Text or WhatsApp (808) 600-9260

Separate Accounts, Funds & Direct Investing for Institutions

Institutions invest in real estate through commingled funds, separate accounts, joint ventures and direct holdings, and most programs use more than one. The vehicles differ mainly in who holds the decisions: in a commingled fund nearly all control sits with the manager, while separate accounts, majority-investor joint ventures and direct holdings leave more of it with the institution. Survey data shows funds and control structures growing side by side, and a single large public plan can hold most of its real estate in separate accounts. The useful test for any structure is which decisions the institution can initiate and which it can block.

Who holds the decisions in a fund, a separate account, a venture and a direct holding

Control moves toward the institution as the vehicle becomes more dedicated. In a commingled fund nearly all control is delegated to the manager, and the institution’s real levers are the decision to commit and the terms negotiated before capital is called.

Separate accounts and majority-investor joint ventures are classed as controlled structures. Commingled funds in which the institution holds a minority interest are typically classed as non-controlled. The line follows who can direct the asset, which is why a small stake in a large pool sits on the opposite side from a majority stake in a venture that owns one building.

A workable test follows an investment from acquisition through financing, capital spending and sale. At each step the analysis records who can propose the decision and who can block it. Ownership percentage cannot answer those questions, and a right to receive reports differs from a right to withhold consent.

The categories also overlap. A joint venture can sit inside a separate account relationship, so calling a holding direct says little about how much discretion the operating partner keeps. Two institutions can both report direct exposure and have very different approval rights.

Planned allocations and actual portfolios measure different things

In one survey series, institutions reported preferring direct investments, along with joint ventures and separate accounts, because these can provide greater control and tailored strategies.

The latest edition of that series shows 81 percent of institutions planning allocations to closed-end funds. In the same edition, planned allocations to direct investment and to joint venture and separate account structures reached five-year highs.

Both results can be true at once because the vehicles are not mutually exclusive. A plan can keep fund commitments for part of its allocation and add controlled structures for another part.

A survey of intentions also differs from a ledger of holdings. Planned allocations say where new capital is headed. They say nothing about the net asset value already sitting in each vehicle, which is the subject of the next section.

Separate accounts as a primary business model at a large public plan

Separate accounts are the primary real estate business model at one large public pension plan, with operating companies as a secondary focus. Its strategic plan also moved toward fewer, larger external-manager relationships.

In its latest reporting, separate accounts represented 72.3 percent of real estate portfolio net asset value, while commingled funds represented 6.8 percent. Those two figures describe one portfolio and cannot stand in for institutions generally.

Fewer, larger relationships change the counterparty count and the size of each mandate. They say nothing by themselves about property diversification, because a single manager can run a dedicated account across many markets or concentrate it in a few.

Portfolio construction and asset approval operate at different levels. An investor can approve every acquisition and still end up with a collection of assets that conflicts with the intended allocation. Separate accounts and other bespoke structures can give an investor more control over portfolio construction than commingled funds or individual co-investments, because the mandate itself is written around the institution’s targets.

The practical test is what happens when an acquisition meets the deal criteria but shifts the overall allocation. A mandate that separates ordinary investment decisions from changes needing investor approval answers that question in advance. A mandate that does not leaves the point to negotiation at the worst moment.

What retained discretion looks like in a direct program

In a case study of one relatively large state pension plan, about 70 percent of real estate net asset value sat in direct vehicles. Staff kept discretion over overall strategy and over the most important portfolio and asset-level decisions, which cover buying and selling as well as recapitalizing.

Recapitalization deserves its own line in any control analysis. It changes the capital arrangement around an asset already owned, such as adding or replacing equity or debt, so authority over the original purchase does not settle who decides it.

Joint ventures and separate accounts can offer institutions more control than commingled funds with limited control, and real estate operating companies can offer similar discretion.

Retained discretion concerns decisions, and property-level execution can still sit with an outside operator. Monitoring produces information, while decision authority determines who can act on it. A strong reporting package leaves the manager in charge if the documents give the institution no approval right.

“With about 70% of real estate net asset value (NAV) held in direct vehicles, staff retains discretion for both overall investment strategy and the most important portfolio and asset-level decisions (buying, selling, and recapitalizing).”

Academics Question the Value of Private Real Estate Funds: What’s an Investor to Do?, RCLCO

Co-investments and sidecars between the two poles

Institutional programs can add co-investments and pre-approved co-invest capital to a fund base, along with sidecar vehicles. Some also build programmatic joint ventures, invest in operating companies and provide co-GP capital. Each is a different route into a manager relationship, with its own answer to who decides.

A co-investment is a decision about one opportunity, made alongside a fund commitment that delegated authority across a whole mandate. The fund decision rests on the manager and the terms. The co-investment decision rests on the asset and the price.

Seeing a deal and receiving an allocation in it are separate from holding approval authority over it. Documents that give only the first two leave the institution with no say in the asset after it participates.

Individual co-investments also have a structural limit. Each one is a one-off, so a series of them does not add up to authority over the manager’s broader portfolio. That gap is part of why separate accounts and bespoke structures can address limitations of both commingled funds and single co-investments.

Fund versus direct is therefore an incomplete question. An institution can delegate part of its allocation through funds and keep more discretion elsewhere. The analytical task is to list which decisions each piece delegates and whether the pieces fit together.

Worked example: a hypothetical fee comparison on $100 million held five and then seven years

Two hypothetical structures hold the same single asset with the same capital. Each starts with $100 million of equity, with no borrowing and no interim distributions, and sells at the end of the hold. Carried interest and all other costs are left out so that only the fee base and the hold length vary.

The fund charges 1.25 percent a year on committed capital of $100 million. The separate account charges 0.80 percent a year on invested capital of $100 million. Both bases equal $100 million because the capital is fully invested from the first day. Fees are totaled and deducted at sale, with no timing effect. The fee rates are illustrative assumptions and carry no implication that separate accounts cost less.

Base case: a five year hold with gross sale proceeds of $140 million. The fund fee is $100 million multiplied by 1.25 percent, or $1.25 million a year, which is $6.25 million over five years. Net gain is $140 million minus $100 million minus $6.25 million, or $33.75 million, a 33.75 percent total return on capital. The separate account fee is $100 million multiplied by 0.80 percent, or $0.8 million a year, which is $4 million over five years. Net gain is $140 million minus $100 million minus $4 million, or $36 million, a 36 percent total return. The separate account finishes $2.25 million ahead.

Stressed case: the sale slips to seven years and gross proceeds are $145 million. The fund fee is $1.25 million multiplied by seven, or $8.75 million, so net gain is $145 million minus $100 million minus $8.75 million, or $36.25 million, a 36.25 percent total return. The separate account fee is $0.8 million multiplied by seven, or $5.6 million, so net gain is $145 million minus $100 million minus $5.6 million, or $39.4 million, a 39.4 percent total return. The gap is now $3.15 million.

The longer hold added $5 million of gross gain, which is $145 million minus $140 million. The fund kept $2.5 million of that, since net gain went from $33.75 million to $36.25 million. The separate account kept $3.4 million, since net gain went from $36 million to $39.4 million. The gap between the two widened by $0.9 million, from $2.25 million to $3.15 million, because each added year carries a fee on the full base.

The mechanics matter more than the rates. Both fees here scale with time because the base stays at $100 million throughout. A committed-capital base would keep charging on money already returned, while an invested-capital base would step down as capital comes back. The example avoids that difference by holding the base constant, and real documents settle it by their wording.

Mistakes that distort comparisons of institutional structures

Equating a majority economic interest with control confuses ownership and governance. Majority-investor joint ventures are classed as controlled, yet the classification does not list which approvals the investor holds. Correct analysis names the decisions reserved to the investor and those left to the operating partner.

Reading planned allocations as current holdings mixes two measures. A survey in which 81 percent of institutions plan closed-end fund allocations reports intentions. A portfolio with 72.3 percent in separate accounts and 6.8 percent in commingled funds reports net asset value at one institution. Neither describes the other, and neither describes the market.

Comparing fee rates without their bases gives a misleading answer. In the example above, 1.25 percent on committed capital and 0.80 percent on invested capital produced fee totals of $6.25 million and $4 million over five years, and $8.75 million and $5.6 million over seven. The dollar gap grew with the hold, which a rate comparison alone would never show.

Treating reported outperformance of controlled structures as proof that control caused it overreaches. The pattern is an observation from practice, and the structures also differ in strategy and in manager. Correct analysis keeps the observation separate from the proposed mechanism.

Treating co-investment access as portfolio construction authority overstates one opportunity. Participating changes the institution’s own exposure and gives no control over the manager’s other investments. Authority over the surrounding portfolio sits in a mandate, and a deal-by-deal right does not create one.

Checklist questions for comparing funds, separate accounts and direct holdings

Which decisions stay with the institution after capital is committed? The answer separates retained authority from manager discretion for the whole life of the investment.

Does the mandate govern the overall allocation as well as each investment? Portfolio construction concerns the composition of the whole, even when every asset meets its own criteria.

Who can approve or block a sale? Sale authority fixes who decides when an investment leaves the portfolio.

Who controls a recapitalization? Authority over a change in the capital arrangement can differ from authority over the original purchase.

What does a co-investment right provide?

Is the interest a majority or a minority position? Majority-investor ventures and separate accounts are classed as controlled, while minority fund interests typically are not.

Who inside the institution can exercise each reserved right? Internal authority connects a contractual approval to an identifiable decision maker.

On what base is each fee charged? A fee on committed capital and a fee on invested capital produce different dollar amounts for the same percentage rate.

Sources
  • CalSTRS publication (link)
  • Rethinking Closed-End Funds, AFIRE (link)
  • Institutional Real Estate Allocations Monitor, Hodes Weill (link)
  • Real estate strategic plan, CalPERS (link)
  • CalPERS publication (link)
  • Co-Investment Sentiment: RCLCO Real Estate Consulting (link)
  • Academics Question the Value of Private Real Estate Funds: What’s an Investor to Do?, RCLCO (link)
  • Institutional Investment in US Real Estate: How to Maximize Performance, RCLCO (link)
  • Park Madison Partners (link)
  • Pitch Decks: 31 Investor Marketing and Pitch Assets, book by Richard C. Wilson
  • The Single Family Office, book by Richard C. Wilson
  • Centimillionaire Strategies, book by Richard C. Wilson

Talk to Richard