Home › Solutions › Waterfall model builder

§ solutions

Waterfalls without the spreadsheet spaghetti.

For sponsors, GPs, and fund controllers who calculate partner distributions in a workbook only one person fully understands.

Most shops run their distribution waterfall in a workbook that grew one exception at a time — a pref accrual here, a side letter there — until the tier logic lives across forty tabs and one analyst's memory. Playgrounds builds it as an app instead. Describe the structure once — preferred return, catch-up, promote tiers, clawback — and it becomes encoded logic that computes every partner's split from live deal performance, shows its math tier by tier, and doesn't depend on the person who built the spreadsheet still being in the building.

How It Works

Describe the structure

Say it the way the operating agreement does: an 8% cumulative pref, return of capital, a 50/50 catch-up until the GP holds 20% of profits, then 80/20 above a 10% hurdle. Playgrounds turns the words into working tier logic, with the conventions — compounding, timing, deal-by-deal or whole-fund — made explicit instead of implied.

Connect the cash flows

Load capital contributions and distributions from your accounting software or the spreadsheets that hold them today, and pull projected proceeds straight from your underwriting model. The waterfall computes from actual numbers, so the splits move when the deal does.

Distribute and document

Run a distribution and see each partner's share with the tier-by-tier math behind it. Test hypothetical exits, export per-partner statements, and keep an audit trail of every calculation for the fund administrator and the auditors.

lp net irr
14.6%
gp promote
$2.3M
active tier
80/20
return of capital
preferred return
gp catch-up
80/20 promote
Fig · a waterfall model, built from a sentence

What You Can Build

  • A full waterfall engine — pref, catch-up, multi-tier promote, and clawback, structured deal-by-deal or across the whole fund.
  • A distribution calculator — enter sale or refinance proceeds and see every partner's split, with each tier's math shown instead of buried in cell references.
  • An investor statement generator — per-partner capital accounts, contributions, distributions, and pref accrual, assembled on demand instead of at quarter-end.
  • A promote forecast — what the GP earns across exit prices and dates, updated as the assumptions in your sell / hold analysis move.
  • A fund-level roll-up — every deal's waterfall in one view, with net-to-LP returns and clawback exposure visible before they become a surprise.

Why Generic AI Falls Short Here

Generic AI can define a promote. It can't hold your operating agreement. Ask ChatGPT or Claude to run your waterfall and you're starting from zero — walking it through your agreement's structure, the pref convention, how your catch-up works — and the answer comes back as a chat message, gone when the conversation ends. Playgrounds encodes your actual structure as reviewable logic, computes from your actual cash flows, and shows the tier math so your controller can check it — a shared tool your team runs each quarter, not a one-off answer.

Frequently Asked Questions

What waterfall structures can it handle?

The structures CRE agreements actually use — cumulative or non-cumulative pref, compounding conventions, GP catch-ups, multiple promote tiers over IRR or equity-multiple hurdles, and clawback provisions. The model is built from your operating agreement's terms, not a fixed template.

How is this different from our Excel waterfall?

The tier logic lives in one reviewable place instead of scattered cell formulas, it computes from live cash flows instead of pasted numbers, and it shows its math tier by tier. Your assumptions carry over — the difference is that anyone on the team can run it, check it, and trust it.

Can partners see their own numbers?

Yes. Role-based access lets you give each partner a view of their own capital account and distributions without exposing the whole fund. Most sponsors keep the full waterfall internal and share per-partner statements.

Is our deal and investor data used to train AI?

No. AI features run under enterprise terms that prohibit training on your inputs. Every record is tenant-scoped, encrypted in transit and at rest, and stays yours.

Encode the waterfall once. Run it at each distribution.

Free to start, no credit card. Describe the structure and see the tier math in minutes.

Build it free → See pricing