Columbia and Bahnsen file interval funds that source private assets two ways

Two interval-fund filings that arrived within days of each other take opposite routes to the same place: Columbia's growth-innovation fund would route at least 40% of its net assets exposure through funds run by Hamilton Lane, while Bahnsen's would be assembled from pooled vehicles selected manager by manager. Both landed days after the SEC voted on Sept. 30 to propose looser interval-fund repurchase rules and a 20% performance-based fee for advisers, and both will be judged by the same event — the first repurchase window each fund has to fill.

Interval funds sell a scheduled repurchase window in place of daily liquidity, so the architecture under the wrapper is a statement about how a queue would be served: hand a sleeve to one platform and the fund's capacity to pay rests on that platform's terms, while building from many pooled vehicles shifts the question to the sponsor's selection — which managers were chosen, and what happens if one has to be replaced. Those are different diligence files, and the two filings put them side by side where a gatekeeper can read them before either fund has a shareholder to answer to.

The SEC's package arrived as the backdrop, and it touches three pressure points: looser repurchase mechanics that define the wrapper, an adviser's ability to charge a 20% performance-based fee, and a companion proposal that would let CPAs and chartered financial analysts qualify as accredited investors, widening the credentialed pool of buyers for funds like these. What exactly loosens — the size of the window, how often it opens, how much notice a shareholder must give — is not spelled out in the published accounts of the vote, and that gap is a design question for anyone drafting a fund this quarter.

Columbia's filing commits at least 40% of the fund's net assets exposure to Hamilton Lane-run funds, a unit worth reading closely because the commitment is expressed as a share of net assets exposure rather than a dollar amount, leaving the sleeve's absolute size to the fund's eventual asset base, and "at least" makes it a floor rather than a target, leaving room to grow past it. For a gatekeeper deciding whether the fund belongs on a platform, that concentrates the diligence into one relationship and one set of underlying terms, the terms that would likely set the limit on what the fund could pay in a stressed window. The account describes the sleeve; it does not describe the remainder of the portfolio, which for a growth-innovation mandate is the part the adviser controls outright.

Bahnsen's filing spreads that labor across a list, building its interval fund from pooled vehicles chosen one manager at a time, which makes the sponsor's selection the product being sold. Each vehicle brings its own manager, strategy and terms for moving money, and the account of the filing does not say whether the sponsor can replace a manager once the fund is running, or how the vehicles' own liquidity lines up against the fund's repurchase window; those are questions a gatekeeper would want answered before the first queue forms rather than after. A list can look diversified and still behave as a single position if its vehicles lean on the same kind of exposure, which makes the sponsor's combining work as consequential as its picking work.

The two shapes carry different risks into a stressed quarter, because a concentrated sleeve concentrates the negotiation: the sponsor sits across from a single platform that can say what its funds will pay and when. A fund of pooled vehicles disperses that negotiation, turning the sponsor's ability to substitute one manager for another into the fund's real liquidity tool and turning the selection record into what investors are underwriting — a job the proposed performance fee would make explicit.

Two filings make a small sample, and the accounts describe their architecture without arguing for it, but the choice is legible: routing exposure through a sleeve buys private-asset sourcing from a firm whose business is sourcing, and building a list buys it with the sponsor's own judgment. The SEC's proposals change the price of both in one direction — an easier repurchase rule makes the wrapper simpler to operate, and a performance fee makes the sponsor's skill at picking a platform or picking managers what it gets paid for.

Ares Strategic Income Fund gave the quarter its clearest reading: investor requests in the third-quarter window equaled 11.6% of the fund, and the fund limited repurchases rather than meeting them in full. The summary it published omits both the payout split — the share of those requests that got paid — and new portfolio-quality figures. The request rate says how many shareholders wanted out; the fill rate says how many did, and only one of those two numbers is public.

PWD's tracking has also reported Ares prorating a third time, paying about $515 million of a $1.35 billion redemption queue while September subscriptions ran at $8 million, with the balance sheet absorbing the difference. That gap decides which semi-liquid vehicles keep their shelf space, and it is the arithmetic a sponsor takes on when it files a fund promising periodic liquidity. Apollo Debt Solutions showed the same shape at a different scale, capping its third-quarter tender at $700 million against requests equal to 14.7% of the fund, which leaves the queue intact and the next window to settle the fill rate.

The disclosure trail is thinner than the stakes, and third-quarter nontraded BDC repurchase data left eight funds yet to report while Ares's summary withheld both the payout split and refreshed portfolio-quality marks. For an interval fund, those are the two numbers that describe what happened when a window opened, and a sponsor that publishes neither leaves the wrapper to be judged by its request rate alone. Neither the Columbia fund nor the Bahnsen fund has a queue to report, and neither will for some time, which is why the architecture is the only evidence available: a filing is a sponsor's forecast of what its own window will look like, written down and made public before any performance exists to argue with it. Both were filed in the same week the SEC proposed to loosen the mechanics that govern the moment the forecast is tested.

A performance fee of 20% changes what a sponsor is paid for, and the two architectures wear it differently: on Columbia's sleeve, the adviser's fee rides on a portfolio Hamilton Lane's funds largely choose, and whether that fee sits on top of the underlying funds' own charges is a question the filing's account does not answer. On Bahnsen's fund of pooled vehicles, the sponsor is charging for selection, and the case it has to make is that the selection earns its keep alongside whatever the vehicles themselves cost.

Multi-manager sleeves and funds of funds are not new; the price list hanging off them is, because a performance fee tied to results, proposed in the same week as looser repurchase mechanics and a wider accredited-investor definition, gives a sponsor a reason to build funds whose returns carry the economics rather than funds whose asset gathering does.

The accredited-investor item bears on the architecture argument more than it first appears, because if CPAs and chartered financial analysts can qualify by credential, the roster of professionals able to discuss these funds with clients widens, putting distribution capacity under the same pressure as portfolio construction. A sponsor building from pooled vehicles has to explain a list to whoever is doing the explaining to clients; a sponsor routing through one platform has to explain a relationship, and neither is the shorter conversation, with the proposal pointing toward having both of them more often.

The number that settles it is a fill rate: Ares reported 11.6% of its fund asking out in the third quarter and limited what it paid, and the payout split showing how much of that queue cleared is the figure its summary withheld. Columbia's fund and Bahnsen's fund have no such number yet, but when one publishes a fill rate the comparison runs back to two questions — how many shareholders wanted out, and how many the sponsor could afford to let go.