The Buy/Build/Sell Equation: How Real Estate Fund Managers Evaluate Deals and Why Returning Capital Sets the Pace
Every real estate fund manager runs the same machine. Buy assets through a defined investment period. Build value through operations, development, or capital structure. Sell into the next cycle's bid. Distribute the proceeds back to LPs. Use the track record of those distributions to raise the next fund. Repeat for thirty years until the franchise either compounds into a multi-fund platform or quietly winds down.
The mechanics look simple. The constraint is not. The number that determines whether the next fund gets raised is not the IRR on the marketing slide. It is DPI, distributions to paid-in capital, the metric that says how much of the LPs' money has actually come back as cash. A fund showing a 22 percent IRR on paper but a 0.4x DPI six years into a seven-year fund life is, for any institutional LP committee that knows what it is doing, an unraisable manager. The promote that exists in the model has not been delivered to anyone, and the model is not the trade.
This article is the operating math of the buy/build/sell cycle, the deal-evaluation framework that fund managers actually run, and the structural constraint, return of capital before the next raise, that sets the cadence of the entire industry.
1. The Fund as a Repeating Machine
A real estate private equity fund is structured around a defined life, typically seven to ten years, with three distinct operational phases that map directly onto the buy/build/sell cycle.
| Phase | Years | Capital flow | Activity |
|---|---|---|---|
| Investment period | 1 to 3 | Capital calls into LPs | Sourcing, underwriting, acquisitions |
| Hold period | 2 to 6 (overlap) | Minimal flows | Operations, value creation, asset management |
| Harvest period | 4 to 7 | Distributions to LPs | Sales, refinancings, promote crystallization |
| Wind-down | 7 to 10 | Final distributions | LPA expiration with optional 1+1 year extensions |
The investment period (years 1 to 3 typically) is when the fund deploys capital, calling LP commitments as deals close. The hold period (overlapping years 2 to 6) is when the GP executes the operational thesis. The harvest period (years 4 to 8 typically) is when assets sell, capital returns to LPs, and the GP's promote crystallizes if the fund hits its hurdles. The fund formally winds down between years 7 and 10, with extension options usually built into the LPA.
The cadence is not optional. It is the structural commitment the GP made to the LPs at fundraising, and it governs every deal-level decision through the life of the fund. A GP who commits to a 7-year fund life and then routinely holds assets to year 9 is operating outside the contractual mandate, even if the LPA permits the extension. The reason most LPs hate extensions is that the fund-vintage clock keeps running, the deployment of fresh capital is paused, and the GP's promote calculations stretch in ways that are rarely beneficial to the LP.
2. The Vintage Concept and Why It Matters
Funds are referred to by vintage, the year of their first close, because the cycle environment in that year structurally shapes the fund's outcomes. A 2007 vintage and a 2010 vintage, deploying into the same asset class with the same GP, produce returns that can differ by 1500 basis points of IRR purely as a function of when the capital was called.
Vintage diversification is therefore the central LP allocation discipline. A pension fund with a $1 billion real estate allocation does not write a $1 billion check to one fund. It writes $50 million to twenty managers across multiple vintages, smoothing the cycle exposure. The implication for the GP is structural. To remain in the LP's allocation rotation, the GP needs to be back in market every two to three years with a new fund, which means the prior fund needs to have demonstrated enough deployment, performance, and capital return for the LP to underwrite the re-up.
This is the hidden treadmill that drives the GP's behavior. A GP raising a new fund every three years is calling capital from Fund III while Fund II is mid-execution and Fund I is harvesting. The performance of Fund I, expressed not as IRR but as DPI, is the most important data point in the Fund III pitch. An LP underwrites the re-up by looking at three numbers: how much capital did the prior fund return, how fast, and at what multiple. IRR is the headline. DPI is the diligence.
3. The Buy Phase: How Fund Managers Actually Evaluate Deals
The deal-evaluation funnel inside a serious institutional GP is structured to compress thousands of opportunities into the eight to fifteen acquisitions that will populate a single fund. For a representative $500 million fund, the annual flow looks roughly as follows:
| Stage | Annual volume | Filter applied |
|---|---|---|
| Sourced opportunities | ~2,000 | Market, product, basis |
| Underwriting candidates | ~200 | Full model, market study, IC pre-screen |
| Active LOIs / bids | ~50 | Sponsor-approved offers in market |
| Closed acquisitions | ~10 to 15 | Awarded and closed transactions |
| Portfolio assets across fund life | ~30 to 50 | Cumulative across 3-year investment period |
The screening criteria at each stage are explicit and largely consistent across institutional managers.
Preliminary screen. Does the asset fit the fund's investment mandate (product type, geography, size, return profile)? Is the basis defensible against current comparables? Is the seller motivated enough to transact at a basis that pencils? Is the local market demonstrating the demographic, employment, or supply trends that support the fund's thesis? Roughly 90 percent of sourced opportunities are eliminated at this stage in under two hours of analyst time per deal.
Full underwriting. The remaining 10 percent receive a complete model, market study, and physical due diligence assessment. The model output that gates the next stage is generally a combination of: leveraged IRR (typically 14 to 20 percent for value-add, 18 to 25 percent for opportunistic), equity multiple (1.7x to 2.2x for value-add, 2.0x to 2.5x for opportunistic), yield-on-cost spread to current cap rates (typically 100 to 200 bps minimum), and breakeven occupancy (typically below 75 to 80 percent for the asset to service debt without operating distributions).
Investment committee. The IC is the GP's internal gating body, typically composed of senior partners, the head of acquisitions, the head of asset management, and the CIO or equivalent. The IC reviews the underwriting, challenges the assumptions, and either approves a binding offer or sends the deal back for further work. A serious IC kills 60 to 80 percent of fully underwritten deals, often on basis or on capital-stack mismatches that the deal team had not adequately stress-tested.
Bid and award. The deals that survive IC enter the bid process, where the GP competes against other capital. The win rate at this stage varies by market and product, but the institutional median runs 15 to 25 percent of bids. Higher win rates often correlate with overpaying. Lower win rates often correlate with underwriting that is too conservative for the cycle.
The overall conversion from sourced opportunity to closed acquisition is roughly 0.5 to 0.75 percent. A $500 million fund that needs to deploy across 30 deals over a three-year investment period is screening between 4,000 and 6,000 opportunities to get there. The implication for the LP is that GP origination capability is itself an underwriting input. A GP without a top-of-funnel pipeline cannot be selective at the bottom of the funnel, regardless of how disciplined the IC process appears.
4. The Underwriting Hurdles That Actually Matter
Most LP marketing materials show projected IRR. The IRR is not the underwriting hurdle. It is the output of a model that hit its hurdles. The underwriting hurdles that actually gate a deal are five.
| Hurdle | Typical threshold (value-add multifamily) | What it tests |
|---|---|---|
| Untrended yield-on-cost | 6.0% to 6.5% | Operating performance against current basis |
| Yield-on-cost spread vs. current cap | 100 to 200 bps | Development / improvement premium |
| Debt service coverage ratio | 1.25x to 1.40x at stabilization | Ability to service debt at projected NOI |
| Breakeven occupancy | 75% to 80% | Resilience to demand shock |
| Equity multiple | 1.7x to 2.0x | Total dollars returned per dollar invested |
A deal that fails any one of these hurdles, even if the projected IRR looks attractive, generally does not survive a serious IC. The reason is that IRR can be manufactured by extending hold, manufactured by levering up, or manufactured by aggressive exit cap assumptions. The five underlying hurdles are harder to manipulate in the model without producing visibly aggressive assumptions elsewhere.
The hurdle that has tightened most aggressively since 2022 is the yield-on-cost spread. In the 2018 to 2021 cycle, sponsors routinely underwrote deals with a 50 to 100 bps spread between yield-on-cost and current cap rates, on the implicit assumption that cap rates would compress further. Post-2022, that assumption has reversed. The current institutional discipline requires a 150 to 200 bps spread minimum, and many GPs have widened to 200 to 250 bps for new-construction development. The deals that pencil under that discipline are dramatically fewer than the deals that penciled under the old discipline. That tightening is what the industry calls "discipline returning to underwriting." It is also what is driving deal volume down by 40 to 60 percent against the 2021 peak.
5. The Build Phase: Where Operational Alpha Actually Lives
The build phase is where the GP earns the promote, in the sense that the value created between basis and exit is the GP's actual product. The build is not always literal construction. For a value-add multifamily acquisition, the build is unit renovation, common-area upgrade, expense optimization, and rent push. For an opportunistic distressed retail acquisition, the build is releasing, rebranding, and capital structure restructuring. For a ground-up development, the build is the construction execution itself.
The operational levers vary by strategy, but the structure of the alpha is consistent. A serious GP's underwriting model decomposes the projected NOI growth into specific drivers, each underwritten separately, each tracked monthly through asset management.
A typical value-add multifamily build decomposition:
| Driver | Contribution to NOI growth | Time to realize |
|---|---|---|
| Unit renovation premium | 35% to 50% | 18 to 30 months |
| Expense optimization (insurance, taxes, utilities) | 15% to 25% | 6 to 18 months |
| Common-area and amenity upgrade | 10% to 20% | 12 to 24 months |
| Operational mgmt upgrade (concession discipline, fee income, lease-up) | 15% to 25% | 6 to 12 months |
| Market rent growth | 10% to 25% | continuous |
The non-obvious point is that each driver carries a different risk profile, and a deal that depends heavily on a single driver is a more fragile deal than the headline IRR suggests. A value-add thesis where 60 percent of projected NOI growth comes from unit renovation premium is, in effect, a leveraged bet on the renovation comp set holding. A thesis that distributes the same NOI growth across five drivers is a fundamentally more diversified trade, and it is the trade that survives a market that does not cooperate with the renovation comp.
Asset management discipline during the build phase is what separates the GPs who hit their underwriting from the GPs who do not. The data is unambiguous. Across institutional value-add multifamily deals over the past two decades, GPs with tight monthly variance tracking, dedicated asset managers per deal, and explicit driver-level dashboards outperformed median GPs by 200 to 350 basis points of levered IRR. The build is not a passive period. It is the period during which most of the deal's actual value creation happens, and the firms with the operational infrastructure to execute it pull away from the firms that treated asset management as a back-office function.
6. The Sell Phase: Timing, Process, and the DPI Imperative
The sell phase is where the GP returns capital. It is also where the LP/GP alignment is most visibly tested.
A disciplined sell process inside an institutional fund follows a predictable cadence. The asset hits stabilization, typically 18 to 36 months after acquisition for value-add. The asset is held through stabilization plus a year or two of stable operations to demonstrate the new NOI run-rate to the buyer market. The asset is marketed through a defined broker process, typically with two to four institutional brokers competing for the listing. The asset closes within four to nine months of listing under normal market conditions.
The decision the GP faces at the sell point is the timing one. Holding longer captures more cash flow but extends the IRR drag. Selling earlier crystallizes promote and returns capital but caps the upside. The institutional discipline most GPs cite is the equity multiple plateau rule. If the projected equity multiple from holding another year is less than 1.10x against the current marked value, the asset is sold. That rule embeds a roughly 10 percent forward-return hurdle to justify continued ownership, which corresponds approximately to the underlying levered IRR profile the LPs underwrote.
The GP's promote crystallization mechanics during sell make this decision economically loaded. A GP whose deal is below the 8 percent pref hurdle has no promote at sale, and the cash fees collected during the hold are the only GP economics. A GP whose deal has cleared the pref but is still below the catch-up hurdle is leaving meaningful promote on the table by selling. A GP whose deal has fully cleared into the upper waterfall tiers is capturing maximum promote per incremental dollar of value, which often pushes the GP toward a sell decision earlier than the LP would prefer.
The DPI consequence of sell discipline is the structural constraint on the GP's franchise. A GP who delays sells across a fund to maximize per-deal IRR often produces a fund-level DPI curve that is too slow for the LP to underwrite the re-up. A GP who sells too aggressively produces a fund-level IRR that is below the marketing target. The sweet spot is the cadence that returns 60 to 80 percent of paid-in capital by the end of year five, with the balance returning by year seven, and the residual upside crystallizing through years eight and nine.
DPI benchmarks for an institutional value-add real estate fund:
| Year of fund life | Strong fund (target) | Median fund | Weak fund (re-up risk) |
|---|---|---|---|
| Year 2 | 0.0x | 0.0x | 0.0x |
| Year 4 | 0.5x | 0.3x | 0.1x |
| Year 6 | 1.2x | 0.8x | 0.4x |
| Year 8 | 1.9x | 1.4x | 0.7x |
The 1.0x DPI line is the soft re-up trigger most LPs apply. A fund tracking on the strong-fund line is back in market on schedule. A fund tracking on the weak-fund line is generally not raisable into the next vintage.
Most LP investment committees apply a soft re-up rule that requires the prior fund to have hit DPI of 0.5x to 1.0x before committing to the next vintage. A GP who fails that threshold by the time of the next raise is, in the LP's underwriting, an unraisable manager regardless of paper IRR. The constraint is structural, and it is the constraint that drives more GP decision-making than most LPs realize.
7. The Constraint That Ends Careers: Returning Capital Before the Next Raise
The fundraising cadence and the capital-return cadence are coupled in a way that creates a hard constraint on every GP's franchise.
A GP raising Fund III with Fund I still mid-life and Fund II early in its investment period is being underwritten on Fund I's DPI. A GP whose Fund I DPI is below 0.5x at the time of the Fund III raise is, in most institutional re-up workflows, a problem. The LP's investment staff will have flagged the slow capital return. The LP's investment committee will have asked why the manager is not returning capital. The LP's allocation team will have downsized the proposed re-up commitment. In severe cases, the LP will pass on the re-up entirely, even if the manager's IRR mark on Fund I is acceptable.
This is the structural reason most institutional fund managers actively manage their DPI curve, sometimes at the expense of marginal IRR optimization. A GP who could hold an asset another two years for an additional 200 bps of IRR will often sell into a slightly suboptimal market because the alternative is a Fund III that does not raise. The LP's preference for capital return over paper performance is, in practice, the binding constraint on the entire industry's hold-period behavior.
The mechanical workarounds to this constraint are limited.
Continuation vehicles are the most prominent. A GP nearing the end of Fund I's life with assets that are not yet ready to sell into the open market can offer the assets to a continuation vehicle, typically capitalized by a different LP base (often a secondary market specialist plus a roll-over commitment from existing LPs). The continuation vehicle pays Fund I, Fund I distributes to its LPs, the GP's DPI on Fund I improves dramatically, and the assets continue to be managed by the same GP under a new fund structure. The continuation vehicle has become a roughly $30 to $60 billion annual market segment in real estate alone over the past five years, growing rapidly as the post-2021 cycle has produced an inventory of assets that cannot be sold cleanly into the open market.
Recycling provisions allow the GP to redeploy returned capital into new investments during the investment period, rather than distributing immediately. Recycling is LP-favorable in concept, since it avoids the tax recognition event of a quick distribution and redeployment, but it does not improve DPI. A GP relying heavily on recycling during a tough exit cycle is, in DPI terms, identical to a GP who has not returned capital at all.
Dividend recapitalizations through refinancing can return capital to LPs without selling the asset. A successful refi during the hold period can return 30 to 60 percent of paid-in capital while the asset continues to compound. The DPI math improves immediately. The asset's leverage profile, however, has worsened, which means the residual equity is more exposed to subsequent cycle moves. Dividend recaps are LP-friendly in good markets and LP-punishing if the cycle turns before the eventual sale. GPs who relied heavily on recap-driven DPI in the 2018 to 2021 cycle are some of the worst-performing GPs of the post-2022 reset.
Strategic secondary sales of LP interests are a separate workaround in which existing LPs sell their fund interests to secondary buyers, returning capital without the underlying assets transacting. The discount to NAV is often 15 to 30 percent depending on the cycle. The GP's DPI is unaffected (since the cash flows did not change), but the LP's effective IRR is improved by the shorter duration. Sophisticated LPs treat secondary sales as a portfolio-management tool, particularly during denominator-effect cycles when their RE allocation is over the target weight.
8. The Denominator Effect and the Allocation Pressure
The 2022 to 2024 cycle introduced a constraint most GPs had not previously had to navigate. When public equity markets declined sharply in 2022 while private real estate marks held up (because they reflect appraisal-based valuations that lag the public markets), institutional LPs found themselves over-weighted in real estate as a percentage of total portfolio. A pension fund with a 10 percent target real estate allocation suddenly carried 14 percent because the public-equity denominator had shrunk.
The mechanical response was a freeze on new commitments to real estate, often lasting 12 to 24 months. GPs raising Fund III in late 2022 or 2023 walked into a market where their best LPs could not commit, even when the fund's track record was strong. The denominator effect compressed a generation of GP fundraising timelines and caused several mid-tier managers to either skip a vintage or wind down their franchise entirely. The Fund III raises that succeeded in this window typically required either pre-existing LP relationships willing to commit through the denominator window or a fundamentally differentiated strategy that LPs could underwrite as additive to the rebalanced portfolio.
The denominator effect compounds with the DPI constraint in a punishing way. A GP whose Fund I DPI is slow at the same time their LPs are over-weighted in RE allocation faces a fundraising environment that is structurally hostile on two dimensions simultaneously. Most of the GP attrition in the 2022 to 2025 cycle has come from this overlap, not from individual deal performance.
9. The Subscription Line and the IRR Distortion Question
A subscription line is a credit facility provided to a private fund, secured by the LP's unfunded commitments, that allows the fund to fund deal acquisitions with debt rather than calling LP capital immediately. The capital is called later, often quarterly, to pay down the line. The mechanic was relatively obscure before 2010 and is now nearly universal in institutional private real estate funds.
The economic effect on IRR is significant. By delaying capital calls, the subscription line shortens the period over which LP capital is at work, which raises the IRR on the LP's invested capital without changing the underlying deal economics. A fund that produces a 13 percent IRR with no subscription line might report a 16 to 18 percent IRR with aggressive subscription line usage, on identical underlying deal performance. The promote calculation, which is typically anchored to LP IRR thresholds, is meaningfully easier to clear under sub line conditions.
This creates a real underwriting question for LPs. A fund's headline IRR may not be comparable to a fund's IRR from a different vintage or a different manager, because subscription line usage varies. The institutional discipline most allocators now apply is to evaluate funds on both IRR with sub line and IRR ex-sub line, and on equity multiple (which is unaffected by sub line timing). A GP whose performance separation between the two metrics is large is a GP whose paper returns are partially financial engineering. A GP whose performance is consistent across both metrics is a GP whose alpha is real.
10. The Closing Equation
The buy/build/sell cycle, viewed as a fund manager's operating equation, has five inputs and one constraint.
The inputs:
- Deal flow quality (top-of-funnel pipeline strength)
- Underwriting discipline (the five hurdles, applied without flexing)
- Build execution (the operational alpha during the hold)
- Sell timing (the equity multiple plateau decision)
- Capital structure (leverage and waterfall tiers calibrated to the cycle)
The constraint: DPI must clear approximately 0.5x to 1.0x by the time the next fund is raised, or the next fund does not raise.
Most GPs optimize the inputs and treat the constraint as something that resolves automatically if the inputs are right. The cycle's hardest lesson, repeatedly relearned, is that the constraint can bind even when the inputs are managed well. A GP who buys disciplined, builds well, and sells at the right time can still produce a slow DPI curve if the cycle does not provide an exit window. The structural reality is that capital return depends on a buyer market the GP does not control, and the absence of a buyer market in years five through seven of a fund's life is the single most common reason institutionally credible managers fail to raise the next vintage.
The successful long-duration franchises (Blackstone, Brookfield, Ares, Lone Star, KKR's real estate platform, the regional specialists who have compounded across thirty years) have built operational structures that respect the constraint as primary, not residual. They sell when they need DPI, not when the model says peak return. They use continuation vehicles, dividend recaps, and recycling provisions strategically rather than reactively. They communicate with their LPs about DPI cadence with the same discipline they communicate about IRR. Most importantly, they raise the next fund when the prior fund's DPI curve is on track, not when the IRR mark is highest.
That is the equation. Buy at a basis that pencils. Build through driver-level execution. Sell at a cadence that returns capital. Raise the next fund into an LP base that has been kept whole by the prior fund. Repeat for thirty years. The franchise that does this becomes a multi-fund platform. The franchise that misses the constraint, even once, almost never recovers.