Press House — Proforma Review: Findings Register

Asset: Press House — 331 N Street NE, Washington, DC (Union Market / NoMa) Deal: Fox Real Estate acquisition · 356 units · $110,000,000 · 24-month value-add lease-up File reviewed: 2) Press_House_Proforma_Model.xlsx (9 sheets, 15,990 cells) Reviewer: internal pre-financing review · Status: DRAFT (external verification in progress) Model headline: Project IRR 61.1% · EM 2.72x · LP IRR 52.0% · going-in cap 4.15% · exit cap 5.0%

Severity key: C = Critical (changes the decision / blocks financing as-is) · H = High (material $ / lender will push back) · M = Medium · L = Low / housekeeping · O = Observation.

A. Decision-level findings (Critical / High)

C-1 — Financing feasibility: 80% LTV / $88M loan is likely not sizable at close

• Going-in DSCR ≈ 0.69–0.76x (T12 vs April-snapshot NOI) against $6.028M annual I/O debt service. The asset does not cover its own debt at acquisition.
• Going-in debt yield ≈ 4.7–5.2%. Bridge/transitional lenders on a lease-up deal typically require a minimum debt yield ~6–7%+ and an interest reserve. At 4.7%, a lender sizes the loan well below $88M, or demands a large reserve (more equity) — either way the 80% LTV and the equity-at-close figure ($26.27M) understate the true day-1 equity.
• No interest-reserve line was found in the Dev Budget or CashFlow. The negative carry is funded only by ~$0.75M of modeled equity cash calls (understated — see C-3, H-1).
• Action: confirm the actual debt term sheet (lender, proceeds, rate, reserve, recourse, debt-yield/DSCR tests, extension options). If proceeds are debt-yield-constrained, re-solve the capital stack. This is the single most important item to resolve before "sending to financing."

C-2 — Property taxes: no reassessment-on-sale (DC reassesses toward purchase price)

• Model holds property taxes at the T12 average (~$1.211M/yr) grown at 2% (Projection E38). The latest month (Apr 2026) is already $130,819/mo ≈ $1.57M/yr — i.e., taxes are rising before any sale.
• DC reassesses real property; a $110M sale typically resets assessed value toward price. At an assumed DC Class-2 rate (placeholder 1.65% — VERIFY), taxes ≈ $1.815M/yr, ~$604K/yr above model.
• Impact: stabilized NOI overstated ~$604K/yr → at a 5% exit cap that is ~$12.1M of overstated exit value. Fixing taxes alone drops Project IRR 61% → 46%.
• Action: verify DC Class-2 commercial rate and the assessor's reassessment-on-sale practice; obtain current TY2026 assessment & any pending appeal. Re-model taxes off purchase price.

C-3 — "Welcome Tax" is a Montreal artifact; DC transfer/recordation taxes are higher

• Inputs B9 charges a "Welcome Tax" of 1.45% = $1,595,000. "Welcome Tax" (taxe de bienvenue) is the Quebec/Montreal transfer duty — this is a template carry-over from Fox's Montreal models, not DC.
• DC levies deed recordation + transfer tax; for commercial/high-value transfers the combined rate is materially higher (placeholder 2.9% — VERIFY; DC commercial >$2M may be ~5% combined). At 2.9% → ~$3.19M, i.e., closing costs / day-1 equity understated ~$1.6M (more if 5% applies).
• Action: verify DC recordation (DC Code §42-1103) + transfer (§47-903) rates for this transaction type/size; correct Inputs. Re-check who pays (often split, but model should reflect the LOI/PSA).

H-1 — Loss-to-Lease dropped from the monthly CashFlow that drives the IRR

• The Projection deducts "Gain/(Loss) to Lease" (row 20; −$93,365/mo in M1 tapering to −$28,009). The Monthly CashFlow revenue build (rows 15–20) omits this line entirely, so interim NOI is overstated by ~$1.39M over the hold.
• Because GPR is built at market $/SF, the loss-to-lease deduction is required to reach in-place rent. Dropping it overstates interim cash, understates equity need, and inflates IRR.
• Impact: fixing it drops Project IRR 61% → 58.4% (and increases true equity required).
• Action: rebuild the monthly revenue to include loss-to-lease consistent with the Projection.

H-2 — Going-in metrics use the most favorable single month (April 2026)

• Going-in NOI uses the April snapshot (Projection col D, NOI $4.567M → cap 4.15%). The T12-average NOI is $4.146M → cap 3.77%.
• April is flattered by anomalously low utilities ($26,390 vs $80,957 T12 avg) partly offset by high taxes. Net: April overstates run-rate NOI.
• Impact: the true going-in basis is a sub-4% cap — very thin for a value-add deal and important context for the appraiser/lender.
• Action: present going-in on a normalized T12 basis; reconcile the utility anomaly (credit/true-up vs seasonality).

H-3 — OPEX savings appear double-counted (payroll cut and contract-services cut)

• Payroll cut to $65,000/mo (from $76,963 Apr / $89,734 T12) — note I29 "≈2.5x the Ledger Payroll" is unclear. Simultaneously Contract Services cut ~75% ($62,611→$23,200) with note "Janitoring to in-house payroll, renegotiate security."
• Moving janitorial in-house should increase payroll, not coincide with a payroll cut. Booking both the contract-services savings and a payroll reduction double-counts the saving.
• Total OPEX falls from ~$379K/mo to ~$355K/mo while occupancy rises 87%→95% and EGI grows ~33%; OPEX ratio drops to ~36%, aggressive for a 356-unit asset (typical 40–50%).
• Action: rebuild a bottom-up staffing/contract plan; reconcile in-house vs contract; expect the lender's appraiser to haircut these cuts.

H-4 — Exit cap (5.0%) is at/through going-in and likely optimistic

• Exit cap 5.0% vs a normalized going-in cap ~3.8–4.2%: the model exits at a higher cap than entry (good discipline) — but 5.0% on a 2028 DC exit needs market support, and returns are highly cap-sensitive.
• Impact: exit at 5.5% → IRR 45%; 6.0% → IRR 29%.
• Action: support exit cap with dated DC Union Market/NoMa comps (see research). Show returns across 5.0/5.5/6.0%.

B. Combined scenarios (independent recompute)

(Corrected.) Property tax = run-rate normalization (Class 1A 0.85% confirmed; reassessment not punitive). Transfer 2.9% is verified but SPLIT — buyer customarily pays only 1.45% recordation — so it is shown as a sensitivity, not baked into the base. Engine reproduces the 61.12% IRR exactly before adjustments. C-2/C-3 above are the original draft; see §G/§H for the corrected conclusions.

C. Model-integrity / housekeeping (Medium / Low)

• M-1 Related-party fees stack: Acq fee 0.75% ($825K) + Broker 1.5% ($1.65M) + Fox disposition 0.75% ($1.23M) + Development fee 1% of a hardcoded $160M target ($1.6M). All flow to Fox/GP. Confirm these are disclosed to the lender/LPs and are market.
• M-2 Development fee hardcodes $160M (=0.01*160000000/24) rather than linking to the model's own stabilized value — disconnected if assumptions change.
• M-3 Retail lease-up +$35K/mo at month 24 ($420K/yr of currently-vacant ground-floor retail) is speculative; ground-floor retail leasing in 24 months is slow. Treat as upside, not base.
• M-4 Insurance ~$9,786/mo (~$329/unit/yr) looks low for US multifamily (typ. $400–700/unit) amid rising premiums — likely understated.
• M-5 "Total Project Cost" definition (Summary H11) mixes acquisition cost + exit selling/disposition costs but excludes the $2.78M dev budget — non-standard denominator for Yield-on-Cost (6.93%).
• M-6 Physical capex is very light (~$70K FFE ≈ $197/unit; no unit-renovation capex). Defensible only if thesis is purely operational; verify deferred maintenance via PCA.
• L-1 Broken formula: Rent Roll market-$/SF column (E6:E14) is #REF! wrapped in IFERROR → silently returns 0. Cosmetic but signals post-build edits; sweep for other #REF!.
• L-2 EM formula (CashFlow B61 =AC58/AC69) ignores interim distributions; minor and conservative.
• L-3 Units shown with fractional counts (e.g., occupied 309.3) because occupancy is derived from $ ÷ avg rent rather than unit status — fine for modeling, label as such.

D. Regulatory / DC-specific items to confirm (research)

• D-1 Inclusionary Zoning: 31 IZ units (8 studio + 12×1BR + 11×2BR) are rent-restricted (MFI-capped); model correctly assumes no uplift on them. Confirm covenant terms, MFI band, and that vacancy/turnover rules are respected. IZ presence implies post-2009 construction.
• D-2 Rent control: DC rent stabilization applies to pre-1976 buildings (5+ units). If Press House is post-2000 construction (likely, given IZ), it is exempt — confirm certificate-of-occupancy date.
• D-3 TOPA (Tenant Opportunity to Purchase Act): affects a future sale of residential property in DC — can delay disposition and requires tenant offers. Material to the 24-month exit; confirm strategy (e.g., TOPA waivers, bulk sale mechanics).
• D-4 Confirm vintage / year built / certificate of occupancy (drives D-1/D-2 and capex).

E. Data governance note

• The Residential Rent Roll (sheet 6) contains real tenant PII (names, tenant IDs, lease dates) sourced from Yardi. This data is kept internal only; it is never sent to web searches/third parties and will be redacted from any shared briefing. Aggregate unit-mix/rent statistics are fine to use.

F. What still must be externally verified (feeds the Verification Ledger)

1. DC deed recordation + transfer tax rate for this transaction (→ C-3).
2. DC Class-2 property tax rate + reassessment-on-sale practice + current assessment (→ C-2).
3. Union Market / NoMa market rents ($/SF by unit type), vacancy, concessions (→ H-2, H-3, thesis).
4. Cap rates & comparable sales 2025–2026 ($/unit, $/SF, cap) for going-in and exit (→ H-4).
5. Supply pipeline in NoMa/Union Market (oversupply → concession risk to lease-up).
6. Asset facts: developer/owner (T12 tree = "foulgerpratt" → Foulger-Pratt?), year built, unit count, retail GLA, why selling.
7. Financing market: realistic bridge terms (rate, LTV, debt yield, reserve) for a sub-1.0x DSCR lease-up (→ C-1).
8. Site risk: FEMA flood, crime, demographics/income trends.

G. POST-VERIFICATION UPDATES (supersedes items above where noted)

↻ C-2 CORRECTED — downgraded Critical → Medium. DC taxes a 356-unit apartment as Class 1A residential @ 0.85% (verified OTR), not Class 2 commercial (1.89%). Back-solving the T12 tax (~$1.21M ÷ 0.85%) implies a current assessment ~$142M — above the $110M price — so reassessment-on-sale is neutral-to-favorable, not punitive. My original $604K/yr NOI hit / $12M value hit is withdrawn. The real, smaller issue: the model uses the T12 average ($1.21M) while the latest month runs ~$1.57M/yr → ~$359K/yr understated on current trajectory. Tax line also bundles the NoMa BID surcharge (~$0.15/SF ≈ $42K/yr) and the small retail/commercial (Class 2) portion. Action: pull the actual TY2026 assessment from MyTax.DC.gov; underwrite to current run-rate; model a (favorable) appeal toward $110M.

↻ C-3 CONFIRMED. DC combined recordation+transfer = 2.9% for commercial (since Oct-2023). "Welcome Tax" 1.45% ($1.595M) understates by ~$1.595M if commercial treatment applies. Confirm residential vs commercial classification for transfer-tax purposes with title/counsel — apartment entity sales are typically commercial 2.9%.

↻ H-2 CONFIRMED via comp. Going-in $308,989/unit ≈ the Belgard comp (33 N St NE, 346 units, $311K/unit, Aug-2024, same sponsor Foulger-Pratt, but stabilized high-90s occupancy). Fox is paying near-stabilized per-door pricing for a 65%-economic-occupancy asset. Normalized going-in cap ~3.8%.

＋ H-5 NEW — Concession/lease-up assumptions are far below current NoMa market (market-verified). NoMa/Union Market vacancy ~10–11% (vs DC ~7%); area concessions up to ~2–4 months free; DC rents fell ~2.2% in 2025. Press House itself currently advertises ~2 months free and prices ~20% below the area average. The model assumes vacancy 10%→4%, concessions 4%→1.5% of GPR, and 3% rent growth — i.e., stabilizing through the current submarket and giving far less than market concessions. This is the single most aggressive operating assumption and the main threat to hitting stabilized NOI. Action: re-underwrite lease-up velocity, concessions (model real free-rent), and rent growth to current NoMa data; this also feeds the exit-NOI downside that the exit-cap sensitivities bracket.

＋ O-7 NEW — Exit liquidity. DC multifamily sales volume −26% YoY (~$5.3B, Newmark); a large NoMa comp (Flats 130, 643 units) is on-market via CBRE with no printed price — a thin, slow buyer pool is a real exit risk for a 24-month business plan.

↻ D-1 IZ CONFIRMED. 31 affordable/IZ units (incl. 11×2BR at 50% AMI) match public record; model correctly assumes no uplift. Implies post-2009 construction.

↻ D-2 RENT CONTROL — CONFIRMED EXEMPT. Press House delivered ~2021 (new construction), so it is exempt from DC rent stabilization (pre-1976 only). The market-rate rent-growth thesis is legally permitted (market softness, not rent control, is the constraint).

↻ D-3 TOPA — RISK LARGELY MITIGATED (was a risk). The RENTAL Act of 2025 (eff. 12/31/2025) exempts buildings with a permanent C of O within 15 years from TOPA's offer-of-sale (Notice of Transfer only). Press House (~2021) qualifies → both this purchase and the 2028 exit avoid the TOPA delay that hit the Belgard in 2024. Net positive vs my initial flag.

＋ O-8 Seller / ops. Seller is the Foulger-Pratt / Juster / ClearRock JV (developer/current owner; model built from their Yardi books). Resident reviews report unstaffed 24/7 desk, weeks without hot water, security/loitering, soiled elevators → supports the "better operations" upside and signals deferred maintenance/reputational drag vs the thin $70K FFE budget.

H. ROUND-2 CORRECTION (transfer tax) + data additions

C-3 transfer tax - DOWNGRADED to Low/housekeeping. DC custom splits the tax: BUYER pays 1.45% recordation, SELLER pays 1.45% transfer. The model's 1.45% is DEFENSIBLE as the buyer's cost; understated (~$1.6M, to 2.9%) ONLY if Fox agreed to pay both sides. Action: rename "Welcome Tax" -> "DC recordation tax"; confirm allocation in the PSA. Audit "lender base" no longer bakes in the transfer add-back (base IRR ~33-41%; +2.9% transfer is a separate ~3-4pt sensitivity). Data added: rent comps (Zumper/Apartments.com), demographics (ZIP 20002 income ~$120K, 62% renters), Gallaudet enrollment (1,260, flat). See RESEARCH_DOSSIER section 8 and VERIFICATION_LEDGER round-2.