--- name: "fix_and_flip_analysis" description: "Run fix and flip numbers before buying: itemised renovation budget from scope and quotes, financing and holding costs month by month, break-even resale price, profit and return metrics, and sensitivity to a price drop or a slow sale, with a tested flip calculator CLI. Triggers on fix and flip analysis, house flipping numbers, ARV and the 70% rule, maximum allowable offer, hard money deal costing, or 'should I buy this flip'." --- # Fix and Flip Analysis ## Purpose Decide whether a flip works before money goes in. The skill builds the full cost stack for one property: purchase, itemised rehab with contingency, financing, holding costs month by month, and selling costs. From that stack it produces the break-even resale price, the net profit, returns on cash and on total cost, and sensitivity tables for the three ways flips usually die: the sale price comes in low, the sale takes longer, or the rehab runs over. A playbook that only lists inputs leaves the arithmetic to hope. This skill ships the arithmetic. The calculator in bin/flip_calc.py computes every figure, asserts its own profit identity on cash and interest-only deals, and prints a worksheet an agent can hand over as the answer. ## Workflow 1. Screen first with the rule, then underwrite in full. Maximum allowable offer is ARV times the rule percentage minus rehab including contingency. At 70% that screen is quick and rough by design. A deal that fails it can still work with cheap money, a fast sale, or low selling costs, and a deal that passes it can still lose money on a long hold. Use the screen to sort, use the full costing to decide. 2. Pin the ARV to comparable sales before anything else. ARV comes from renovated homes nearby that sold recently, matched on size, beds, baths, and finish level. If there are no renovated comps, there is no ARV worth trusting and the honest answer is to pass or to price the uncertainty with a lower rule percentage. Never let a seller asking price or a listing estimate stand in for ARV. 3. Build the rehab budget from scope and quotes, line by line. Write one line per job: kitchen, baths, flooring, paint, roof, HVAC, electrical, plumbing, permits, exterior. Use contractor quotes where they exist and label every other line as an estimate. A scope of work for each line states what is included, what is excluded, and the schedule, because exclusions are where budgets leak. Put the lines in rehab_items in the deal file, schema in references/deal-schema.md. 4. Add contingency on top of the rehab base, never inside a padded line. Ten percent is the floor experienced flippers use. Older houses, gut jobs, and any scope where walls have not been opened need 15 to 20% or more. The calculator applies contingency to the whole base, so a thin contingency shows up as its own flag. 5. Cost the financing as a total, not as a rate. Hard money for flips commonly runs about 9 to 13% interest with 1.5 to 3 points and a 6 to 12 month term, plus document, underwriting, appraisal, draw, and extension fees. A lower rate with more points can cost more on a short hold. Enter the actual quote: rate, points, fees, loan to cost, whether rehab is funded in draws, and whether interest runs on the full commitment or only on money drawn. That last term alone moves the interest bill by thousands. For cash deals set the financing type to cash. For a conventional loan the calculator amortises the payment and pays off the remaining balance at sale. 6. Build holding costs month by month from real bills. Property taxes, vacant property insurance, utilities, HOA, lawn and upkeep, and security where it applies. Vacant insurance costs more than a standard homeowner policy, so quote it as vacant. Multiply by a hold that includes rehab time, time on market, and closing time. ATTOM put the median completed flip at 161 days in Q2 2026, so a plan under five months needs a reason, not just optimism. 7. Cost the sale before calling anything profit. Agent commissions, commonly 5 to 6% of the sale price, plus title, escrow, and transfer charges, plus fixed costs such as staging, photography, and cleaning. Sellers commonly pay 6 to 10% of the sale price in total. Enter commissions and sale closing as percentages and fixed sale costs in dollars so the break-even math stays exact. 8. Run the calculator. `python3 bin/flip_calc.py ` prints the worksheet. Add `--markdown report.md` to save it. The sample deal at assets/sample-deal.json runs as shipped. 9. Read the output in this order: break-even price and cushion first, then net profit and margin, then return on cash, then the sensitivity tables. Return on cash is flattered by leverage, so read it beside return on total cost. Annualized return is a comparison figure only, and on a short hold it inflates fast. 10. Decide off the downside, not the base case. Check the combined downside line: sale down 5%, hold two months longer, rehab up 10%. If that case loses money the buyer cannot afford to lose, the deal needs a lower purchase price or a pass. Renegotiate from the break-even: the gap between the asking price and the price that survives the downside is the negotiation, in dollars. ## Output Contract Every analysis returns: - Screen: maximum allowable offer at the chosen rule percentage, and whether the purchase price sits over or under it, in dollars. - Rehab budget: every scope line with its cost, the subtotal, the contingency amount and percentage, and the rehab total used in the math. - Funding schedule: a month by month table with rehab spent, net rehab cash after lender draws, debt or closing cash, holding costs, cash out, and cumulative cash invested. It also states the loan split between purchase and rehab draws, points, lender fees, total interest, total cash invested before sale, and the monthly burn for each extra month held. - Sale and profit: selling costs, loan payoff, total project cost, net pre-tax profit, profit margin on the sale price, return on cash invested, annualized return, equity multiple, and return on total project cost. - Break-even: the sale price at which profit is zero, and the cushion between the expected sale and break-even in dollars and percent. - Sensitivity: profit and return at sale prices down 10%, down 5%, at plan, and up 5%. Profit at holds extended by two and four months, with the extra cost against plan. Profit at rehab overruns of 10% and 20%. One combined downside case. - Flags: purchase over the screen, cushion under 10%, margin under 10% of sale price, contingency under 10%, or a base case that loses money, each with the number behind it. Worked example, verified by running the calculator on assets/sample-deal.json. Purchase $150,000, ARV $280,000, nine rehab lines totalling $48,000 plus 10% contingency for a $52,800 rehab total, hard money at 80% loan to cost, 11% interest, 2 points, $1,500 in fees, $600 a month holding, six month hold, selling costs of 6% plus $1,500 fixed. The screen gives a maximum offer of $143,200, so the purchase sits $6,800 over the screen and the full costing has to carry it. It does, on plan: loan $162,240, total interest $8,536, total cash invested $60,441, total project cost $240,981, net profit $39,019, margin 13.9% of the sale price, return on cash 64.6%, break-even sale price $238,490 with a 14.8% cushion, and a monthly burn of $2,087 once rehab ends. The combined downside still clears $16,049. That is what a complete answer looks like: the screen said caution, the costing said the numbers hold, and the downside said by how much. ## Operating Rules 1. Never present gross profit as profit. Gross in published flipping reports means sale minus purchase, with no rehab, financing, holding, or selling costs deducted. Every profit figure this skill produces is net of all of those, and pre-tax. Say pre-tax every time. 2. Never invent an input. Missing tax, insurance, quote, or loan figures get asked for, or taken from the benchmark ranges in references/formulas.md and labelled as assumptions in the worksheet. A deal file full of quiet defaults is a wrong answer with good formatting. 3. Never trim a real cost to make a deal pass. If the base case fails, the levers are purchase price, scope, or walking away. Cutting contingency, shortening the hold on paper, or lowering commissions in the model changes nothing at closing. 4. Treat the 70% rule as a screen with a known blind spot: it bundles holding, financing, and selling costs plus profit into the 30% it holds back. Long holds, expensive money, and high sale costs need a lower percentage. Short holds with cheap money can support a higher one. The full costing settles which case a deal is. 5. Keep ARV honest. Use the lower end of a comp range for the base case and let the sensitivity table show the top end. An ARV chosen to make the numbers work is the most common way flips lose money on paper first and in cash later. 6. Quote both returns. Return on cash shows what leverage does. Return on total project cost shows what the project itself earns. A deal that only looks good on the leveraged figure is a financing trade, and it breaks first when the hold stretches. 7. State the draw timing limit. The schedule treats rehab draws as reimbursed in the month the work is spent. Real lenders release draws after inspection, so the buyer fronts more cash early than the schedule shows. Flag this whenever the buyer's liquid cash is close to the cumulative cash line. 8. This skill is analysis, not advice. It does not recommend buying any property, does not predict prices, and does not do tax work. Tax treatment of flip profit depends on the buyer's situation and belongs with their tax professional.