> ## Content Index > Fetch the complete content index at: https://thesilentpartner.ghost.io/llms.txt > Use this file to discover other available public pages before exploring further. # Your Underwriting Is Always Wrong - URL: https://thesilentpartner.ghost.io/your-underwriting-is-always-wrong/ - Published: 2026-09-28T00:00:53.000Z - Updated: 2026-09-28T00:00:52.000Z - Author: TheSilentPartner Every deal I have ever bought underperformed the model I bought it on. Every single one. Not because the models were careless. Because that is what models do. And once you accept it as a permanent condition rather than a series of unlucky exceptions, you start buying differently. ## The number is a ceiling, not a forecast When you underwrite a property, you build a picture of what it will produce. Rents, expenses, vacancy, the capital you will put in, what is left over. People treat that output as a prediction. It is not. It is the best case that survived your assumptions. Here is what actually happens after you close. Insurance renews higher than you modeled, sometimes dramatically. Something structural fails that the inspection did not catch, or caught in language vague enough that you discounted it. A tenant leaves and the unit takes longer to fill than your vacancy assumption. Taxes reassess. A law changes. A pipe does what pipes do. None of these are catastrophes. Individually, each is a rounding error. Together, they are the difference between the number in the spreadsheet and the number in the bank account, and that gap is consistently in one direction. The gap is not a sign you did the analysis badly. It is structural. Your model contains everything you thought of, and reality contains everything. ## Where the 40% comes from So here is what I do, and it is the most useful mechanical rule I have. I take forty percent off actual gross income. Then I look at whether there is still a deal. Two words in that sentence carry the weight. **Actual.** Not projected, not stabilized, not market. What the property is collecting right now, verifiable, from real tenants who are really paying. The gap between actual and pro forma is where most bad acquisitions live. **Forty.** It is not a precise figure derived from a formula. It is what decades of watching the gap between model and outcome has taught me is roughly the size of that gap plus enough margin that I sleep. If a deal only works at a ten percent haircut, it does not work. It is relying on nothing going wrong, and something always goes wrong. I run this on everything before I look closely at anything. It is a first screen. If the deal survives forty percent off actual gross, it is worth my attention. If it does not, I am done, and I have spent four minutes instead of four weeks. ## I never look at pro forma rent Pro forma rent is what the seller believes the property could produce under a set of conditions that have not occurred. It is not a lie, exactly. It is a sales document wearing the clothing of an analysis. Somebody built it to make a property look like something, and the assumptions inside it were selected accordingly. You are permitted to look at it the way you would look at a resume, as a claim about capability, not a statement of fact. Then go find out what is actually true. If a broker package leans heavily on pro forma numbers, that itself is information about the deal. ## The buffer is the position There is a way of buying where everything works if things go as planned. Financing is stretched to make the numbers clear. Reserves are thin because the money went into the purchase. The deal pencils, barely. That is not a deal. That is a bet that the next three years will be uneventful. The years are never uneventful. Rates move. Laws change. The insurance market has a bad decade. A tenant stops paying and the process to do anything about it takes far longer than you assumed. If you are tight, any one of those forces a decision at the worst possible moment. If you have room, the same event is an annoyance you handle in a Tuesday afternoon phone call. The cushion is not being conservative. It is what lets you stay in the position long enough for the thing to work, and time in the position is where the return actually comes from. ## What this means if you are not buying buildings The principle generalizes, and it is worth stating plainly. Every projection you are handed was built by someone with a reason to build it. The startup growth curve. The fund target return. The renovation estimate. The salary progression in the offer letter. None of them are lies. All of them are best cases that survived a set of assumptions chosen by an interested party. Take a haircut. Then decide. You do not need my number. You need a number, applied consistently, that reflects the honest distance between what people project and what shows up. Once you have one and you use it every time, an entire category of expensive mistake stops being available to you. ## The rule The model tells you the ceiling. Buy as though you will land well below it, because you will. If the deal only works at the ceiling, it is not a deal. It is a hope with a spreadsheet attached.