REITs vs. private multifamily — the actual tradeoff

A family-office CIO I’ve known for years asked me last week whether public Sun Belt multifamily REITs at 6.5 to 7 percent implied cap rates were a better deal than the private opportunities crossing my desk at 5.5 to 6 percent. The instinct behind the question — public is offering 50 to 130 basis points of pickup for the same underlying buildings — is the standard frame. I think it’s missing more than half the picture, and I’ll show why.

The honest answer if you have both options on the table: REITs and private multifamily are not the same product priced differently. They own similar buildings, but the wrappers around those buildings deliver materially different economics to whoever holds them. The relevant comparison is not gross cap rate. It’s what you keep after taxes, fees, liquidity discounts, and the optionality each structure actually provides.

This piece walks through how I think about that decision, with current numbers from Q1 2026 supplementals for the three Sun Belt names that anchor most institutional discussions: MAA, UDR, and CPT.

What you’re actually buying with each

A REIT share gives you:

A diversified pool of stabilized properties under professional asset management
Daily price discovery and continuous liquidity
Quarterly cash distributions averaging 80 to 90 percent of AFFO
Governance protections — independent board, public reporting, securities law oversight
Portfolio-level tax efficiency at the corporate layer (REITs avoid corporate-level income tax provided they distribute 90 percent of taxable income)
Capital gains optionality on the share itself (long-term cap gains rates if held over a year)
Ordinary-income tax treatment on the dividend portion that doesn’t qualify as return-of-capital

A private multifamily LP commitment gives you:

Direct fractional ownership of specific properties through a partnership structure
An illiquid 5-to-7-year hold (sometimes 7-10 with extensions)
Quarterly cash distributions tied to actual property cash flow
A K-1 each year showing your share of partnership-level income, expenses, and depreciation
The full benefit of depreciation passthrough — typically 25 to 35 percent of basis depreciated over the hold period, often accelerated through cost segregation
1031 exchange optionality at exit, allowing tax-deferred rollover into the next deal
Step-up in basis at death, which can permanently eliminate the deferred capital gains
Sponsor carry that takes 20-30 percent of upside above a preferred return

These are not substitutes priced differently. They’re complements priced for what they actually deliver.

The pre-tax picture, accurately

Start with what’s actually trading at what cap rate today. Here are the three Sun Belt multifamily REITs at Q1 2026, computed using same-store Property NOI directly from each company’s supplemental disclosure deck rather than rule-of-thumb assumptions:

	Q1 2026 SS Revenue	SS NOI Margin	Implied Cap Rate (EV-based)	vs 10Y (4.39%)
MAA	$516.98M	63.58%	6.74%	+235 bps
UDR	$398.65M	66.96%	6.51%	+212 bps
CPT	$360.01M	64.56%	6.74%	+235 bps

Trio at 6.5 to 6.8 percent. UDR’s coastal-mix premium drives its 23 bps tighter print; MAA and CPT cluster.

Private Class A/B Sun Belt multifamily has been clearing in the 5.5 to 6.0 percent range based on operator transaction context the last 12 months. (On average.) The public-private spread on a gross-cap-rate basis is 50 to 130 basis points wider on the public side — historically normal. Public has always carried this premium because of liquidity, governance overhead, and the structural cost of being a publicly traded vehicle.

If the analysis stopped here, you’d conclude that REITs offer roughly 100 basis points of pickup for accepting daily-marked volatility, which is roughly the right framing for pure pre-tax yield. It’s not the relevant framing if your capital is taxable.

The tax overlay changes the picture

If your capital is fully taxable — high-bracket individual, family office, HNW capital outside a tax-exempt structure — the math runs differently for each vehicle.

REIT cash flow path:

Property-level NOI flows up to the REIT
REIT pays no corporate income tax on distributed amounts
Investor receives a 1099-DIV
Most REIT distributions are taxed as ordinary income (REIT dividends generally don’t qualify for the 20 percent qualified-dividend rate; QBI deduction provides 20 percent deduction so effective top federal rate is roughly 29.6 percent rather than 37 percent)
A portion may be classified as return-of-capital, which reduces basis and defers tax
A portion may be classified as long-term capital gain (from REIT-level property sales), taxed at 20 percent federal plus state

Private LP cash flow path:

Property-level NOI flows to the partnership
Partnership records depreciation and amortization at the property level
Investor’s K-1 shows their share of NOI minus depreciation — for new acquisitions with cost segregation, this is typically negative or zero in the first 3-7 years
Cash distributions exceed taxable income; the difference is shielded by depreciation
At exit, gain is generally long-term capital gain (20 percent federal plus state), with depreciation recapture at 25 percent on the depreciated portion
1031 exchange allows deferral of all gain into the next deal, indefinitely
Step-up at death eliminates deferred gain permanently for heirs

Putting stylized numbers on it for a $1M allocation, top federal bracket plus 5 percent state, 5-year hold:

REIT path: $1M into MAA at 6.74 percent implied cap rate flows through to roughly 4.0 percent dividend yield (the gap is debt service, recurring capex, and corporate G&A; not “reinvestment”). Annual cash $40K taxed at roughly 34 percent blended (federal QBI-adjusted plus state) = $26.4K after-tax. Plus capital appreciation: assume 7 percent gross total return going forward (FTSE Nareit Equity REITs 10-yr trailing through April 2026 = 6.8 percent; Sun Belt apartment subsector lower); that’s a 3 percent annual capital gain component, taxed at 25 percent on exit = 2.25 percent after-tax appreciation. Total after-tax return: roughly 4.9 percent annualized.

Private path: $1M into a private Sun Belt multifamily deal at 5.7 percent cap rate, 65 percent leverage, 5-year hold, target deal IRR 14 percent gross of carry. Cash distributions roughly 6 percent of equity, mostly shielded by depreciation in the early years (taxable income years 1-3 near zero; after-tax cash ≈ pre-tax cash $60K). Year 5 exit at flat 5.7 percent cap and 3 percent annual NOI growth: equity multiple ~1.7x, gross IRR ~14 percent. After 20 percent carry above 8 percent pref: net IRR ~12 percent. After exit-year cap gains (20 percent federal + state) and depreciation recapture (25 percent on the depreciated portion): net after-tax IRR roughly 9.5 to 10 percent.

The 100 basis points of cap rate “advantage” the REIT appears to offer on a gross basis becomes a 460 to 510 basis point disadvantage on a fully-taxable after-tax basis. This is not because the REIT is poorly managed — MAA, UDR, and CPT are all run by capable operators with long track records. The partnership structure’s tax wrapper just delivers materially better economics for tax-paying investors in stabilized cash-flowing real estate. The credit-side instinct that’s served me well for fifteen years says: if a structure looks 100 bps better on the surface and 500 bps worse after the tax overlay, the surface number isn’t the one to trade on.

Both paths above are stylized. Actual REIT total return depends on cap-rate trajectory and capex intensity; actual private deal economics depend on sponsor execution, leverage cost, and exit market. The structural conclusion — partnerships materially outperform REITs after-tax for tax-paying investors in stabilized multifamily — survives reasonable assumption sensitivity.

If your capital is tax-exempt — endowment, foundation, sovereign wealth, qualified retirement plan — this flips. The REIT’s no-corporate-tax pass-through is itself the efficiency for tax-exempt holders, who don’t benefit from depreciation passthrough. Tax-exempt capital generally finds REITs more efficient for stabilized core multifamily and partnerships more efficient for value-add or development, where upside isn’t structurally linked to depreciation timing.

There’s a quieter point underneath the math. The partnership wrapper’s deferral-and-step-up sequence — depreciation through the hold, 1031 into the next deal, basis step-up at death — is a multi-generational tool, not a yield-pickup trick. It rewards the holder who plans to own real estate across cycles and pass it on, not the one renting exposure for a 5-year window. The post-tax math favors private structurally, but the structural reason is that the wrapper was designed for stewardship, not for trading.

What this means in practice

The decision rests on the tax status of the capital itself — whether you’re deploying your own money, family money, or capital you’re stewarding for someone else.

If your capital is taxable — own money, family office, HNW capital, anything outside a tax-exempt structure:

Private LP commitments to stabilized multifamily are typically the better risk-adjusted choice for the core allocation. REIT exposure makes sense for liquidity management, tactical positioning, or as a barbell against private illiquidity — but not as the core multifamily holding. Today’s roughly 100 bps wider public cap rate is not nearly enough to compensate for the tax disadvantage on stabilized cash flow. The math would have to flip another 300 to 400 basis points wider on the public side — REITs trading at 9.5 to 10.5 percent implied caps, a recession/distress scenario — before public becomes the obviously better choice on an after-tax basis.

If your capital is tax-exempt — endowment, foundation, sovereign wealth, qualified plan:

REITs are the more efficient vehicle for stabilized core multifamily, and private partnerships make sense for value-add, development, or market access not available in public form. The 50 to 130 bp pickup REITs offer over private today is a real yield differential after structural costs. It sits at the wider end of the historical range relative to spread vs Treasuries, and is defensible to lean into at the margin.

The framing this piece adds is that for taxable capital, the question isn’t “which one yields more pre-tax.” It’s “which structure gets me the most after-tax cash plus terminal value, given my actual tax situation, hold preference, and reinvestment options.” The post-tax framework consistently favors private partnerships for stabilized cash-flow real estate held in taxable hands. It consistently favors REITs for tax-exempt capital. Both can be true; both usually are.

What would change my mind

I’ve laid out a strong claim — that the gross cap-rate gap is misleading and the post-tax math favors private partnerships decisively for tax-paying investors in stabilized multifamily. Howard Marks has a habit I try to imitate: state the view, then state what would falsify it. Here’s where I’d flip:

REIT implied cap rates widen 300–400 bps without private clearing wider in tandem. A move toward 9.5–10.5 percent on the public side, while private stays in the 5.5–6.0 range, makes the after-tax math work for REITs even after the QBI haircut. That’s a recession/distress scenario, but it’s the cleanest invalidation.
Cost segregation gets curtailed by Treasury or Congress. The depreciation-passthrough advantage in the LP path depends on aggressive accelerated depreciation that’s been administratively friendly for a decade. A reversal — bonus depreciation phasedown that bites, or a TCJA-style rewrite that limits cost-seg — compresses the partnership advantage materially.
A specific REIT trades at a meaningful discount to NAV with idiosyncratic catalyst. Vehicle-vs-vehicle math is the general case. A 20 percent discount to NAV at one of the trio with a clear path to closure (activist, balance-sheet repair, asset sale) is a different question — that’s a single-name trade, and the REIT can win it on issuer-specific grounds even when the structural math favors private at the asset class level.
My deal flow becomes unrepresentative. The 5.5–6.0 percent private cap range is anchored in deals I see; a tightening of that range past 5.0 percent without my pipeline adjusting would mean I’m seeing a stale slice. If I notice I’m citing the same cap range three quarters in a row while market commentary moves, I’m wrong about the comparable.

If none of those break, I think the framing in this piece holds. If any of them break, the math needs redoing — not just the conclusion softened.

A note on intra-REIT dispersion

UDR’s 23 bp premium to MAA and CPT reflects real portfolio differentiation — coastal mix (NYC, Boston, SF, OC) drives a 67 percent same-store NOI margin vs 64 percent at the Sun Belt-pure names. For an active REIT book, picking between names matters; per-name deep-dives launch at Compound’s paid tier next quarter. For an LP-vs-REIT allocator, this is mostly noise — the first-order question is whether the public structure makes sense for your tax profile at all.

What we’re watching next

Q2 2026 earnings drop in late July. Three diagnostics for the public-private decision specifically:

Same-store NOI growth dispersion across the trio. If UDR continues to print growth meaningfully above MAA and CPT, the coastal premium thesis sustains and UDR’s 6.51 percent print becomes the right comparable. If the trio converges, the dispersion case weakens.
Recurring capex per unit trends. AFFO is what flows to dividends; recurring capex is what determines AFFO from NOI. Older portfolios face mounting capex that compresses FCF yields independent of NOI margin. Watch the trio’s capex disclosures.
Private clearing prints from MSCI Real Capital Analytics, RCA, Green Street commentary. If private cap rates print tighter than 5.5 percent in Q2, the public-private gap widens at the public end and the decision math shifts further in favor of private allocation for tax-paying investors. If private compresses below 5.0 percent, REITs start to become attractive even for tax-paying allocators on yield-basis-alone.

—

Compound is the bi-weekly post on bottom-up CRE analysis and macro trends, informed by operator-grade private-market context. Free bi-weekly; with deep-dives planned for the future. Subscribe at matthewd.com/letter.

—

Methodology + sources

Implied cap rate = Annualized NOI ÷ Enterprise Value. For each REIT: NOI is annualized as Q1 2026 same-store NOI from the supplemental × 4. EV is market cap (shares outstanding × close on 5/4) plus total debt (long-term debt or notes payable per the company’s XBRL filing). Cash is excluded from EV (typically <2 percent for multifamily REITs).

Same-store NOI margins are extracted directly from each REIT’s Q1 2026 supplemental disclosure deck, with full per-record source provenance:

MAA Q1 2026: Same-store NOI $328.7M / Same-store revenue $516.98M = 63.58 percent. Source: COMPONENTS OF NET OPERATING INCOME table, MAA 8-K Exhibit 99.2 filed 2026-04-29.
UDR Q1 2026: Same-store NOI $266.9M / Same-store revenue $398.65M = 66.96 percent. Matches UDR’s disclosed “Operating Margin: Same-Store Communities 67.0 percent.” Source: 1Q 2026 NOI section, UDR 8-K Exhibit 99.2 filed 2026-04-29.
CPT Q1 2026: Same-store NOI $232.4M / Same-store revenue $360.01M = 64.56 percent. Source: NET OPERATING INCOME table, CPT 8-K Exhibit 992 supplement filed 2026-04-30.

Total debt sourced from each company’s most recent 10-Q via SEC EDGAR XBRL.

Equity prices from Polygon for May 4, 2026 close.

10-year Treasury from FRED series DGS10 for May 4, 2026.

Tax framework assumptions: Top federal bracket (37 percent ordinary, 20 percent long-term capital gains), 5 percent state. REIT dividend effective rate of 29.6 percent reflects Section 199A QBI 20 percent deduction (made permanent under OBBBA 2025; verified May 2026). Private partnership analysis assumes typical Sun Belt multifamily structure: 65 percent leverage, 8 percent preferred return, 20 percent sponsor carry above pref, 5-year hold, exit at flat or modestly compressed cap rate, full cost segregation at acquisition.

*REIT total return assumption (7 percent gross going forward) is a blended estimate. Backward reference: FTSE Nareit Equity REITs trailing 10-year compound annual through April 2026 = 6.8 percent (Sun Belt apartment subsector below the index). Forward reference: implied cap rate 6.5–6.8 percent + 2–3 percent NOI growth − 1–2 percent recurring capex drag.

Private cap rate context (5.5 to 6.0 percent for Sun Belt multifamily) is from the principal’s direct deal flow at Uplift Capital, anonymized at the property and counterparty level. No specific transaction is referenced.

The structured dataset of MAA, UDR, and CPT supplemental metrics from 2010 to present (167 quarterly supplementals, 4,657 metric records, full per-record source provenance — URL, filename, page or section, raw quoted text, sha256, extractor version) is maintained internally and backs every number in this piece. Trustworthiness rated 8.5/10 in independent redteam audit (Q1 2026 trio cross-checks dollar-perfect against EDGAR XBRL). Every number cited can be traced to its source document page in under 60 seconds.

What this is not: Investment, tax, or legal advice. A recommendation to buy or sell any specific REIT or to enter or pass on any specific private deal. Tax framework assumptions vary substantially by individual situation; consult your tax advisor before allocating. A complete analysis of either public or private multifamily — both deserve more depth than a single piece can offer, and subsequent issues will dig in.