Idle cash between note maturities can cost about 1.7 points a year

Cash drag is what 9 month paper costs you. You bought the right to get your money back four times faster than you can find a use for it and you're surprised the option has a price. When I hold land for eight years there's no reinvestment problem because there's nothing to reinvest. If the drag bothers you, buy longer duration and stop touching it.

Before anything else, does your 7.8% include the extension interest as received or as accrued? If one note extended twice and you haven't been paid the extension period yet, your realized number is understating you and your drag is smaller than you think. I got that wrong on a private note last year and spent a month solving a problem that wasn't there.

Worth pinning down where that 31 days average dead time is measured from. Is it from payoff posting to the next funding date, or from payoff posting to the day capital was committed? Those can be two weeks apart on some platforms because committed capital sits uninvested until the deal closes. If it's the second one, part of the drag is structural and no ladder fixes it.

Turning on auto-invest is a common fix for exactly this problem, and it typically cuts idle time to something like six days. It also tends to place capital into deals an investor would not otherwise pick, and one of those can end up well past due on interest while still sending cheerful status updates. So it solves the idle cash metric and creates a different problem in its place. Auto-invest with extra steps is a fair way to describe option one, and it is worth being clear eyed about that trade.

Sorry to slow the thread, what is cash drag exactly? Is it just money sitting in the account not earning anything, or is there a fee involved too?

@karen_whitfield That framing understates it. Short duration is doing real work here, tender gets to reprice every nine months and re-underwrite the sponsor each time. That's worth something in a market where nobody trusts a two year old appraisal. @kevin_lombard It's idle money. No fee involved, the balance simply earns nothing while it waits, so the return on the whole pot ends up lower than the rate on any single note. The number worth chasing isn't the average dead time, it's the distribution. If most payoffs redeploy in a week and one sat for four months, that's one bad month, not a systemic problem, and the fix is a rule for that one case. Pulling the actual gaps note by note usually beats giving up 140 basis points on 25k without checking.

I track this as weighted average deployed percentage per month and it's the only number on my dashboard I actually look at. Mine runs about 91%. On 85k that's 7.6k of idle capital on average, which at 9.4% is roughly 715 a year of foregone interest. Your gap is bigger than that, which makes me think something else is in there too. Check whether the platforms are paying you from the funding date or from the closing date.

It depends on how the note is structured. An extension fee typically flows to whoever is bearing the extended risk, usually the lender or lenders of record, not the platform. If it isn't showing up in a distribution, ask the platform whether the fee is retained as a servicing charge or passed through, since both structures exist and the loan documents should say which applies.

@chriswalton8 it depends entirely on the offering documents and it varies a lot. In most sets of documents, the extension fee goes to the platform or the affiliated originator as servicing income, and the investors get the same coupon for a longer period. Sometimes the coupon steps up 100 to 200 basis points on extension. Sometimes it doesn't move at all. The clause is usually two paragraphs from the default interest language, and default interest is often split the same way. Read those two clauses side by side before funding anything, since they tell you whose interests the extension serves. Tender, that also matters for the ladder question. If extensions are free to the borrower and profitable to the platform, a 9 month term is really a 9 to 18 month term and the ladder spacing is fiction.

@marcus_hale41 is pointing at the right thing. On many platforms, interest to the investor starts on the closing date, and funds are called and held for anywhere from 3 to 20 days before that. Nobody advertises this. It's in the subscription mechanics, usually one sentence about funds being held in a non-interest bearing account pending closing. Twenty days a year on four rotations is nearly a month of dead capital that never appears in a dead-time calculation because the account shows the money as invested. Worth adding up those pending days across a full year of rotations before choosing between the two options. A third of the gap likely lives there and neither option touches it.

The pooled fund idea is the part worth understanding fully. If redemption is quarterly, what price do investors get out at? Is it whatever they put in, or does the fund mark the loans down if some go bad? The phrase net asset value gets used constantly in these structures, and it's still worth confirming plainly whether it moves.

@raj_mehta14 it moves, or it should. Pooled debt funds carry the loans at some value, and if borrowers stop paying, that value comes down, and a redemption is at the marked price on the request date. Whether the manager marks aggressively is a real question, and the answer is in the valuation policy in the offering docs. On the math: 6.6k against an average balance of 85k is 7.76%, fine. But average balance is doing hidden work. If the balance was 85k in month one and 61k in month seven because payoffs had not redeployed, then average deployed capital was well under 85k and return on deployed capital was closer to the stated rate. A 1.6 point gap like that is a return on committed capital, not a return on the notes. Both numbers are true and they answer different questions. If the alternative use of idle cash is a savings account, committed capital is the right denominator and it matters. If the alternative is nothing, the gap is really an accounting choice, not a loss.

Reading all this I'd take the 8% and stop. You're spending real hours to defend 1.4 points on 25k, which is 350 a year. I know that's the unimpressive answer.

@stephen_ok9 350 a year for the rest of an investing life compounds into something less unimpressive, though the point about the hours involved still stands.