How the preferred equity waterfall on a ground-up industrial shell actually works at a refi versus at a sale

No market standard, that ambiguity is the negotiating room.

Push for language that says the multiple tracks on deployed capital, not returned capital, so a partial refi paydown doesn't reset the clock.

Dayton spec industrial at a 7.0 exit cap in this rate environment is already thin; you cannot afford to give that 80bps back on a waterfall technicality.

The two structures you are describing are sometimes called "return of capital" and "return on capital" resets, and sponsors exploit the vagueness on purpose. The well-drafted version says a partial refi return reduces the outstanding preferred basis for accrual purposes while the multiple is calculated on the original committed capital rather than the remaining basis, so a mid-hold refi never lets the sponsor off the hook for the full 1.5x on $1.8M. Where the language goes soft, the sponsor argues at refi that the multiple has been partially satisfied pro rata, and an investor in that position can end up roughly $40k below what the model said.

The multiple should survive a partial return. That is what a multiple on the full preferred means, and any sponsor who argues it resets at refi is redefining the deal mid-stream. That argument tends to surface when the term sheet says 1.5x on invested capital without specifying that returned capital still counts toward the denominator, so the exact noun matters more than the number. At a 6.1 yield on cost against a 7.0 exit cap you have basically no cushion for the sponsor to also be eating into your multiple through waterfall language. What to nail down before committing is whether the refi proceeds at month 30 are projected to return all $1.8M or just a partial, because if it is partial, the remaining basis question you are asking stops being a hypothetical and becomes real money.

The 1.5x surviving a partial refi return is the structure I have seen hold up in the deals I have reviewed most carefully, but the mechanism that makes it work is narrower than people think. What you want the waterfall language to say is that the multiple applies to contributed capital, not outstanding capital. Those are different numbers the moment any refi proceeds come back. If the language ties the multiple to outstanding basis, the sponsor has a legitimate argument that returning $900K of your $1.8M means your remaining 1.5x target is now calculated on $900K, not $1.8M, and you just gave away roughly $900K in future promote protection for free. I have seen that exact reframe cost a preferred investor around 140 bps on a similar exit cap scenario, which is in your range.

The thing I am not seeing in your post is what the current-pay structure does at refi. If 9% is current pay and the refi is at month 30, you have already received roughly $486K in cash interest by then assuming full current pay from day one. Does that $486K count toward the 1.5x multiple or is the multiple calculated only on return of capital, with the current pay sitting outside the waterfall entirely. That is the question your lawyer actually needs answered before the draft even matters, because if current pay counts toward the 1.5x, your real exposure on the multiple gap at a tight refi is smaller than your model probably shows, and the sponsor may have already priced that in when they set the 1.5x number.

Sponsors almost never volunteer clarity on ambiguous waterfall language before they have to. Whenever an investor waits on the full draft to see how it reads, the draft comes back with the language that favors the sponsor, and the negotiation then starts from their position instead of yours. Push for a term sheet rider that defines return of capital explicitly before your lawyer even looks at the draft, because the term sheet is where the real negotiation happens and the LLC agreement mostly records the outcome. On the specific mechanics, consider a comparable preferred position in a 62,000 sf cold storage shell where this exact fight plays out at the term sheet stage. The workable language says any capital returned at refi reduces the preferred basis dollar for dollar, while the multiple keeps accruing on the original committed amount until full realization. Under that language a $900K refi return leaves the clock on the remaining $900K untouched and simply reduces the balance the current pay is running on. The 80 to 120 bps swing you are modeling is real. On a similar yield-on-cost spread, letting the sponsor define the term their way typically costs the preferred investor about 90 bps.

The version worth guarding against runs like this. A sponsor returns $1.1M of a $1.4M preferred at refi on a small mixed-use in Columbus, then argues the 1.5x has been satisfied on the returned portion and resets the clock on the remainder. That costs the preferred holder about $60k in blended return, purely because nobody defined whether the multiple applied to the tranche as a whole or to each dollar as it came back. Push for language saying the multiple is calculated on the original committed amount, full stop, regardless of when or how capital is returned, and that any partial return reduces the outstanding basis without reducing the total multiple obligation.

At a 6.1 yield on cost against a 7.0 exit cap you have almost no cushion for a reset argument to go the wrong way. What needs answering before anyone commits: does the waterfall define "return of capital" as return of your contributed basis specifically, or does it sweep in any distribution to the preferred tranche regardless of label.