AMT — American Tower Corporation

Sell-side sentiment is genuinely constructive: 40 Buy ratings, 9 Holds, zero Sells, a "Buy" consensus, and an average price target of $209.89 (recent-average $205.43, low $188, high $240) versus the $177.85 price as of 2026-09-13 - roughly 18% implied upside. That was reinforced by the July 28, 2026 Q2 print (revenue +4.6% YoY, EPS $1.86 vs. ~$1.56 expected), a second consecutive guidance raise (property revenue up ~$110M, CoreSite growth lifted to 15% from 13%), completion of the Philippines and Bangladesh exits, and a $1.6B senior notes pricing. But the tape does not confirm the enthusiasm: -4.98% over three months, -3.56% over six months, and a +0.17% position versus the 200-day SMA as of 2026-09-13. That gap between a bullish analyst community and a stock that has gone nowhere for half a year is why sentiment reads Neutral rather than Bullish - the market is waiting for the 2027 AFFO inflection before paying up.

Revenue is stable but decelerating. After negative YoY comps through mid-2025 caused by the India/APAC divestitures (Q4'24 2547.6M vs Q4'23 2786.7M, -8.6%; Q1'25 2562.8M vs Q1'24 2834.1M, -9.6%), growth turned positive and then began to fade: Q3'25 +7.7% YoY, Q4'25 +7.5%, Q1'26 +6.8%, Q2'26 +4.65% (2749.1M vs 2626.9M). Sequentially the top line has been almost frozen at ~$2.74B for three straight quarters (2738.0M, 2737.5M, 2749.1M) - the classic steady-but-unexciting tower lease model. GAAP EPS is noisy and should not be read straight: Q3'24 printed -$1.69 (a $792M net loss from impairments) and Q4'24 printed +$2.62 (asset sale gains), while the last four quarters have been remarkably clean and consistent at $1.82, $1.75, $1.84 and $1.86. TTM operating margin actually contracted modestly, from 45.8% to 44.4%, and TTM free cash flow rose ~4.6% YoY to $3.96B (Q2'26 alone was $1.166B on $320.9M of capex). The one unambiguous balance-sheet trend is positive: long-term debt fell from $35.19B at 2025-06-30 to $31.96B at 2026-06-30, a $3.2B paydown. Verdict: not deteriorating, not accelerating - a genuine trough year with a flat top line, stable margins and improving cash flow, which is precisely what management has guided.

Fair, not cheap. As of 2026-09-13 the stock carried a 24.43x trailing P/E and a 25.5x forward P/E ($177.85 / next-FY consensus EPS of $6.97) - note the forward estimate is BELOW the $7.25 trailing EPS, so the market is explicitly pricing an earnings decline, not growth. PEG of 0.149 is a statistical artifact of a REIT's distorted GAAP EPS and should be ignored. Versus its own history the stock does look discounted - P/S of 7.57 vs. a 5-year average of 10.10, and P/E of 24.43 vs. a 5-year average of 49.02 - which is what drives the deterministic score's "Cheap vs. its own history" label (90.6). But that 49x average is inflated by impairment-depressed earnings years and should not be treated as a fair anchor. The cleaner read is EV/EBITDA of 17.96x against ~$43.2B of net debt, plus a 4.8% trailing FCF yield and a 3.73% dividend yield that consumes roughly 78% of trailing free cash flow. P/B of 22.3x and ROE of 69.3% are both meaningless here - equity is only $3.72B against $63.3B of assets - and ROIC of 6.7% is the number that actually matters, and it sits at or below a reasonable cost of capital. Fair value, with the discount justified by leverage and the trough, not a bargain.

The setup is a classic trough-year mismatch: the market is paying ~25x forward earnings for a year in which consensus EPS declines from $7.25 to $6.97, meaning it has already conceded 2026 and is refusing to pay for 2027. That is a double-edged read. If the guided AFFO reacceleration arrives - DISH churn laps, CoreSite compounds at 15%, refinancing costs stabilize and rates ease - the multiple re-rates down against recovering earnings and the stock works from both directions, which is the mechanism behind the $205-$240 analyst targets. If it does not, you own a levered, 4-5% organic grower at 25x trough EPS with a 78% FCF payout and no multiple cushion. Watch the estimate revisions on 2027 more than the 2026 prints: the Q2 beat ($1.86 vs. ~$1.56) was a real positive surprise, but it beat a bar the market had already set low. Buying here is a bet on the inflection showing up in guidance, not a bet on the current earnings power.

Range-bound with a mild short-term uptick. As of 2026-09-13, RSI(14) was 55.89 - neutral, nowhere near overbought or oversold - and the price sat 1.35% above the 20-day SMA (~$175.5), 3.59% above the 50-day (~$171.7), and just 0.17% above the 200-day (~$177.6). Reclaiming the 200-day is constructive, but the medium-term context is weak: -4.98% over three months and -3.56% over six months, with the stock at roughly the 40th percentile of its $160.06-$199.13 52-week range. The practical read is that $177.6 (the 200-day) is the exact line in the sand - holding above it keeps the recovery alive toward $190-$195; losing it re-opens $165-$160. There is no urgency to chase at 55.89 RSI, and a pullback toward $170-$172 (the 50-day) would offer a materially better entry with the same thesis.

• Interest rates and the Fed path: AMT carries roughly $45B of total debt against $3.72B of equity (D/E of 12.09x) and a 0.352 current ratio, so every basis point of refinancing cost flows straight into AFFO. A genuine easing cycle is the single largest mechanical tailwind; a 'higher for longer' regime is the largest headwind.
• US carrier capex and 5G densification: organic tower growth depends on the big three carriers plus DISH continuing to add/amend leases. Carrier capex has been subdued post the initial 5G build, and DISH's well-documented troubles are a specific, quantifiable churn drag that management has flagged as running through 2026.
• Data center / AI demand: CoreSite is the growth engine, with management lifting 2026 data center revenue growth guidance to 15% from 13% on record leasing. AI and low-latency edge workloads are a genuine secular tailwind, but CoreSite is still a small slice of total revenue.
• REIT sector rotation and rate sensitivity: as a long-duration, dividend-paying, capital-intensive REIT, AMT is bid up when rates fall and growth is scarce, and sold when rates rise or when mega-cap AI/tech trades soak up all the equity inflows.
• FX and geographic mix: with the APAC exits (Philippines, Bangladesh) now complete, revenue is more concentrated in the US and Europe, but a strong dollar still compresses reported international revenue, and further portfolio pruning dilutes headline growth even when it improves returns on capital.

• Q3 2026 earnings on October 27, 2026 - the key test of whether the second consecutive guidance raise holds and whether leverage/refinancing commentary improves.
• 2027 guidance and the AFFO inflection: management has explicitly framed 2026 as a trough year. Confirmation of reaccelerated AFFO growth into 2027 is the thesis-defining catalyst.
• DISH Network churn lapping and any improvement in US carrier leasing/amendment activity as 5G densification resumes.
• CoreSite data center leasing momentum and whether the 15% growth guidance is raised again on AI/edge demand.
• Continued deleveraging (long-term debt already down $3.2B YoY to $31.96B) toward target leverage, plus any Fed easing that lowers refinancing costs.
• Further portfolio simplification / asset sales that recycle capital into US towers and data centers.

• Leverage is the dominant risk: ~$45B total debt, D/E of 12.09x, long-term debt/equity of 8.59x, and a current ratio of just 0.352. This is a business that must continuously access credit markets, and an adverse rate or spread move hits AFFO directly.
• ROIC of 6.7% sits at or below a reasonable cost of capital, raising the question of whether incremental capital is actually creating shareholder value - the deterministic score's quality pillar (56.9) is anchored on this.
• Revenue growth decelerating to ~4.65% YoY in Q2 2026 from ~7.7% in Q3 2025, with sequential revenue flat at ~$2.74B for three quarters - if organic growth keeps fading, the 2027 'inflection' never arrives.
• Dividend coverage: ~$3.09B of annual dividends against $3.96B of trailing free cash flow is a ~78% payout, leaving limited cushion if AFFO disappoints or capex needs rise.
• REIT-GAAP distortion cuts both ways: tiny book equity ($3.72B) and heavy depreciation make P/B (22.3x) and ROE (69.3%) uninformative, and one-off impairments/gains have whipsawed reported EPS (Q3'24 -$1.69, Q4'24 +$2.62), making headline earnings an unreliable guide.
• Tenant concentration and US carrier consolidation: a merger among major carriers could reduce future amendment/co-location revenue and trigger churn.

At $177.85 as of 2026-09-13, AMT is a genuinely high-quality asset - irreplaceable tower infrastructure, 73% gross margins, a 3.73% dividend, and a data center segment compounding at 15% - but it is not an attractive buy here. The fundamentals are weak on any trailing basis: revenue growth has decelerated to 4.65% YoY, TTM operating margin slipped from 45.8% to 44.4%, ROIC of 6.7% is at or below cost of capital, ROE and P/B are meaningless due to a $3.72B equity base, and the current ratio of 0.352 against ~$45B of debt makes this a structurally rate-sensitive, refinancing-dependent equity. The stock is cheap versus its own inflated 5-year history (P/S 7.57 vs. 10.10; P/E 24.4 vs. 49.0) but only fairly priced against its actual risk profile. The bull case requires the 2027 AFFO inflection to actually land, and the market has already set a low bar for 2026 (forward EPS of $6.97 is below the $7.25 trailing figure). That is a reasonable bet but not a compelling one at 25x trough earnings with no technical confirmation and the 200-day moving average serving as the dividing line. My call: not a buy at this price; wait for either a pullback toward the $170-$172 area or hard evidence of the 2027 inflection in guidance before committing capital.

Primary method: forward EV/EBITDA. I applied 18.0x to FY2027E EBITDA of roughly $7.30B (implied TTM EBITDA of ~$7.02B derived from the 17.96x EV/EBITDA on a $126.1B EV, grown ~4%), giving an enterprise value of ~$131.4B; subtracting ~$43.2B of net debt (total debt ~$45.0B implied by the 12.092x D/E times $3.72B equity, less $1.76B cash) leaves ~$88.2B of equity, or ~$189/share on 466M shares. Cross-checks: 18x FY2027E AFFO/share of ~$11.20 gives ~$202, and a Gordon-growth dividend discount ($6.63 dividend, 8.75% cost of equity, 4.75% growth) gives ~$178. Blending the three (189 / 202 / 178) lands at ~$195, a ~10% premium to the $177.85 price as of 2026-09-13 and below the $205-$210 analyst consensus - the gap reflecting my higher weight on leverage and the risk that the 2027 AFFO inflection slips.