Modeling a Debt Waterfall When You Have Multiple Tranches

A waterfall lays out the order in which available cash or sale proceeds get applied across each layer of debt, based on the priority each lender negotiated. Senior secured debt typically sits at the top, getting paid first from available cash or collateral proceeds, with subordinated or mezzanine debt sitting behind it, and equity holders only seeing anything once every debt layer above them has been fully satisfied. The model exists to answer a very practical question: if a specific amount of cash or proceeds becomes available, who actually gets paid, in what order, and how much is left when you reach the bottom.

Companies with multiple tranches often need two separate waterfall models, not one. An ongoing cash sweep waterfall governs how excess operating cash flow gets applied to mandatory prepayments across tranches during normal operations, as spelled out in each credit agreement. A sale or liquidation waterfall is a different exercise entirely, modeling how proceeds from selling the company or its assets get distributed if that event happens, which usually involves larger numbers, different triggers, and intercreditor agreements that only matter in that specific scenario.

Before you plug in a single dollar figure, map out the priority order itself: which tranche is senior, which is subordinated, and whether any intercreditor or subordination agreement changes that order in specific circumstances. Getting the sequence wrong makes every number downstream wrong too, and it's a surprisingly easy mistake when a company has added tranches over time from different lenders who each negotiated their own priority language rather than a single coordinated capital structure.

A practical decision rule: build the cash sweep waterfall first if the company is operating normally, and the sale waterfall first if a transaction or refinancing is under discussion. For example, a company with a revolver, a term loan and a mezzanine note can list the three tranches in priority order, then add one row for each fee, accrued interest item and prepayment trigger the agreements describe. Only then should dollar amounts go in. Have someone who did not build the model trace one tranche back to its credit agreement, because a second reader most often catches the missing trigger or the priority clause that a later lender negotiated differently.

The most common modeling mistake isn't a math error; it's leaving something out entirely. Accrued but unpaid interest, prepayment penalties that only apply under specific triggers, and fees owed to a specific tranche at payoff all have to be captured layer by layer, not assumed away as immaterial. A waterfall that looks clean because it only tracks principal will consistently overstate what's actually left for equity once every real obligation is accounted for.

Go back to the actual credit agreements rather than relying on a summary term sheet when building each layer, since some of the details that matter most, like exactly when a prepayment penalty applies, live in the definitive documents rather than in the shorter summary most people keep on hand day to day.

A complete waterfall captures each of these for every layer:

• Accrued but unpaid interest on each tranche, since a model that tracks only principal overstates what remains for equity.
• Prepayment penalties that apply only under specific triggers, checked against the definitive credit agreement rather than a summary term sheet.
• Fees owed to a specific tranche at payoff, captured layer by layer instead of assumed away as immaterial.
• Any intercreditor or subordination language that changes the priority order in specific circumstances.

Run the model under a downside scenario, not just your base case, since a waterfall that looks fine when cash flow meets expectations can look very different when it doesn't. This matters most for equity holders and management with equity upside, since a downside case often reveals that subordinated tranches or equity see materially less than a simple pro rata assumption would suggest, well before an actual sale or distress scenario forces everyone to find that out at once.

Share the stress-tested version with your board alongside the base case, not instead of it. A board that has only ever seen the optimistic waterfall is far more likely to be caught off guard by a real downside outcome than one that has already walked through what a weaker scenario does to each tranche and to equity.