Waynick Underwriting Start underwriting
  1. 1The model
  2. 2The app
  3. 3Audit trail
  4. 4Teams
  5. 5Pricing

Checks and audit trail

It says when your own numbers disagree, keeps where each one came from on the number itself, measures the deal after it closes, and checks every export against a second run of the model.

When the inputs disagree

It tells you when your own numbers disagree

The rent roll adds up leased and vacant area and sets it against the building's net rentable area. Recoveries are billed on net rentable, so when the two disagree every reimbursement in the model is billed against the wrong denominator. Most tools carry on. This one says so, on the screen where the mistake was made.

The rent roll with leased and vacant area equal to net rentable area, and no warning shown. The same rent roll where leased plus vacant is 52,600 square feet against 47,000 net rentable, with an amber line naming the 5,600 square foot gap.

52,600 SF listed against 52,600 SF net rentable. No banner, because there is nothing to report.

Occupancy over 100% is flagged the same way and deliberately not clamped. The ratio is right; a roll that lists more area than the building has is what is wrong, and hiding the symptom would leave the recovery shares that depend on the same number still broken.

Where a number came from

Where a number came from, written on the number

A model gets passed around. By the time it reaches committee, nobody remembers why the exit cap is 6.5% or which email the price came from. Here the answer is kept on the field itself, where the next person will look.

Right-clicking the purchase price opens a menu: edit note, calculation and history, references, confirmed, estimate, placeholder, flag for attention, unlock field, copy and paste. The field shows a lock, a green confirmed dot, a note marker and a paperclip.

Right-click any field

Every input has the same menu, on a phone as a long press.

  • A note with who wrote it and when, ticked into the memo if committee should read it.
  • Confirmed, Estimate or Placeholder, shown as a stripe down the field, and counted above the returns so an unfinished model looks unfinished.
  • A flag for the figure someone needs to look at, and a lock for the one nobody should move. The goal seek leaves a locked field alone.
  • Its history, every change with the value before it, restorable in a click.
  • A source: a photo from the deal’s references, or a link to the OM in your data room.
  • A calculation, the way Excel keeps one. Type =52600+4900 and the field shows the figure; click back in and the working is still there.

The note follows the number, and every figure opens

Write on Tenant C’s free rent that it comes from the signed LOI, and Tenant C’s first year in the statement carries a marker with that note, annual or monthly. The note is written once, on the input. And every figure opens its own explanation: the items that make it up, each with the rate behind it, and a link to the tab where those inputs live. Subtotals open onto the rows above them, all the way down.

  • Notes and flags on any cash flow line, at any depth, and on any single figure: the 2027 development cost, one lot’s sale. A folded heading says what is flagged inside it.
  • The same on each phase of a land deal: a phase’s cash flow opens to its categories, budget lines and lots.
  • The memo prints the notes, the status of every input and the sources, in their own sections.

Try it · the real engine, running in this page

After it closes

The underwriting stops being a guess and starts being a record

Close a deal and one revision is pinned as the underwriting of record. As months are entered, the tab also shows where the deal lands now: the same plan with your actual NOI in place of the planned NOI to date, its IRR, multiple and NPV beside the underwriting’s. Everything on the Performance tab is measured against that revision, not against the model as it stands today, so a number nobody can quietly improve after the fact.

Buildings you already own come in the same way. Start from what you paid and what it is worth today, with the loan as it stands now, worked out from the dates on it, and track it from this month. Nothing is charged for buying what you already have. And when the question is whether to sell, the Analysis tab answers it: selling now against selling at the end of each year, after the loan, any prepayment penalty and the cost of sale, and which year is worth the most at your discount rate, with the return since you bought it beside each.

Two charts on the Performance tab: NOI against plan month by month for six months, as bars, three months behind plan and three ahead, with a running total ending $1.4 thousand ahead; and what moved NOI, each line's effect adding up to that: base rent a little behind, credit loss, recoveries, other income and operating expenses each a little ahead.

Six months of ownership against the revision this deal closed on: each month against plan, the running total, and what moved it, with each line’s effect adding up to the NOI variance.

Enter what actually happened

Monthly actuals, expenses line by line; the totals are derived, never typed.

Rent, credit loss, recoveries, other income and operating expenses, by month, and expenses and recoveries line by line against the underwriting’s own lines, taxes, insurance and CAM each with a budget, plus one line for spend nobody budgeted. Effective gross income and NOI are never entered; they are derived, so a total cannot fail to foot.

A saving is good news even though it is negative

The color follows the engine’s verdict, not the sign of the figure.

Spending less than underwritten is a variance of minus four thousand dollars and a favorable one. The color comes from the engine's verdict about the line, never from the sign of the figure. A committee reading a cost saving as a problem is the wrong way to be embarrassed.

Both sides count the same months

Underwritten and actual always cover the same period.

If the underwritten column ever summed the whole elapsed period while the actual column summed only complete months, the report would say the deal was failing and nothing on the page would explain why. An assertion in the engine refuses to produce a report where the two disagree.

Two documents, one calculation

The committee PDF and the workbook come from one calculation.

A committee PDF and an Excel workbook, both built from the same call; neither gathers its own figures. The workbook writes real numbers under a percentage format rather than text, because a spreadsheet of strings looks identical and cannot be summed, sorted or charted.

Nothing arrives from the browser when you export these: the server reads the pinned revision, the actuals and your branding, and runs the model itself. Coverage is stated rather than assumed: how many months are in the total, which are missing inside the run, and whether the first month was a partial one.

When it leaves the app

The exports check themselves

When you export, the server does not format the numbers your browser sent it. It re-runs the entire model from your assumptions, compares its own answer to the browser's, and prints any disagreement into the document.

Browser

$6,319,061

Computed as you typed

agrees

Server

$6,319,061

Re-run from the assumptions

Net operating income over the hold, from the sample office deal. Where the two ever differ, the workbook and the memo say so on the page rather than quietly picking one.

Excel

A workbook, not a data dump

  • Cash flow transposed: periods across, line items down
  • Accounting format, costs negative so a column foots
  • Tenants and phases as collapsible outline groups
  • Native charts pointing at real cells, restyleable
  • Sensitivity grids shaded, landscape, fit to width

PDF

An investment committee memo

  • Overview, returns, financing, annual statement
  • Five charts drawn from the same series the app draws
  • Sensitivity grids and a tornado, priced per step
  • Rent roll, fee schedule, solved improvement plan
  • Your notes, input status and sources, in their own sections
  • Legends print figures, so it survives a photocopier
  • And an executive summary: the deal on one page, with what to watch

PDF

A comparison, two to six deals

  • Ranked on the numbers, not the formatted strings
  • Turns landscape from four deals up
  • Sections adapt to what is in the set
  • Each model validated separately, notes by position
  • Ties and single columns go unmarked