When: FOMC meetings or Polymarket 10Y probability shift

Then: Reassess trim architecture

REDUCE exposure; trim 30-50% of existing position; cap retained position at 25% of normal specialty REIT allocation; hard stop at 173.15 on residual.

• Death cross convergence is a bear regime continuation signal, not a reversal: 50 SMA at 171.09 and 200 SMA at 173.15, both declining (200 SMA rolled from 176.97 in June to 173.15; 50 SMA from 177.06 in July to 171.09), with price 4.8% below the 200 SMA — a 1.2% gap between MAs signals compression ahead of trend continuation, not reversal
• MACD bullish divergence is a condition not a signal: RSI bounced from 31.5 to 41.6 but 41.6 is still BELOW 50, marking a dead-cat bounce off oversold; divergences fail more than half the time in sustained downtrends and mark counter-trend bounces that get sold into
• AI re-rating thesis rejected by market: AMT down 18% YTD despite 18 months of AI capex narrative driving Nvidia from $400 to $1000+; Data Center REIT cohort (Digital Realty, Equinix) has re-rated materially while AMT has been actively rejected as an AI vehicle
• Yield peak not confirmed: Polymarket shows 43% probability of 10Y testing 5.5% (down 7.9 percentage points), still nearly a coin flip; 8% probability of hitting 6.0%; Fed cut odds December 3%, October 1%; Fed officials publicly stating inflation still too high — this is a 'higher for longer' environment with marginal softening only
• FCF coverage of 0.90x is unsustainable in tightening credit: 200-300 million quarterly gap funded by debt issuance; in tightening credit environment (like 2022-2023), AMT faces binary choice of dividend cut, frozen growth capex, or balance sheet deterioration; 6% earnings hit from 50 basis points is not manageable at 23.7x forward P/E with 0.90x coverage
• Light sentiment is not a coiled spring: 52-week low at $160.06 versus current price $164.78 provides only 3% buffer before fresh lows; no sentiment floor to catch falling knife; light positioning tells you positioning won't amplify a move, not which direction the next move goes

→ vs aggressive: Death cross convergence signals bear regime continuation, not reversal — both MAs declining, price below both, compression signals continuation of dominant downtrend, not breakout

→ vs aggressive: MACD bullish divergence is a condition not a signal — RSI at 41.6 is still below 50, marking dead-cat bounce not momentum confirmation; divergences fail more than half the time in sustained downtrends

→ vs aggressive: AI re-rating already priced — 18 months of AI capex narrative, AMT down 18% YTD while Data Center REIT peers re-rated; market has actively rejected AMT as AI vehicle

→ vs aggressive: 43% probability of 10Y at 5.5% is still nearly a coin flip; Polymarket shows 8% odds of 6.0%, Fed cuts priced at 3% December/1% October; 'higher for longer' not yield peak

→ vs aggressive: FCF coverage of 0.90x in tightening credit environment is fundamentally different risk than in falling-rate environment; 6% earnings impact from 50bps not manageable at current leverage

→ vs aggressive: Light positioning tells you positioning won't amplify, not which way next move goes — 3% buffer to 52-week low is not a floor, it's a cliff

Trim approximately 20% (not 30-50%), restructure stop at 173.15 from exit trigger to re-entry trigger, maintain residual position through Q3 earnings with 159 as downside invalidation, consider put spread hedge at 160-155 strikes as alternative to aggressive trim

• Polymarket 43% probability of 10Y hitting 5.5% is nearly a coin flip, not a yield peak signal—the aggressive analyst cherry-picks direction while ignoring absolute level
• Light positioning in high-rate environment with confirmed downtrend indicates no bid, not coiled spring for upside
• MACD bullish divergences in sustained downtrends fail more often than succeed but don't always fail
• Bear case more compelling than bull case directionally, but base case is range-bound chop between 159 and 175, not directional move to extremes
• AMT's 0.90x FCF coverage is structurally stable, not fragile—consistent operating cash flow over years supports this
• Conservative analyst's downside to 150 doesn't account for support zone at 159-161 (lower Bollinger band and early October low)
• 30-50% trim locks in losses near lower range while surrendering participation in relief rally to 173-176
• Stop at 173.15 on sell position functions as contrarian re-entry trigger if price closes above, confirming 200 SMA has held
• Q3 earnings in late October is the binary catalyst for re-evaluation
• 159 is the structural line in the sand—daily close below on volume warrants full exit
• Put spread at 160-155 strikes expiring after Q3 earnings provides downside protection at lower cost than trimming a high-quality income name at 12-month low
• Neither full long nor full trim is correct—reality is high-quality name with structural concerns in difficult macro requiring balanced risk management

→ vs aggressive: Full-size position with stop below 159 in confirmed downtrend asks to be wrong with maximum pain if any catalyst fails; this is symmetric pain, not asymmetric upside

→ vs aggressive: Apathy-as-bullish thesis interprets absence of positioning as opportunity rather than evidence nobody wants to defend the long side—two different reads of same data with no justification for optimistic choice

→ vs conservative: Framing that AMT has been definitively rejected by the market as AI vehicle ignores forced selling dynamics and rate pressure; when those shift, AI narrative could matter again

→ vs conservative: Dividend coverage at 0.90x is stable and structural, not deteriorating—binary framing of dividend cut vs. balance sheet deterioration is overly dramatic

→ vs conservative: Downside path to 150 prices in worst case as likely case, mirroring aggressive analyst's error of pricing best case as likely case

→ vs conservative: Dismissing entire catalyst because it's been quiet for 18 months is the kind of thinking that misses generational buying opportunities