Profit, loss and the two waterfalls

Most capital structures state a ranking and then discover, under stress, that it was a policy rather than a mechanism. Here both orderings are established in advance, and they are not the same ordering reversed.

The loss waterfall

Loss falls on Equity first, to the extent Equity is available. Only loss that Equity cannot absorb reaches the Bond book, and there it falls on the risky component. Liquid kUSD is not impaired by the loss-allocation stages at all.

That last point is the design. A liquid holder is compensated for lending balance-sheet capacity, not for carrying asset risk, so the loss stages do not reach the position that was never paid to absorb them.

The profit waterfall

Profit is allocated in the other direction. The Bond Hurdle is met first, so a Bond recovers its own accrued preference. Equity then receives the balance of accrued preference. Only what remains after both is surplus, and surplus is divided between Bond and Equity.

The asymmetry is the point. Running loss upward from the residual and profit downward through the preferences means neither position can be senior in both directions, which is the arrangement most structures quietly adopt.

Hazard, the price of capacity

Bond holders use balance-sheet capacity to carry a leveraged position, and they are charged for it. Hazard accrues daily against the Bond book and is credited to liquid kUSD holders.

Its level is formulaic rather than negotiated. More deployed risk relative to the Reserve raises the rate; longer Bond duration lowers it. Nobody sets it in a meeting, and it is published on the balance sheet rather than advertised.

Bond holders recover the Hazard they paid through the Hurdle, at twice the amount paid, before any surplus is divided. So the full sequence is: capacity is charged for, the charge is recovered, and only then is there something to share.

Preference claims, and why they decay

An absorbed loss is not written off. Both Equity and Bonds accrue a preferential recovery claim, carried as a prior claim on subsequent profit.

Those claims decay by half every cycle. That places the structure between two conventional extremes. A hard high water mark can leave a position permanently underwater, so that no achievable performance clears it and the rational move is to walk away. No mark at all forgets losses entirely, so a position that destroyed value last period competes on equal terms with one that did not.

Halving each cycle keeps a claim economically meaningful in the periods immediately after the event that created it, and lets it fade thereafter rather than accumulating into an overhang that no future performance can clear.

The distribution gate

Absorbing first loss does not entitle the Equity seat to withdraw at will.

A minimum capital threshold is set at the start of each cycle, before that cycle’s outcome is known. Only the excess above it can be withdrawn, and then only subject to the state of the system at the point of withdrawal.

Three consequences follow, and each is a question a committee asks directly. A new deposit does not enlarge the distributable amount. Marks cannot manufacture a larger allowance. And Equity cannot defer a senior obligation in a cycle while extracting its own capital in the same window.