QTS Realty

Phase A — Understand the business

Lens 1 · Company Overview

QTS (legally Quality Technology Services / QTS Realty Trust) is a wholesale and hyperscale data-center developer-operator — the crown-jewel real-estate asset inside Blackstone's ~$1.3T AUM machine. It designs, builds, owns and operates large-scale, purpose-built data-center campuses leased to the biggest cloud and AI tenants in the world under long-dated, investment-grade contracts.

What it actually is (not the "colo REIT" of its listed past): Pre-2021 QTS was a mid-cap public REIT (NYSE: QTS) doing a mix of retail colocation, hybrid-colo and hyperscale. Blackstone took it private on 31 August 2021 for ~$10B ($78.00/share, 21% premium, debt included) via a stack of perpetual-capital vehicles — Blackstone Infrastructure Partners (BIP) and Blackstone Real Estate Income Trust (BREIT). Since then the strategy has been re-pointed almost entirely to build-to-suit hyperscale: anchor tenants pre-commit to entire buildings before a shovel hits dirt, on 15–20 year leases with CPI-linked annual escalators.

Scale today (company's own disclosure):

• >3 GW total contracted power capacity
• >$3B annual leasing revenue
• 75+ data centers in operation or under development, US + Europe
• $18–25B pre-leased development pipeline — QTS/BREIT cite "$18+ billion," Blackstone commentary cites "$25 billion," extending through 2028. Treat the range as $18–25B; the delta is disclosure timing, both are "substantially pre-leased."
• Growth under Blackstone: commissioned capacity ~400 MW → ~3 GW (roughly 7–8x); development pipeline ~$1B → $18–25B; leased capacity up ~14x since the 2021 buyout.

Customers: the named hyperscalers — Microsoft, Google (Alphabet), AWS, Meta — plus Fortune-1000 enterprise and government. Contracts are the whole story: 15-year leases, blended AA- tenant credit, ~3.6% annual escalators on the Northern Virginia assets that just changed hands. Take-or-suit, not merchant — revenue is contracted before capex is committed.

Suppliers / dependencies: utilities (Georgia Power, PJM utilities), GC/construction, power-and-cooling OEMs, and — increasingly — the grid interconnection queue itself, which is now the binding constraint on the whole model (see Lens 2).

, unaudited per public sources. This is correct: QTS is a private Blackstone portfolio company with no CIK and no SEC financials.]

Lens 2 · Supply Chain

Map, upstream → QTS → end customer, every named stakeholder:

Upstream — power (the real chokepoint):

• Georgia Power (Southern Co.) — a 20-year PPA for ~350 MW underpins the Atlanta/Suwanee/Fayetteville campuses. The Georgia PSC in Dec 2025 approved 9,885 MW of new energy, most earmarked for large loads (data centers) — the supply that QTS's Georgia growth rides on.
• PJM Interconnection — the mid-Atlantic grid (Northern Virginia, the world's densest data-center market). PJM's capacity auction cleared at $329.17/MW-day for 2026-27, up 11x from $28.92 in 2024-25 — a direct, brutal read on how scarce firm power has become. FERC has ordered PJM to write large-load colocation rules.
• Construction / GC, power & cooling OEMs — QTS is moving to water-free cooling and liquid-cooling retrofits for GPU-dense (>50 kW/rack) halls.

Midstream — QTS itself: land assembly → entitlement/permitting (the newly-exposed failure point) → power procurement → build → operate. QTS's differentiator is land + power + entitlement banked ahead of demand at campus scale.

Downstream — the buyers: the four hyperscalers above, contracted on 15–20 yr AA- paper. Concentration is high and undisclosed at the name level, but structurally the buyer set is <10 counterparties for the bulk of the pipeline.

Chokepoints / single-source dependencies:

1. Grid interconnection & firm power — the true single point of failure. QTS can build faster than the utility can energize; a slipped interconnection date strands capex.
2. Local entitlement / community consent — newly proven to be a hard constraint (Prince William, see Lens 8/13).
3. Tenant concentration — a handful of hyperscalers; a capex-pause at one moves the whole book.

Names or it didn't happen — the chain is: Southern Co./Georgia Power + PJM utilities → QTS campuses (Atlanta, Suwanee, Fayetteville GA; Manassas/Sterling VA; Groveport OH; Cedar Rapids IA; Cambois UK) → Microsoft / Google / AWS / Meta.

Lens 3 · Competitive Advantages (moats)

QTS's moat is not technology — data-center shells are commoditizing. The moat is a stacked scarcity position + a balance sheet no independent can match:

1. Banked power + land at campus scale. In a market where CBRE put primary-market vacancy below 3% and pre-leasing above 90% in H2 2025, the scarce input is not floor space — it is entitled land next to firm megawatts. QTS's decade of land/power assembly (350 MW Georgia PPA, multi-GW campuses) is a genuine, hard-to-replicate moat. This is a switching-cost + scarcity moat, not a brand one.
2. Blackstone's cost of capital and perpetual capital. QTS can fund $18–25B of development because it sits inside BIP/BREIT (perpetual, no fund-life forced-sale clock) and can term out stabilized assets into investment-grade CMBS at will — it raised ~$4.5–4.6B across four securitizations in 2026 alone and carries a Baa2 (investment-grade) corporate rating from Moody's. No standalone wholesale operator (Vantage, Aligned, CyrusOne, Switch) has this exact combination of scale + IG rating + perpetual equity.
3. Switching costs / contract lock. 15–20 yr leases with AA- tenants and CPI escalators create bond-like, sticky cash flows; a hyperscaler that has fitted out a hall does not move.
4. Bargaining power — mixed. Over suppliers (utilities): weak-to-improving — QTS needs the grid more than the grid needs any one campus, though its load size gives it a seat at the PPA table. Over customers: weak on price, strong on availability — in a <3% vacancy market the hyperscaler needs QTS's ready megawatts more than QTS needs any single tenant, which is why pre-leasing runs >90%. Net: QTS's power is that it can deliver energized capacity on a timeline nobody else can — availability, not pricing, is the lever.

Where the moat is thinner than bulls think: the shell/build layer is replicable given capital, and capital is flooding in (Aligned bought by NVIDIA/Microsoft/BlackRock for $40B Oct 2025; Vantage raised $9.2B equity from DigitalBridge/Silver Lake). The durable edge is the power + entitlement + IG balance sheet triple, not the buildings.

Lens 4 · Segments

QTS does not report public segment financials (private, no filings) — n/a — private, not disclosed on formal segment breakouts. Structurally:

• By product: the business has shifted decisively from its legacy 3-tier model (hyperscale / hybrid-colo / retail colo) to ~majority build-to-suit hyperscale under Blackstone. The pre-leased pipeline ($18–25B) is overwhelmingly hyperscale. Legacy enterprise/government colo persists but is no longer the growth engine.
• By geography: US-centric (Atlanta/Suwanee/Fayetteville GA hub; Northern Virginia; Groveport OH; Cedar Rapids IA at ≥$750M) with a European push — a new UK campus at Cambois (first phase completing 2026). The Northern Virginia (PJM) exposure was partly monetized to Digital Realty in June 2026 (Lens 8).
• Trend & cause: accelerating, and the cause is singular — the AI training/inference buildout. Hyperscaler 2026 capex guided to $635–690B, +67–74% YoY, ~75% AI-directed. QTS is a direct beneficiary of that flow, contracted ahead via build-to-suit.

and a wholesale IG-tenant NOI margin of ~65-70%, run-rate NOI is roughly $2.0-2.1B — arithmetic only, not disclosed. Flag as, do not treat as a QTS figure.]

---

Phase B — Measure performance

Lens 5 · Funding & valuation trajectory (+private lens-swap for "Earnings Result")

QTS has no quarterly print — the equivalent scoreboard is its valuation mark-up and financing cadence:

The mark: ~$10B → ~$60B in under four years is a ~6x paper markup. It is a reported/estimated private mark, not an audited transaction value — caveat heavily. The Digital Realty deal is the one hard, arms-length data point: 288 MW of 100%-leased AA- hyperscale product sold at ~$27M/MW and a >6.5% stabilized cap rate. That is the number to anchor QTS's asset value on, not the $60B headline.

Financing (the real "earnings" — access to cheap term debt):

• Nov 2025: record $3.46B CMBS on 10 data centers
• Jan 2026: $750M ABS; Feb 2026: ~$1.95–2.05B CMBS (BX Trust 2025-VLT6, Goldman-led, refi'd ~$1.76B + funded expansion reserves)
• Apr 2026: 10-yr IG green bond, first-time Moody's Baa2, part of a $4.6B raise for the Fayetteville campus (T+162.5 bps)
• ~$4.5–4.6B raised across four deals in 2026

Balance-sheet flag: the pace of securitization is both a success signal (assets are financeable at IG spreads) and a warning — GlobalCapital itself framed "rapid QTS dealmaking" as a warning and a sign of success. QTS is levering stabilized assets hard to fund the next tranche of development; the model depends on continuous access to the CMBS bid.

Guidance / outlook: no formal guidance (private). Directional: pipeline pre-leased through 2028; UK and additional US campuses in build. Tone is unambiguously expansionary.

Lens 6 · Founder / sponsor commentary trend (+private lens-swap for "Earnings Calls")

No earnings calls. Proxy signal = Blackstone's public posture + the leadership change:

• Blackstone's stated conviction has escalated, not softened: it calls itself "the leading investor in data centers and power," is building the $80–100B GAIIP partnership, committed $5B to a Google-TPU AI-infra JV (first 500 MW by 2027), and launched a public data-center REIT (BXDC). The through-line across 18 months: demand is exceeding their 2023 underwriting.
• The tell in the leadership change: founder Chad Williams stepped down 18 Apr 2025 after 20+ years; Blackstone paid ~$3B to buy out his shares and installed co-CEOs David Robey (ex-COO) and Tag Greason (ex-CGO) — operators, not a founder-visionary. Read: Blackstone wants execution-and-monetization management for the harvest phase, not empire-building. The "things they stopped saying" is founder-led narrative; the new recurring phrase is pre-leased, investment-grade, long-term contracts.
• The one discordant note (Jul 2026): QTS/Blackstone walked away from the Prince William Digital Gateway — the would-be largest data-center campus on earth (see Lens 8). That is a rare public retreat and cuts against the all-gas narrative; it is a permitting/consent retreat, not a demand retreat, but it matters.

Lens 7 · Cap table & public comps (+private lens-swap for "Comps")

QTS cap table (the syndicate quality — an IPO-proximity / durability tell):

• Owned by Blackstone Infrastructure Partners (BIP) + BREIT. As of 30 Sep 2025, BREIT held ~35% of QTS, representing ~17.4% of BREIT's total real-estate value. (A separate DCD figure cites "20.4%" — treat 17.4% as the BREIT-sourced primary, 17–20% as the range. Either way QTS is BREIT's single largest position — a concentration risk for the vehicle, see Lens 10/13.)
• Founder equity bought out (~$3B) in 2025 — the cap table is now essentially all-Blackstone.
• This is not an IPO-track cap table. There are no crossover funds (Fidelity/T. Rowe/Coatue) taking QTS primary — the "crossover = IPO-proximity" tell is absent. QTS's path-to-tradeable runs through Blackstone's own vehicles (BXDC, CMBS), not a QTS S-1 (see Lens 11).

Public comps (there is no QTS multiple — use the listed wholesale/colo peers):

Read: the public wholesale/colo group trades ~18–22x AFFO at ~2–3% yields — you buy for AFFO growth, not income. QTS, were it public, would likely command the top of that band (fastest growth, best pipeline, IG tenants) — but it is not public, and the honest cross-check on its private mark is the >6.5% stabilized cap / ~$27M/MW arms-length DLR trade, which is more conservative than the ~$60B headline implies on a look-through basis. Do not fabricate a QTS EV/EBITDA — n/a, not sourced.

Lens 8 · Stock-Price / valuation Catalysts (re-framed: value-mark catalysts, no ticker)

Events that have moved QTS's value or narrative materially:

1. Take-private (Jun–Aug 2021): the +21% premium / $10B deal reset QTS from public-REIT multiple to Blackstone's growth-capex model.
2. The valuation re-rate to ~$60B (by Apr 2025) on the AI-capex wave — a ~6x mark.
3. Founder ouster + $3B buyout (Apr 2025): signaled the shift to harvest-mode management.
4. The 2026 securitization blitz (~$4.6B, first IG rating Baa2): proved QTS assets are financeable at investment-grade spreads — a validation event.
5. Digital Realty buys BX's 64% of 288 MW NoVA for $7.8B GAV (30 Jun 2026): the first hard, external valuation print — a partial exit at a rich cap, and a template for future monetization.
6. BXDC public REIT IPO (14 May 2026, $20.00/sh, $1.75B, 87.5M shares): the public drip-vehicle that will buy QTS-developed pre-leased assets — QTS's real path to public-market capital.
7. Prince William Digital Gateway collapse (early Jul 2026): QTS abandoned the ~$100B / 37-building / 22M-sq-ft Northern Virginia mega-campus after Compass withdrew (May 2026) and legal/community/regulatory opposition proved insurmountable — not a power or demand problem. The single most important recent negative catalyst; proves local consent is now a hard limit on the largest projects.

Pattern: QTS's "stock" is a private mark set by Blackstone, so the value-moving events are financings, asset sales, and the founder transition — plus, uniquely, permitting outcomes. The market (here, the CMBS and M&A market) reacts most to IG-tenant credit + power availability; the Prince William retreat shows it also now reacts to entitlement risk.

---

Phase C — Judge people & books

Lens 9 · Management

Founder era (2003/2005–Apr 2025): Chad Williams. Built QTS from a single Kansas facility into a top-tier national platform, took it public (2013), and sold to Blackstone at ~$10B (2021). Genuine builder track record. Blackstone paid ~$3B to exit him in 2025 — a clean, well-compensated founder exit, not a scandal departure.

Current: co-CEOs David Robey (ex-COO) & Tag Greason (ex-CGO), since Apr 2025.

1. Track record: operators risen from within — Robey ran operations, Greason ran growth. They have executed the 400 MW→3 GW / $1B→$18-25B build under Blackstone's mandate. No independent P&L-CEO history to grade (private, no filings).
2. Tenure & skin in the game: ~15 months as co-CEOs; insider ownership is n/a — private, Blackstone-owned, so "skin in the game" runs through Blackstone incentive comp, not disclosed equity. This is a governance weakness vs. a founder-owner: management is agents of the GP.
3. Capital allocation: disciplined by design — build-to-suit means capex follows contracted revenue, not speculation. The monetization playbook (CMBS + DLR sale + BXDC drip) is sophisticated and de-risking. ROIC not disclosable.
4. Red flags: the structural one is agency — QTS management serves Blackstone's LPs (and carry), whose interests (harvest, mark-up, term-out debt) are not identical to a hypothetical minority holder's. Also: related-party density is inherent — QTS develops assets that Blackstone's other vehicle (BXDC) is designed to buy, and QTS is BREIT's largest position that BREIT's sponsor also manages. All arm's-length in theory; all one house in fact.
5. Founder vs professional manager: decisively shifted to professional-manager/steward archetype. Correct for the harvest-and-monetize phase; the empire was already built.

Lens 10 · Forensic Red Flags

, unaudited per public sources.]

No audited financials exist — that is itself the headline forensic caveat. What to scrutinize instead (all structural, all /):

1. Leverage layering / continuous-refi dependence. ~$4.5–4.6B of CMBS/ABS in 2026 on top of a Baa2 corporate rating. The model requires the securitization market to stay open at IG spreads. If CMBS spreads blow out (rate shock, DC-CMBS fatigue — GlobalCapital already flagged the pace as a warning), QTS's development funding tightens fast. This is the single biggest financial risk and it is off-balance-sheet by design.
2. Private-mark opacity. The ~$60B figure is a reported/sponsor-influenced mark, not a transaction. The one arm's-length print (DLR, >6.5% cap, ~$27M/MW) is more conservative than a look-through of the headline implies — watch for the gap between BREIT's carried NAV on QTS and where third parties actually transact. BREIT's valuation methodology has drawn independent criticism as "aggressive" vs. public-REIT implied cap rates.
3. BREIT concentration & liquidity loop. QTS is ~17–20% of BREIT's real-estate value and its largest single position. BREIT is a semi-liquid vehicle with a redemption gate history (the 16-month gate, 2022–23). If BREIT faced a redemption wave, an illiquid mega-position like QTS is exactly what you cannot sell into it — hence the CMBS/DLR/BXDC monetization is partly a BREIT liquidity-management exercise, not only growth funding.
4. Related-party web. QTS develops → BXDC (same sponsor) is built to buy; QTS sits in BREIT (same sponsor manages the mark). Every leg is disclosed and arm's-length in form, but the incentive alignment is Blackstone-first.
5. Tenant concentration (undisclosed, structurally <10 hyperscalers) — a single hyperscaler capex pause would be material and is not visible in any public number.

Regulatory / legal:

• SEC: no findings — QTS has no CIK, no EDGAR presence (private). Verified via regulatory-findings.md.
• Non-SEC / litigation: the Prince William Digital Gateway carried active legal challenges (administrative-notice violations, community litigation) that contributed to QTS abandoning it in Jul 2026. Not an enforcement action against QTS, but material project-level litigation. Broader sector overhang: ratepayer/consumer-advocate opposition and moratorium bills in Georgia, New York, Oklahoma targeting data-center power-cost socialization — regulatory sector risk QTS is squarely exposed to.
• Summary: No SEC enforcement (private, no CIK). Material project-level litigation at Prince William (now abandoned). Rising sector-level regulatory/ratepayer risk in QTS's key power markets. Verified via SEC EDGAR EFTS (0 findings, no CIK), web search, and public reporting as of 2026-07-06.

---

Phase D — Project & stress-test

Lens 11 · IPO-readiness & path-to-tradeable (+private lens-swap for "Forward Projection")

The honest answer: QTS will (almost certainly) not IPO — and that is the whole point of owning it privately. Blackstone took QTS off the public market specifically to escape quarterly discipline and run a $18–25B capex program. It has instead engineered three monetization rails that let it harvest value without a QTS S-1:

1. CMBS/ABS securitization — term out stabilized, IG-tenant assets into the bond market (~$4.6B in 2026, Baa2). Recycles equity into the next development tranche. The primary, continuous rail.
2. Strategic asset sales — sell mature, 100%-leased campuses to strategics (Digital Realty, $7.8B GAV, Jun 2026). Crystallizes the private mark against a public buyer. The event-driven rail.
3. BXDC (Blackstone Digital Infrastructure Trust) — Blackstone's own public data-center REIT, IPO'd 14 May 2026 ($20.00/sh, $1.75B, NYSE), a blind pool explicitly designed to buy into QTS-led, pre-leased developments. This is the retail-access rail — the closest thing to "buying QTS" a public investor can do.

Path-to-tradeable verdict: there is no direct tradeable QTS security and no near-term S-1 catalyst. The +private "IPO window" lens resolves to: monetization is already live via CMBS + DLR-style sales + BXDC; a standalone QTS IPO is unlikely while the private/perpetual structure is working. Milestones that could force a QTS-specific listing (none imminent): a BREIT liquidity crisis needing a public exit, or a Blackstone decision to spin QTS as its own public co rather than feed BXDC.

[No forecast.ts EPS forecast logged — QTS has no EPS, no ticker, and this is an unattended --watchlist run. The scoreable prediction, if one were logged, would be a binary: "Blackstone files/announces a standalone QTS IPO or take-public by YE2027" — p≈0.15, because the three existing rails make a QTS S-1 unnecessary. Not logging per --watchlist rules.]

Lens 12 · Bull vs Bear

Bull case. QTS is the best-positioned wholesale hyperscale platform in the United States, sitting on the scarcest asset in tech — entitled land next to firm megawatts — with a $18–25B pipeline that is already pre-leased to AA- hyperscalers on 15–20 yr CPI-escalating contracts. It is funded by the deepest-pocketed, cheapest-cost-of-capital sponsor in private markets, can term out assets at IG spreads on demand, and rides a hyperscaler capex wave guided to $635–690B in 2026 (+70%). In a <3% vacancy, >90% pre-leased market, availability — QTS's edge — is worth more than pricing. The mark went ~$10B → ~$60B in <4 years and the arms-length DLR trade (>6.5% cap) says the underlying assets are genuinely, conservatively valuable. The secular tailwind is real and QTS is the purest private-market expression of it.

Bear case (permanent-impairment vectors).

1. Power/permitting is the binding constraint, and consent is turning against it. PJM capacity 11x'd; ratepayer backlash and moratorium bills are spreading; and QTS just abandoned the largest project on earth (Prince William) over community/legal opposition. Growth is now gated by grids and city councils, not demand — and that gate is tightening.
2. The demand is concentrated in <10 hyperscalers whose own capex is under scrutiny. Three of four hyperscalers lost market value after recent prints on capex-sustainability fears; AI assets depreciate ~20%/yr against ~$400B of implied annual depreciation. A capex digestion pause would hit QTS's next pipeline tranche even though the signed book is contracted.
3. Financialization/leverage fragility. The model depends on continuous IG-CMBS access and a rich private mark. A CMBS-spread shock or a BREIT redemption wave (QTS = BREIT's largest, least-liquid position) could force a de-risking sale into a weak market and re-rate the mark down.

Pre-mortem (18 months out, thesis broke): Late 2027 — hyperscaler AI-capex growth decelerated from +70% to +15% as model-efficiency gains cut compute-per-dollar; a couple of build-to-suit LOIs on the un-signed pipeline slipped or shrank; simultaneously a rate/spread shock widened DC-CMBS and a BREIT redemption bump forced Blackstone to sell a QTS campus at a >7% cap. The ~$60B mark quietly became ~$45B; new-development starts paused. Nothing "blew up" — the secular story just normalized while the private mark and the leverage were priced for it not to.

Are multiples too high? The public comps (DLR ~22x, EQIX 19x AFFO) already price aggressive AFFO growth; the private QTS mark ($60B) is sponsor-set and looks full vs. the >6.5%-cap arm's-length print. Not obviously cheap on any measure.

Contrarian view (what the market refuses to see): the crowd treats "AI datacenter" as one uniformly bullish trade. The truth for QTS specifically: the equity upside is fully captured by Blackstone and its carry — every public rail (CMBS bonds, BXDC, DLR) hands the outside investor the beta and the yield while Blackstone keeps the alpha and the optionality. And the binding risk is not demand — it is the newly-proven power that a Virginia county board can kill a $100B project. Being right about AI does not make you money on QTS; you can't buy it, and the wrappers you can buy are engineered to give the sponsor the upside.

Lens 13 · Devil's Advocate (short-seller)

Dismantling the bull case:

• You cannot short it and you cannot own it — so the "trade" is a category error. The only tradeable proxies are BX (diversified GP, QTS is a sliver), BXDC (a blind-pool that hasn't bought the assets yet — you're underwriting Blackstone's future allocation, not QTS), and DLR/EQIX (which QTS competes with and out-grows — long them and you're partly long QTS's rivals). There is no clean expression.
• Revenue concentration is a loaded gun. <10 hyperscalers underpin the book. Microsoft's public willingness to walk away from data-center leases when its own capacity math changes is documented sector behavior; the signed leases hold, but the pipeline (the source of the ~$60B mark) depends on those same buyers keep-signing at +70% capex growth. That is not contracted — it is a bet on a rate of change.
• The moat is weaker on the build layer than bulls claim. $40B (Aligned) and $9.2B (Vantage) of fresh capital just poured into direct competitors. Capital is not the moat's protection — power and entitlement are — and those are exactly what Prince William proved can be denied.
• The related-party structure should make a skeptic uneasy. QTS develops → BXDC (same sponsor) buys → BREIT (same sponsor marks) holds the largest slug. Every valuation touchpoint is inside one house. The $60B mark is not adversarially tested; the one time an outside buyer set the price (DLR), it printed a >6.5% cap — hardly a trophy multiple.
• What must hold for the mark: hyperscaler capex stays near +70%, CMBS stays open at IG spreads, and grid/permitting doesn't tighten further. If pipeline growth disappoints 20–30%, the signed NOI is fine but the development NAV — the bulk of the $10B→$60B markup — compresses hard, because that value is future-pipeline, not in-place cash flow.
• The single permanent-impairment scenario: a coordinated hyperscaler capex-digestion pause (efficiency-driven, à la a DeepSeek-style step-change in compute economics) coincident with a CMBS-spread shock, forcing distressed campus sales out of BREIT into a bidless market. Plausibility: moderate — each leg is real; the coincidence is what makes it a tail, not a base case.

Lens 14 · Management Questions (ordered by information value)

1. What share of the $18–25B development pipeline is signed lease vs. LOI/pipeline, and what is the weighted-average remaining lease term and tenant credit on the signed portion?
2. What is QTS's customer concentration — top-1 and top-3 tenants as a % of contracted revenue?
3. How much of the ~$60B valuation is in-place NOI vs. development/pipeline NAV, and at what cap rate is the in-place book carried?
4. What is total debt / net debt and net-debt-to-EBITDA, including the 2026 CMBS/ABS stack, and how much development funding depends on continued securitization access?
5. If DC-CMBS spreads widened 150–200 bps, how many quarters of development starts could QTS fund without new securitization?
6. How does QTS's carried NAV inside BREIT get marked, and how is it reconciled to arms-length prints like the Digital Realty >6.5%-cap sale?
7. What is the firm, energized-and-contracted power (MW) vs. announced capacity, and what is the interconnection backlog by market (PJM, Georgia, UK)?
8. Post-Prince William, what has changed in the site-selection/entitlement playbook, and how much banked land is now at permitting risk?
9. What is the intended flow of assets from QTS into BXDC — pricing mechanism, governance of the related-party conflict, and expected annual volume?
10. What are QTS's assumptions for hyperscaler capex growth 2026–2028, and what happens to development starts if that growth halves?
11. What is the development yield-on-cost on build-to-suit vs. the cap rate at which stabilized assets are sold/financed (the value-creation spread)?
12. How exposed is QTS to power-cost escalation and ratepayer/regulatory backlash in Georgia and PJM, and are PPAs fully hedged for the pipeline?
13. What is the plan for liquidity if BREIT faces a redemption wave — would QTS assets be a forced-sale source, and at what discount?
14. How is management compensated and incentivized, and how is that aligned with any future minority/public holders vs. Blackstone's carry?
15. Under what conditions would Blackstone pursue a standalone QTS IPO or take-public rather than feeding BXDC and the CMBS market?

---