ~80%, the gold standard of the sector.

Leverage and the debt structure — the central financial risk. Net debt/EBITDA was ~6.4x at year-end 2025, ~6.6x at Q1-2026, against a stated target of 6.0–7.0x (revised to that range in late 2024) — operating mid-range, lower than CCI’s ~8.7x-pre-paydown but higher than AMT’s ~5x. Interest coverage is ~3.87x EBITDA/interest — better than CCI’s ~2.95x. The debt is predominantly securitized Tower Revenue Notes (ABS) plus two senior note tranches, a term loan, and a revolver — total ~$13.0B face (~$15.3B including ~$2.4B finance leases), net debt ~$12.7B, cash ~$269M. The capital structure’s great strength is its legacy cost: large ABS tranches locked at 1.6–2.6% in 2020–21. That is also the source of the air-pocket’s financial leg — the refinancing wall: the 2020-1C $750M was repaid in January 2026; the 2021-1C $1,165M at 1.631% has a November 2026 anticipated-repayment date and guidance assumes refinancing at ~5.25% (more than 3x the old coupon); and ~$1.5B of 3.875% senior notes mature in February 2027. Rolling sub-3% debt into ~5.25% is a structural, multi-year headwind to AFFO/share — the single most important financial dynamic to track. The planned 2026 investment-grade transition (an inaugural IG bond) is the mitigant, intended to lower the marginal cost of debt over time.

The securitization model — a structural advantage and its expiry. SBA pioneered financing towers with securitized Tower Revenue Notes — investment-grade-rated ABS backed by the cash flows of a pool of towers, which historically priced well inside what the parent’s unsecured credit would command. Through 2020–21 SBA termed out large tranches at 1.6–2.6%, an extraordinarily cheap, mostly-fixed, long-dated cost of capital that powered a decade of accretive buybacks and M&A. That structural edge is now partially expiring on schedule: the maturity/anticipated-repayment ladder forces refinancing of those sub-3% tranches into a ~5.25% market over 2026–2027. The planned investment-grade transition — issuing an inaugural unsecured IG bond and migrating toward an IG capital structure — is management’s response, intended to broaden funding access and lower the marginal cost over time. The net effect through 2027 is still a rising blended cost of debt; the question is the magnitude, and it is the single most trackable driver of the AFFO/share path.

Cash flow. Operating cash flow was ~$1,291M in 2025; the business is genuinely capital-light to maintain (sustaining capex is a small fraction of revenue), with discretionary capital going to builds, M&A, and land buyouts. FCF comfortably funds the ~41%-payout dividend with ample room left for buybacks and growth — the hallmark of the low-payout model.

Verdict: high-quality economics on a levered, negative-equity balance sheet, currently pressured by FX and a refi step-up. SBA has the best margins and returns in the sector and a well-laddered, mostly-fixed, historically cheap debt stack — but it is meaningfully levered (~6.6x) and faces a genuine multi-year refinancing headwind that, together with Brazil FX, is the proximate cause of the AFFO/share air-pocket. Economics improve with scale here in a way they demonstrably do not at CCI; the caveat is the balance sheet’s sensitivity to rates as the cheap legacy debt rolls off.

7. Capital Allocation

This is SBA’s defining strength — and the sharpest contrast with Crown Castle. Where CCI spent a decade destroying ~$7–12B in a fiber misadventure, SBA has run one of the most disciplined and shareholder-aligned capital-allocation programs in the REIT universe. The model: use cheap securitized debt and the high-return US base to fund (a) high-return build-to-suit towers and disciplined bolt-on M&A, (b) a fast-growing dividend off the lowest payout in the group, and © consistent buybacks — while pruning sub-scale assets.

Capital returned. Over the last five years SBA returned ~$3.65B to shareholders — ~$1.81B of buybacks and ~$1.83B of dividends — while simultaneously funding M&A and builds. Buybacks have run consistently (~$498M in 2025, ~$200M in 2024, ~$100M in 2023, ~$432M in 2022, ~$583M in 2021), shrinking the share count from ~112M (2018) to ~106M today — the only negative net issuance in the big three (versus EQIX-style dilution elsewhere in REIT-land). ~$1.1B of buyback authorization remains. SBC is modest at ~$76M/yr.

Dividend. The dividend has compounded from $1.86 (2020) to $4.47 (2025) to $5.00 annualized in 2026 (~13% raise, ~$1.25/quarter) — yet still represents only ~41% of AFFO, by far the lowest payout of the three (CCI ~90%, AMT ~93%). That low payout is the engine of optionality: it funds buybacks and growth from retained cash rather than relying on external capital, and it leaves enormous room for future dividend growth.

M&A discipline. SBA acquires, but at sensible terms — the Millicom Central America portfolio (~$975M for ~7,110 towers, largely USD cash flows, lease-up ahead of plan), land buyouts at ~7x cash flow — and, crucially, it sells sub-scale assets (Canada, Philippines, Colombia in 2025). This willingness to prune, not just acquire, is the mark of genuine discipline and the opposite of empire-building.

The M&A track record in context. SBA’s acquisition history is one of measured, returns-driven expansion rather than transformational bets. It built its international footprint over a decade of bolt-ons (Brazil via Vivo/Oi/TIM portfolio purchases, Central America via América Móvil and now Millicom), paying multiples that — net of the organic lease-up that follows — have generally been accretive to its high corporate ROIC. The Millicom deal (~$975M for ~7,110 towers, ~$89M of first-year tower cash flow, implying ~11x going-in and lower as lease-up seasons) is representative: a contracted, largely-USD cash-flow stream bought at a reasonable multiple with a build-to-suit pipeline attached. Critically, SBA has shown it will not chase growth at any price — the 2025 exits from the Philippines, Colombia, and Canada removed sub-scale operations that diluted returns and management attention. That two-way discipline (buy when accretive, sell when not) is the behavioral signature that distinguishes a genuine capital allocator from an empire-builder, and it is the single sharpest contrast with Crown Castle’s one-way fiber accumulation.

Compensation — best-in-class alignment. The 2026 proxy shows an annual bonus on Adjusted EBITDA (50%), site-leasing revenue (25%), and individual goals (25%), and — the part that matters — long-term PSUs weighted 60% to AFFO-per-share, 20% to relative TSR, and 20% to ROIC over three years. Tying the majority of long-term pay to per-share AFFO and ROIC directly discourages the scale-for-scale’s-sake empire-building that the absolute-EBITDA metrics elsewhere (including CCI’s recent re-weighting) reward. Hedging and pledging are prohibited. This is the comp design an owner would write.

The per-share arithmetic. The buyback’s value is best seen per-share: shrinking the count from ~112M to ~106M while AFFO holds means each remaining share owns more of the cash flow, and at a ~17x AFFO multiple (a ~5.9% AFFO yield) buying back stock is itself a ~5.9%-return reinvestment that compounds with the dividend and organic growth. This is the engine that drove the 2016–2023 per-share compounding, and it is why the low payout matters — a REIT distributing 90%+ of AFFO (CCI, AMT) has little left to repurchase shares or self-fund growth, whereas SBA’s ~41% payout leaves more than half its AFFO for buybacks, builds, M&A, and deleveraging. The catch in 2026 is that the same cash now competes with a higher cost of debt and the desire to hold leverage in range, so the pace of buybacks may moderate until the refi wall is past.

The one caveat. The strategy runs on ~6.6x leverage and a negative-equity balance sheet, and the buyback-plus-M&A model depends on continued access to cheap debt — which the refi wall is making more expensive. Capital allocation has been excellent, but it is levered excellence, and the cost-of-capital tailwind that powered it (sub-3% securitized debt) is reversing.

Verdict: excellent — among the best in the REIT universe, and the textbook counter-example to Crown Castle. Per-share-focused, return-disciplined, willing to prune, with a comp plan that enforces it. The only asterisk is that the model is leverage-dependent and now faces a higher cost of debt. This is a management team that has earned the benefit of the doubt — the rarest verdict in this framework.

8. Changes and Headwinds — Last Two Years

• CEO transition (Jan 2024). Long-time CEO Jeff Stoops handed the reins to 28-year company veteran and former CFO Brendan Cavanagh, with Marc Montagner becoming CFO and Stoops moving to active Chairman. Orderly, internal, telegraphed a quarter ahead — the opposite of Crown Castle’s four-CEOs-in-two-years chaos.
• The Millicom/Tigo Central America acquisition (2024). ~$975M for ~7,110 towers — SBA’s largest recent M&A, materially expanding the Central American footprint with largely USD cash flows; lease-up is running ahead of plan.
• Portfolio pruning (2025). Exited the Philippines and Colombia entirely and substantially all of Canada — concentrating capital on higher-return markets.
• The AFFO/share air-pocket (2024–2026). The peak-to-trough ~9% decline in AFFO/share, driven by Brazil FX, peak international churn, US Sprint/EchoStar churn, the Canada sale, and the refi step-up (/) — the single biggest change to the near-term thesis.
• The EchoStar/DISH default (Jan 2026). All EchoStar revenue removed from 2026 guidance; SBA terminated, accelerated, and sued. Its exposure is smaller than CCI’s, but it removes a US growth contributor.
• Refinancing wall + IG transition (2026–2027). Rolling ~1.6–2.6% legacy ABS into ~5.25% paper, partly mitigated by a planned move to an investment-grade capital structure to lower the marginal cost of debt.
• Private-equity take-private speculation (2026). TMT Finance reported multiple PE firms studying a take-private at ~$250/share (~22% above current); management was non-committal. A soft optionality floor, not a thesis.

Verdict: the changes are net stabilizing, with one genuine negative. The CEO handoff was clean, the M&A and pruning were disciplined, and the IG transition is sensible balance-sheet management. The genuine negative is the AFFO/share air-pocket — but its causes are largely identifiable, cyclical, and (on management’s telling) troughing in 2026. This is a quality business managing through a cyclical/FX/rate squeeze, not a company in distress or strategic disarray.

9. Risk Analysis

The dominant risks are Brazil FX and the refinancing wall — together the proximate cause of the air-pocket — amplified by leverage. These are the variables that determine whether 2026 is a trough or a plateau. International churn and US deceleration are secondary, cyclical drags. The catastrophic-loss probability is low (irreplaceable assets, investment-grade US tenants, ~98% retention); the disappointing-total-return probability — if FX and the refi step-up persist and AFFO/share keeps grinding — is the real risk.

How the risks interact. SBA’s risks are correlated through the AFFO/share line and the balance sheet. A weaker real lowers reported AFFO and (via lower EBITDA in dollar terms) nudges leverage higher; higher leverage and a higher refinancing cost compress AFFO further; a softer AFFO trajectory both delays the buyback that drives per-share growth and risks the cheap-vs-history multiple proving justified rather than opportunistic. The chain runs FX/rates → AFFO/share → leverage/buyback capacity → multiple. The mitigant is the quality of the underlying assets and the low payout: even in a prolonged air-pocket, the dividend is covered ~2.4x by AFFO, the US cash flows are hard-currency and contracted, retention is ~98%, and the ~$1.1B buyback authorization provides per-share support — so the realistic bear case is dead money (yield with stalled growth), not impairment. As with CCI, the equity is long irreplaceable real assets and short a large, now-repricing fixed-income book; rates and the BRL are the hidden second and third factors alongside leasing fundamentals — but unlike CCI, the operating engine underneath is the best in the sector, which is what makes the air-pocket more likely to be temporary.

The valuation-versus-AMT risk deserves its own note: because SBA is not absolutely cheaper than AMT on forward AFFO, a broad tower-sector de-rate (on rates, or on a structural-demand scare like satellite direct-to-device) would pressure SBA alongside the group regardless of its operating quality — the “cheap vs own history” cushion is relative, not absolute, and does not immunize the stock against a sector-wide multiple reset.

10. Valuation Discussion (Embedded Expectations)

Where it trades (2026-06-13, ~$204.79). Market cap ~$21.7B (~106M shares); enterprise value ~$34.4B on ~$12.7B net debt. On REIT metrics: ~17x forward AFFO (~$12.06 2026E) and ~18.5–19x EV/EBITDA (~$1.92B 2026E EBITDA; 19.4x on 2025 EBITDA); ~2.4% dividend yield (low yield, fast growth, lowest payout). The GAAP P/E (~21x) and P/B (negative) are both meaningless here. On its own ten-year history, SBA sits near the bottom of its valuation range — EV/EBITDA of ~19x against a 2018–2023 band of 22–37x, and the 7th percentile of its own composite valuation history (1.9th percentile on P/E, 12th on P/S). This is the cheapest SBA has been relative to itself in a decade.

The peer comparison.

The table yields a nuanced conclusion. SBA is not dramatically cheaper than AMT on forward AFFO (~17x vs ~16x) or EV/EBITDA (both ~19x) — the two highest-quality tower names trade roughly in line, and both are well below the expensive outlier CCI (~21x). What distinguishes SBA is quality at that multiple: the highest ROIC and margins, the lowest payout (most reinvestment optionality), the only shrinking share count, the best comp alignment, and — the crux — a valuation at the low end of its own decade-long history. AMT carries lower leverage (a real advantage) but a near-maxed payout and slower per-share dynamics; SBA carries more leverage but far more capital-return flexibility. The honest framing: SBA is the best operator in the group, trading roughly in line with the other high-quality name and cheaply versus its own history — but it is not an absolute-screaming-cheap multiple, and the discount-to-own-history is partly earned by the AFFO/share air-pocket.

Embedded expectations / reverse read. At ~$34.4B EV and ~$1.92B 2026E EBITDA growing only modestly near-term, with AFFO/share declining into 2026, a ~17x forward AFFO multiple embeds an expectation that 2026 is the trough and AFFO/share re-accelerates — that US organic returns to 4–5%, international churn normalizes after its 2026 peak, Brazil FX stabilizes, and the IG transition contains the refi cost. If that holds, the math is attractive: a low multiple on resuming mid-single-digit-plus AFFO growth, plus a ~2.4% fast-growing dividend and ~1–1.5%/yr buyback accretion, supports a low-double-digit total return with potential multiple re-rating toward the historical mean. If instead the air-pocket extends — BRL stays weak, the refi step-up bites harder, churn lingers — AFFO/share grinds sideways-to-down and the cheap-vs-history multiple proves justified rather than opportunistic.

The SBA-versus-AMT choice. Because the two highest-quality names trade so close on forward AFFO, the genuine question for a tower investor is which to own. AMT offers lower leverage (~5x), a larger and more globally diversified footprint (including India-exit cleanup and the CoreSite data-center optionality), and a higher current yield (~4%) — but a near-maxed ~93% payout that leaves little room for buybacks, slower per-share dynamics, and its own large EM exposure. SBA offers the higher ROIC, the far lower payout (hence the buyback-driven per-share model and more dividend-growth runway), the cleaner comp alignment, and the cheaper-versus-own-history multiple — at the cost of higher leverage and a more concentrated (Brazil-heavy) international book. The choice is essentially lower-leverage-higher-yield-slower (AMT) versus higher-quality-capital-allocation-cheaper-vs-history-but-more-levered (SBA). For an investor who weights capital-allocation discipline and per-share compounding — and who believes the air-pocket is cyclical — SBA is the more attractive of the two; for one prioritizing balance-sheet safety and current income, AMT. Both are clearly preferable to CCI on price-for-quality.

Asset-value sanity check. Private-market tower transactions clear at high-teens-to-low-20s EV/EBITDA; SBA’s ~18.5–19x public multiple sits at the low end of that range, which is precisely why private-equity interest (the rumored ~$250/share, ~22% above current) is credible — a financial sponsor could underwrite SBA’s hard-currency US cash flows and disciplined model at a premium to the depressed public multiple. The take-private chatter is therefore not idle: it reflects that the public market is valuing the best operator in the group below where private capital would, a tension that either resolves through a bid or through a public re-rating as the air-pocket clears.

Scenarios (return drivers; no price target).

• Bear: the air-pocket extends — Brazil FX deteriorates further, the refi wall lifts interest expense more than guided, international churn doesn’t trough in 2026, and AFFO/share stagnates or declines again in 2027. The multiple holds at ~16–17x (already low), so the damage is muted but the total return is just the ~2.4% yield with little growth — a dead-money outcome, not a collapse.
• Base: 2026 is the trough; US organic recovers toward 4–5%, international churn improves, the IG transition lowers debt cost, AFFO/share resumes ~mid-single-digit growth. Total return ~= ~2.4% yield + ~5–7% AFFO/share growth + modest buyback accretion ≈ ~low-double-digits, with optionality on a re-rating toward the historical mean.
• Bull: the trough proves shallow and the recovery faster — Brazil FX rebounds, the 2027+ spectrum cycle drives US organic to the high end, the multiple re-rates toward the historical ~22–25x EV/EBITDA, and/or a private-equity take-private materializes near the rumored ~$250. Mid-teens-plus total return.

Verdict: the best operator in the group at the low end of its own valuation range, with a genuine but largely cyclical growth air-pocket. Unlike CCI — where the premium multiple caps the upside on the weakest franchise — SBA offers a high-quality franchise at a reasonable (and own-history-cheap) multiple, with the downside cushioned by the low multiple, the buyback, and the take-private floor, and the upside levered to a growth recovery management says is already underway. The distribution is asymmetric to the upside if 2026 is the trough.

11. Variant Perception

Consensus. The Street views SBA as a high-quality but currently challenged tower REIT: best-in-class operator and capital allocator, but with reported growth depressed by Brazil FX, international churn, and the refi step-up, and carrying more leverage than AMT. The debate is squarely about when AFFO/share re-accelerates, and the stock has de-rated to the low end of its history while the market waits for proof.

The strongest bull case. SBA is the best-run tower company — highest ROIC and margins, lowest payout, only shrinking share count, fastest-growing dividend, cleanest capital allocation, best comp alignment — trading at the cheapest multiple relative to its own history in a decade, at a point where the headwinds depressing AFFO/share (FX, refi, churn) are largely cyclical and management says 2026 is the trough. Buy the best operator in an air-pocket at a trough multiple, collect a fast-growing (if small) dividend and buyback accretion, and let the recovery plus a re-rating compound — with a rumored ~$250 PE take-private as a soft floor.

The strongest bear case. The “cheap vs own history” multiple is earned, not opportunistic: AFFO/share has fallen ~9% from its 2024 peak and may keep grinding if Brazil FX stays weak, the refi wall lifts interest expense, and international churn lingers past 2026. The model runs on ~6.6x leverage and negative equity as its cheapest debt rolls off into ~5.25% paper, and on forward AFFO SBA is no cheaper than the lower-levered AMT. If the re-acceleration slips to 2028+, this is dead money — a great operator whose per-share metric simply isn’t growing, on a balance sheet that gets riskier as rates bite.

The 3–5 assumptions that matter most. (1) Is 2026 the trough, or does the AFFO/share decline extend into 2027? (2) Does Brazil FX (BRL ~5.20 assumed) stabilize or deteriorate? (3) How much does the refi wall actually lift the cost of debt, and does the IG transition offset it? (4) Does US organic genuinely recover to 4–5% in 2027–2029? (5) Does the multiple re-rate from the low end of its history, or is the de-rate permanent?

What would falsify each side. Bull falsified by: a 2027 US organic guide below 4%; the BRL weakening materially past ~5.50; leverage drifting above ~7x; AFFO/share guided down again for 2027. Bear falsified by: a confirmed 4–5% 2027 US organic guide; a successful IG bond at a tight spread; BRL stabilization; or a take-private bid near ~$250.

12. Fact vs. Interpretation

13. Open Questions

1. Is 2026 truly the trough? The entire constructive case hinges on AFFO/share inflecting in 2027 — is the 4–5% US organic re-acceleration real and contracted, or aspirational?
2. Where does the Brazilian real go? FX is the largest single driver of the air-pocket; a sustained move past ~5.50 would deepen and extend it.
3. How much does the refi wall actually cost, net of the IG transition? The spread on the inaugural IG bond and the blended cost of debt post-2027 are decisive for AFFO/share.
4. Does international churn genuinely peak in 2026? Brazil/Central America consolidation churn is a management forecast, not a fact.
5. Is the take-private speculation real? A ~$250 bid would crystallize value; its absence leaves the stock dependent on the operating recovery.
6. Can SBA delever while still buying back stock and acquiring, or does the higher cost of debt force a choice?

14. What Must Be True

Bull case — what must be true:

• 2026 is the trough; US organic re-accelerates to 4–5% in 2027–2029 and international churn normalizes after its 2026 peak.
• Brazil FX stabilizes (or improves) and the IG transition contains the refinancing step-up, so AFFO/share resumes growth.
• The buyback + low payout + fast-growing dividend continue to compound per-share value; the multiple re-rates from the low end of its history.
• Falsification test: a 2027 AFFO/share guide flat-to-down, a BRL break past ~5.50, or leverage above ~7x would break the bull case.

Bear case — what must be true:

• The air-pocket extends — FX, the refi step-up, and churn keep AFFO/share grinding sideways-to-down through 2027.
• On forward AFFO, SBA is no cheaper than lower-levered AMT, so the “cheap” multiple is earned and doesn’t re-rate.
• Leverage and the negative-equity balance sheet amplify any cash-flow or rate disappointment.
• Falsification test: a confirmed 4–5% 2027 US organic guide, a tight-spread IG bond, BRL stabilization, or a ~$250 take-private bid would break the bear case.

15. Source Appendix

See SBAC_source_appendix.md for the full source list. Primary sources: SBA Communications FY2025 Form 10-K and Q1-2026 Form 10-Q (SEC EDGAR, CIK 0001034054); 2026 DEF 14A proxy; Q4-2025 and Q1-2026 earnings releases and 2026 guidance (8-K exhibits); Q1-2026, Q4-2025, and Q2-2025 earnings-call transcripts; the Millicom/Tigo Central America acquisition disclosures; Form 4 insider-transaction corpus (CIK 0001034054); American Tower and Crown Castle Q1-2026 results and Crown Castle’s FY2025/Q1-2026 results for peer comparison; and quantitative cross-checks via third-party financial-data aggregators. All figures reconciled to filings where possible; management commentary is treated as hypothesis and validated against filings and external evidence.

APPENDIX A — Standard Diligence Questionnaire

SBA Communications Corporation (NASDAQ: SBAC) — supplemental to the research memo. Report date: 2026-06-13.

Answers are grounded in the underlying analysis; Fact/Interpretation/Assumption labels applied where it matters.

General

What thoughtful questions have other investors asked about this company? The dominant questions: (1) Is 2026 the trough of the AFFO/share air-pocket, or does the decline extend into 2027? (2) Where does the Brazilian real go, and how much is reported AFFO an FX story? (3) How much does the refinancing wall (1.6–2.6% legacy debt → ~5.25%) cost net of the investment-grade transition? (4) Is SBA cheap, or just cheap versus its own (formerly elevated) history — and is it actually cheaper than AMT? (5) Is the rumored ~$250 private-equity take-private real? (6) Can SBA delever while continuing to buy back stock?

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? Interpretation: AFFO/share is at a cyclical low — it peaked in FY2024 (~$13.21) and is guided down ~9% to ~$12.06 in 2026, the trough of an air-pocket driven by FX, churn, and the refi step-up. The underlying assets are not impaired; the per-share metric is cyclically depressed.

Driven by the external environment or internal actions? Predominantly external (Brazilian real, the higher-rate refinancing market, carrier-consolidation churn) plus one internal pruning decision (the Canada divestiture). The operating model itself is intact and high-performing.

How stable are revenues? Very stable at the contract level — ~98% retention, long MLAs, ~3% (US) / CPI-linked (intl) escalators. Reported dollar revenue is destabilized only by FX translation and episodic merger-driven churn, not by demand volatility.

Outlook for products/services? Site leasing is a durable annuity; management guides US organic growth to recover to 4–5% in 2027–2029 after a sub-4% 2026 trough, with faster (FX-eroded) international growth.

How big is this market — growing, shrinking, domestic/international? US towers: mature, slow-growing, three-carrier. International (Brazil, Central/South America, Africa): less mature, faster-growing in local terms, FX-exposed. SBA is the most internationally exposed of the big three by tower count.

Business Quality & Competitive Moat

Is the industry getting more or less competitive? Structurally stable (3-player US oligopoly; concentrated tower ownership abroad), with no new-supply threat; competition is for incremental carrier dollars, intensifying as the US leasing cycle matures.

How profitable is the business (ROIC, ROE)? Fact: ROIC ~12.4% (2025) — the highest of the three peers. ROE is not meaningful on negative equity. Domestic tower cash-flow margin ~80%; consolidated EBITDA margin ~65.6%.

How profitable is the industry — competitors, barriers? Highly profitable (65%+ EBITDA margins across all three); high barriers (zoning, site scarcity, anchor tenant). SBA earns the best returns of the three through cost discipline and capital allocation.

Can the business be easily understood? Yes — a tower-leasing annuity with a US profit core and an international growth leg.

Can it be undermined by foreign low-cost labor? No — physical real-estate assets; labor is minor.

Do brands matter? No — location, reliability, and price drive tower selection.

Nature of competition? Location, footprint, cost-to-lease, and execution; structural differentiation among the three operators is limited, so SBA’s edge is operational/allocative.

Customers’ switching costs? High — relocating integrated, propagation-tested antennas is costly and risky, and alternatives are often unavailable nearby. The core of the moat.

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? The franchise value of ~46,000 site-monopoly towers and the contracted lease backlog far exceed book value (negative). A large tower tranche is fully depreciated, so book understates economic asset value.

Off-balance-sheet liabilities? Ground-lease commitments are substantially capitalized (~$2.4B finance leases); standard long-dated operating-lease and purchase commitments exist.

How conservative is the accounting? Reasonable, with two flags: reported depreciation has collapsed as towers fully depreciate (flattering GAAP EPS — use AFFO), and ~$174M of asset impairments ran through 2025. The AFFO definition is standard for the sector.

How CapEx-hungry is the business? Capital-light to maintain (sustaining capex a small fraction of revenue); discretionary capital goes to builds, M&A, and land buyouts, all return-driven.

Capital Allocation & Management

How much FCF, and how is it used? Strong FCF (~$1.29B OCF 2025). Uses, in priority: a low-payout (~41% AFFO) but fast-growing dividend, consistent buybacks, accretive builds/M&A, land buyouts, and deleveraging — the best-balanced capital-return model of the three.

Philosophy? Per-share-focused, returns-disciplined, willing to prune (sold Canada/Philippines/Colombia in 2025). Funded historically by cheap securitized debt; now managing a higher cost of debt via an IG transition.

Significant acquisitions recently? The 2024 Millicom/Tigo Central America portfolio (~$975M, ~7,110 towers, largely USD cash flows, lease-up ahead of plan).

Buying back shares? Yes, consistently — ~$1.8B over five years; share count shrank ~112M (2018) → ~106M (2025). ~$1.1B authorization remains.

Issuing stock to insiders? No — SBC modest (~$76M/yr).

Compensation policy? Best-in-class: long-term PSUs weighted 60% AFFO/share, 20% relative TSR, 20% ROIC; annual bonus on Adjusted EBITDA / site-leasing revenue / individual. Per-share and ROIC keyed — discourages empire-building. Hedging/pledging prohibited.

Motivations of management? Long-tenured, internally promoted team (CEO Brendan Cavanagh, 28-year veteran/ex-CFO; Chairman Jeff Stoops). No code-P open-market purchases and <1% insider ownership (a neutral signal), but the comp design enforces shareholder alignment.

Valuation & Market Data

ADR, MLP, or K-1 issuer? None — US C-corp taxed as a REIT; issues a 1099-DIV.

Dividend policy? $5.00/yr annualized in 2026 (~13% raise), ~2.4% yield, ~41% of AFFO — the lowest payout of the three, with the most room for future growth.

How profitable is the business? Highest margins and ROIC of the three; FCF strongly positive.

Is net income diverging from cash from operations? Yes — GAAP net income is distorted by the depreciation collapse and FX/impairment items; OCF and AFFO are the honest measures. Use AFFO, not P/E.

Risks & Downside

What would cause the stock to decline? A persistent AFFO/share air-pocket (Brazil FX, refi cost, lingering churn); leverage stress; a broad tower-sector multiple de-rate (rates, satellite-substitution scare) that pressures SBA alongside peers despite its quality.

Risk of catastrophic loss? Low — irreplaceable assets, investment-grade US tenants, ~98% retention. Downside is to the multiple, the AFFO trajectory, and (in an extreme) the dividend’s growth rate — not the franchise’s existence.

Chance of total loss? Very low. The leverage means equity holders bear amplified cash-flow/rate risk, but the asset base is durable and cash-generative.

Recent News & Events

Has the business environment changed recently? Yes: a 2024 CEO transition (Stoops → Cavanagh); the 2024 Millicom Central America acquisition; 2025 exits from the Philippines, Colombia, and Canada; the EchoStar/DISH default (Jan 2026, revenue removed from guidance, in litigation); a 2026 refinancing wall + planned IG transition; and rumored private-equity take-private interest (~$250/share).

Significant acquisitions? Millicom/Tigo Central America (2024). Plus ongoing build-to-suit and land buyouts.

Change in accounting policies? No material change; the depreciation decline reflects assets becoming fully depreciated, not a policy change.

Recent changes — new markets, facilities, management? Expanded Central America (Millicom); exited sub-scale markets (Philippines/Colombia/Canada); orderly management transition; IG capital-structure transition underway.

APPENDIX B — Source Appendix

SBA Communications Corporation (NASDAQ: SBAC). Report date: 2026-06-13. Primary sources prioritized over secondary; management commentary treated as hypothesis and validated against filings and external evidence.

Primary — SEC filings (EDGAR, CIK 0001034054)

• SBA Communications FY2025 Form 10-K (filed 2026-02-27) — segment detail (domestic vs international), tower counts (46,328 total; ~17,394 US / ~28,934 intl), tenant concentration (T-Mobile 31.1% / AT&T 20.3% / Verizon 15.1%), international footprint and 2025 exits (Philippines, Colombia, Canada), churn disclosure, debt schedule (securitized Tower Revenue Notes, senior notes, maturities/ARDs and coupons), Schedule III. https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001034054&type=10-K
• SBA Communications Q1-2026 Form 10-Q — Q1 organic growth, FX, leverage, EchoStar litigation status, balance sheet. https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001034054&type=10-Q
• SBA 2026 DEF 14A (proxy statement) — compensation design (LTI PSUs 60% AFFO/share, 20% relative TSR, 20% ROIC; annual bonus on Adj EBITDA / site-leasing revenue / individual), board, hedging/pledging prohibition, security ownership.
• Q4-2025 earnings release (8-K EX-99.1) — FY2026 guidance: AFFO $1,260–1,308M, AFFO/share $11.84–$12.29, Adj EBITDA $1,912–1,932M, net income $774.5–827.5M; BRL and refinancing assumptions; dividend.
• Q1-2026 earnings release (8-K EX-99.1) — raised AFFO/share guidance, FX update, organic-growth commentary. https://www.sec.gov/Archives/edgar/data/0001034054/000119312526191545/d49075dex991.htm
• 8-K material-event record — CEO transition (Stoops → Cavanagh, eff. Jan 1, 2024; 8-K dated 2023-09-13), debt issuance/refinancing, dividend increases, M&A.
• Form 4 insider-transaction corpus (CIK 0001034054) — transaction-code tally since Jan 2024 (grants/exercises/withholding dominate; zero code-P open-market purchases; <1% insider ownership).

Primary — Company disclosures and transcripts

• Earnings-call transcripts (public): Q1-2026 (2026-04-29), Q4-2025 (2026-02-26), Q2-2025 (2025-08-04) — management framing of the AFFO/share trough (“2026 is the bottom”), domestic vs international organic growth, Brazil FX, the 2026 peak in international churn, Sprint/EchoStar churn, the refinancing wall and IG transition, buyback and dividend policy, and the Millicom integration.
• Millicom/Tigo Central America partnership disclosures (Oct 2024) — ~$975M for ~7,110 towers, 15-year MLAs, build-to-suit exclusivity. https://www.globenewswire.com/news-release/2024/10/28/2970332/0/en/Millicom-Tigo-and-SBA-Communications-agree-to-long-term-partnership-in-Central-America.html

Primary — Peer comparables

• American Tower (AMT) Q1-2026 results — AFFO/share (~$10.99), organic growth, EBITDA margin, leverage (~5x), dividend (~$7.16). https://www.sec.gov/Archives/edgar/data/0001053507/000105350726000094/pressreleaseq12026.htm
• Crown Castle (CCI) — FY2025/Q1-2026 results and SEC filings, for the three-way peer comparison and the US tower-industry framing.

Secondary — Trade press and financial media

• Total Telecom, “SBA Communications buys 7,000 Millicom sites for $975m,” 2024. https://totaltele.com/sba-communications-buys-7000-millicom-sites-for-975m/
• TMT Finance — reporting on private-equity take-private interest in SBA (~$250/share). (Speculative; treated as optionality, not thesis.)
• Trade-press coverage of Brazil tower-market dynamics, the Oi unwind, and the EchoStar/DISH default.

Quantitative cross-checks

• Third-party financial-data aggregators — income statement, balance sheet, cash flow, profitability/credit ratios, enterprise value, and valuation multiples for SBAC, AMT, and CCI (reconciled to filings; SEC filings remain primary).
• Public market-data sources — live prices, market cap, share count, and quick comps.
• Own-history valuation percentiles (third-party): P/E 1.9th, P/S 12.3th, composite 7.1th (P/B null on negative equity) — the “cheap vs own decade” tell.

Note on EV reconciliation: ROIC’s headline SBAC market cap uses a stale low share count; the report uses the current ~106.06M shares × $204.79 ≈ $21.7B market cap and ≈ $34.4B EV on ~$12.7B net debt.