For a first check into a debt fund, is an open end structure with a gate or a three year closed end fund the safer one?

The gate is not the scary part, lantern. A gate is what keeps the fund from selling loans at a discount to pay the first people who panicked, and if it fires, everyone still in the fund benefits from it. What deserves three readings is the order of the queue and how partial fills work. Pro rata across all requests, or first in first out? Pro rata means a large institution asking for 40 million and you asking for 25k get the same percentage, which is fine. First in first out means the people with the fastest lawyers are ahead of you. Also ask what happened in the worst quarter they have had. Whether the gate has ever been used, and what the queue looked like the month after, tells you far more than whether it exists.

What is a gate? Is it a penalty for leaving early, or is it the fund just saying no? I'd rather ask than read this thread twice and learn nothing.

And if the closed-end one has "no redemptions at all," what happens if I need the money in year two? Do I just not get it?

Ember, a gate is the fund saying no, or saying not all of it. It caps how much money can leave in a quarter, usually as a percentage of the fund's net asset value. So if you ask for your whole 25k and the gate limits redemptions that quarter, you might get a slice and stay in line for the rest. It isn't a fee.

And on the closed-end, yes, mostly you don't get it. Some funds let you try to sell your interest to another investor with the GP's consent, which in practice means a discount and a slow process, and it varies with what the LPA actually says.

Which is why I lean closed-end for money I've already decided is asleep. I hold land for eight years at a time. A three year lock doesn't frighten me the way it frightens people who might need the cash. The lock only hurts if you mispriced your own timeline.

Opposite vote from arbor, and for a boring reason. For an investor who wants income without a second job, a closed-end fund creates a job every three years. At the end of the term the capital comes back and the investor has to find the next fund and diligence the next manager, sitting in cash the whole time. Do that four times and a decade has gone to re-underwriting the same decision. The open-end one just keeps paying. If the investor wants out, notice goes in and the position joins the line. Managing a queue is easier than managing a reinvestment date.

Marlow's reinvestment point is the one that changed my thinking, because the closed-end return you get quoted isn't the return you live with. Three years at 9 and then four months in a savings account while you pick the next one is not 9. The gap between the two is real money and nobody puts it in the deck.

Does the fee load differ between the two structures? I'd assume the closed-end has a promote over a hurdle and the open-end just charges a flat management fee, but I've only skimmed one of each and I might be inventing that.

Beacon, I've seen promotes in both, the closed-end ones just measure against the whole fund life and the open-end ones measure per period, which matters more than it sounds.

My actual reason for liking closed-end comes from the renovation side. Every project I've run, the thing that made me finish was a date. When there's no date, a bad loan becomes an extension, and an extension becomes a modification, and eighteen months later it's still on the books at par because nobody had to test it. A term is the forcing function. Open-end funds can carry a problem loan basically forever and mark it however their valuation policy allows.

In an open-end fund you buy and sell at a NAV somebody else calculates. Your entry price and your exit price are both the manager's opinion of what the loan book is worth. In a closed-end fund there is no NAV transaction, you get whatever the loans actually pay when they pay it. That cuts both ways. If the manager's marks are honest, open-end is fine and more flexible. If they are generous, the person redeeming early is taking value from the people who stay, and the person subscribing today is overpaying. So the open-end question is really who does the valuation and how independent they are. Harrow's forcing-function point is the same problem from the other end.

Reading kestrel's post I realized I'd been treating NAV like a fact and not a number someone produced.

I read both kinds of documents twice this month and the gate language is not standardized at all, which surprised me. One said redemptions "may be suspended in the manager's sole discretion," full stop, no percentage, no test, no time limit. The other spelled out a 5 percent cap, pro rata fills, and a requirement to disclose the queue length to all investors quarterly.

Those are both described as "quarterly liquidity with a gate" in the summary. They are not the same product. If I ended up in the open-end camp I'd want the second one and I'd want the queue disclosure in writing.

On the closed-end side, watch the extension language, because it is how the fixed term stops being fixed. Two one-year extensions at the GP's discretion means your three year fund is legitimately a five year fund and you have no vote. Take a single private lender whose borrower asks for a six month extension. That lender usually grants it because the alternative is worse. A fund manager faces the same choice on thirty loans at once, and if the term is closing in, some of those loans get extended past it. So the honest comparison is five years against forever. Three years was never the real term.

Who signs off on the NAV in the open-end case? Is that an outside valuation firm or the manager's own team with an annual audit checking the arithmetic?

Halyard, in most of these documents it is the manager valuing loans against a written policy, with an auditor testing the policy once a year. That is a long way from an independent quarterly mark, and when a manager is asked directly whether any third party touches the quarterly number, the usual answer is a long one that means no. Neither yet gets my vote. Few of these funds have been through a quarter where a chunk of the tape stopped paying, so everything an investor would be relying on is a description of how the gate would work rather than a record of how it did. Being late is cheaper than being the person who finds out what the queue does. Lantern, the fact that both of yours quote high single digits net with such different structures says the structure is not priced yet, and things that are not priced usually get priced eventually.

Split, and treat that as a design rather than a compromise. Different money has different jobs. The open end piece is where cash goes that might be wanted back inside two years, with the understanding that you are trusting somebody's marks for that convenience. The closed end piece is money you have written off touching, and there the discipline of a term and a single identifiable vintage is the point. What does not work is splitting the same 25k in half. That means two managers to diligence properly and two sets of reports to actually read, with neither position big enough to matter.

There is a fair argument that the structure debate is a distraction from the coupon. If one fund is netting 9.2 and the other 8.4, that 80bp is measurable, and the liquidity question is a story about a quarter that may never come.

Windlass, that 80bp is 200 dollars a year on 25k. The queue is the whole principal. Those aren't the same size of question. For a first check the practical constraint is often different anyway. Most closed-end funds want a 250k minimum and a capital call schedule a small investor can't commit to, while the open-end ones will take 25k and allow monthly additions. For small checks the structure question sometimes gets decided by the minimum before the investor gets to have an opinion.

Read it from the loans themselves rather than the wrapper. If a fund does construction lending, it has to fund draws for two years on a project it committed to eighteen months ago. That is a promise of future capital. An evergreen fund with continuous subscriptions is a more natural fit for that than a fund whose term ends in the middle of a vertical build. Lantern says both of these are residential bridge, which is shorter, six to eighteen months a loan, so a three year closed-end term actually lines up with the loan lives. If either manager said they were doing ground-up out of a closed-end fund with a term inside four years, the thing to ask for is how the last one landed.

Girder's duration point is the one that decides it. Match the wrapper to the loans. Twelve month bridge loans inside a three year fund means the book turns over twice and then winds down naturally, and a plan that ends by itself beats a plan that needs a decision. The open-end version never ends by itself, so the quality of the manager has to hold for as long as the money is in it. That is a longer bet on one person than a first check usually justifies.

Nobody's mentioned the ramp. In the closed-end one your money gets called as loans are originated, so for the first several months you might be 30 percent deployed and earning on 30 percent while the fee clock runs on committed capital in some structures. Check whether the management fee is on committed or invested, because it changes the first year materially.

The open-end fund is already deployed on day one. You buy into a book that's earning. That's a real advantage that gets talked about as if it's nothing.

Distribution timing matters when the money is being spent as it arrives. For an investor covering payroll or any other fixed monthly obligation, quarterly distributions with a two month lag are a different thing to plan around than monthly ones. Ask both managers when the cash actually lands rather than when it is declared.

So in the closed-end one, do the distributions stop when the term ends, or is the last payment just bigger?

Rigging, distributions stop because the loans are gone. Through the term you're getting interest as borrowers pay, and toward the end the fund stops making new loans and starts returning principal as the existing ones pay off, so the payments get lumpier and then they're your capital back rather than income. That final stretch is where the term risk lives, because if a few loans don't pay off on schedule the wind-down drags.

And reading all of this, I think the structure argument is mostly a proxy for how much you trust the manager. Closed-end limits how long you're exposed to being wrong about them. Open-end gives you a chance to leave if you change your mind, assuming the door works when you push it.

Closed end tends to win the argument, mainly on the extension math being knowable and on the fact that open end gates have rarely been tested through a genuinely bad quarter. One thing still stands out though. In the open end fund, the people who redeem in a good quarter get out at a mark set before anything went wrong, and the people who stay own whatever is left. Does anything in a typical LPA slow that down before the gate trips, or is the gate the only protection the people who stay have? Lantern, if either manager has an answer to that, it would be worth reading.