A quality of earnings review is good at what it tests. Revenue recognition, addbacks, normalization, whether the earnings are real and repeatable. Then it closes, the deal closes, and a number nobody decomposed starts drifting: how much of each booked dollar actually becomes a banked dollar, month by month, across the hold.
Seven points, no line item
At one practice we measured, booked-to-banked conversion sat at 94 cents on the dollar around close. Two years in, it was 87. Seven points of margin, gone, with no line item anywhere in the board pack explaining it, because the board pack reported revenue and reported cash and reported nothing about the space between them.
Where a booked dollar waits
Every booked dollar is in one of five states. Aging with a payer. Denied, and either being fought or already dead. Clawed back after payment. Received and sitting unrecorded in the billing system. Or banked. Only the last one is working capital. Standard reporting lumps the first four into one receivables number, which is why conversion can bleed for two years without producing a variance anyone has to own.
The miss is structural rather than careless. Revenue comes out of the billing system, produced by the practice on the accrual calendar. Cash comes off the bank feed, assembled by finance on the deposit calendar. Two sources, two owners, two schedules, and the reconciliation between them is nobody’s deliverable, so it never becomes a page. The board pack faithfully reports both endpoints of a pipe and nothing about the pipe, which is how a portfolio can review a practice monthly for two years and never once discuss its conversion.
Structural or discount, nothing in between
The multiple math here is unforgiving. A conversion gap you find and close structurally is margin expansion that prices into the exit, because the improvement survives diligence: the buyer’s QoE tests it and it holds. A conversion gap you never find gets found anyway, by the next buyer’s team, at the worst possible moment, and comes off the price with a story attached about operational control. The gap gets priced exactly once. The only variable is which side of the table gets to price it. Conversion is an operating number, it moves during a hold, and that is why it belongs in how a sponsor reads a practice’s operating numbers rather than in the closing binder.
The instrument
The hold-period arithmetic makes the stakes plain. Conversion leakage is margin leaving every month of the hold, and at exit the multiple applies only to whatever margin survived. Recovered early, the same points compound twice: as cash collected through the hold, and as structural margin priced into the exit. Recovered never, they compound the other way, and the last year of a hold is the most expensive possible time to learn a practice has been quietly donating seven points back to its payers.
The fix is a monthly bridge, one page per practice. Billed, expected after contractual discounts, banked against actual deposits, and the difference decomposed to the dollar: aging by payer, denials worked versus abandoned, clawbacks, received-but-unrecorded, prior-period cash. Anchored to deposits, not to what the billing system claims it recorded, because deposits are the one number in the pack nobody can argue with. Any line of the gap that grows two consecutive months flags with an owner. Run monthly across a portfolio, the same page becomes comparable practice to practice, and conversion stops being a year-two surprise.
Run the bridge on one portfolio practice. Seven days against its own data. You will see the model, the conversion number, and every dollar of the space between the P&L and the bank. Ask us to run it.