Most real estate calculators ask you to trust a black box. Mine shows its work, and you can reproduce every number with a hand calculator.
I am a licensed Florida commercial real estate broker, and I built the calculation engine behind every tool and worked example on this site. I built it because the wave of AI-written real estate content gets the numbers wrong in ways that cost real money. An IRR figured as a shortcut instead of on the actual cash flows. A waterfall that splits the wrong tier. A depreciation schedule run on the wrong basis. One bad assumption in a model is the gap between a deal you should buy and one you should walk away from.
One engine, no second math path.
Every figure, in every calculator, in every lesson, comes from a single calculation engine written from first principles. The formulas are the canonical ones a lender or an appraiser uses. The engine carries no hidden defaults. Where this site uses AI to draft a scenario, the AI never touches the arithmetic: it proposes the assumptions, and the engine computes the result.
The conventions that separate a right number from a plausible one
- IRR runs on dated cash flows. A 1.6x return over five years produces a 9.9% IRR only when the cash lands in a lump at year five. Collect some of it along the way and the answer climbs. The engine solves the true internal rate of return on the real stream.
- Exit value uses forward NOI. You capitalize next year’s income at your exit cap rate, the way a buyer underwrites your sale.
- Waterfalls solve tier by tier. Preferred return, GP catch-up, and promote each compute on their own tier, never blended into one split.
- 1031 depreciation respects carryover basis. The exchanged basis keeps depreciating on its original schedule, and only the new money starts fresh. Depreciating the full new price overstates the deduction.
Seventy-five independent checks
The engine ships with 75 tests. Each one has an answer known from first principles or hand computation, so the engine has to match an outside source, not its own output. Every test passes. A sample you can reproduce with a hand calculator:
| Inputs | What it computes | Result |
|---|---|---|
| NOI $89,805 / price $1,200,000 | Cap rate | 7.48% |
| NOI $100,000 / debt service $80,000 | DSCR | 1.25x |
| $750,000 loan, 6%, 30-year amortization | Annual debt service | $53,960 |
| $8,320,000 loan, 6.5%, 25-year amortization | Annual debt service | $674,127 |
| Out $100 today, in $20/yr for 4 years, $80 at year 5 | True IRR (dated cash flows) | about 14% |
| The same stream, by the multiple shortcut | The WRONG number a shortcut hands you | about 9.9% |
| $2.0M out, $4.0M back over the hold | Equity multiple | 2.0x |
The IRR rows show the gap between a shortcut answer and a correct one. The shortcut, about 9.9%, is what a lot of models and a lot of AI will hand you. The right answer, about 14%, counts the cash you collect each year. I re-solved every worked example in the catalog against this engine and corrected the ones that used the shortcut.
What this means for your deal
When a number on this site says your deal clears a 1.25 debt coverage, or returns a 14% IRR, or leaves you $470,000 after a refinance, it came from the same tested engine, computed the same way every time. You can reproduce it. That is the whole point.
Educational tools and content. Not financial, investment, tax, or legal advice. Confirm every assumption against your own deal and your own advisors before you act.