Gross flipping profit in that dataset is resale price minus purchase price, nothing else subtracted. Gross return is that figure divided by the purchase price. So 325,000 minus 260,000 gives 65,000, and 65,000 over 260,000 is 25 percent. Renovation, financing, carry, agent commissions and closing costs all sit outside it. The number is a spread, not earnings.
Run a cosmetic version through. Buy-side closing and points, 5,000. Cosmetic scope at that price point, 25,000 to 35,000. Short-term financing on 200,000 borrowed at double-digit rates for five months plus origination, 12,000 to 16,000. Taxes, insurance and utilities while vacant, 3,000. Selling at 7 to 8 percent of 325,000, 23,000 to 26,000. That totals 68,000 to 85,000 against a 65,000 spread. The median deal in that shape doesn't clear, which is consistent with the share of flips falling to low single digits as sales.
Two things keep it viable. Cash or partner equity strips out most of the financing line, and a genuine off-market discount moves the purchase well below the median rather than paying the median. That's why the strategy has shifted toward off-market sourcing and short timelines. Where the reported returns run far above the national average, in the affordable Midwest and Northeast metros, the spread as a percentage of a low purchase price is doing the work.
One caution on reading the medians at all. That data counts arms-length resales inside a twelve month window, so it includes wholesale flips with no renovation and light retail resales alongside real rehabs, and the median buy and the median sale are not the same houses. Treat it as a direction of travel and underwrite your own deal from the cost stack up.