Rental Yield Reality Check: What Medellín Actually Returns

"Medellín properties return 7–8%+!" You've seen the claim — it's on half the broker sites in this market. It isn't exactly a lie. It's a gross number, computed at optimistic occupancy, on revenue before a single peso of the expenses that actually run a rental. The honest version of this conversation starts with subtraction.

Gross vs net: the entire argument in one distinction

Gross yield = annual rental revenue ÷ purchase price. It's the number in the marketing because it's the biggest one available.

Net yield = (revenue − operating expenses) ÷ all-in cost (price + closing costs + furnishing). It's the number that pays your bills — and in Medellín the gap between the two is routinely 30–50% of gross revenue, wider for short-term than long-term.

The expense lines the yield claims skip

• Administración — the HOA fee, due whether the unit earns or sits. In the amenity buildings foreigners buy, commonly COP 250,000–550,000+ monthly — on a smaller unit this line alone can consume one to two points of yield.
• Predial — annual property tax, scaled to cadastral value.
• Vacancy — the silent killer of gross math. Market-wide short-stay occupancy in Medellín has been running roughly 47–55% — meaning the average listing earns nothing almost half the nights, in a market of 25,000+ competing listings. Strong operators do materially better; spreadsheets assuming 75%+ occupancy are describing the top decile, not you in year one.
• Management — ~20–30% of revenue for short-term (guest ops, cleaning coordination), ~10–15% for mid-term, roughly a month's rent per placement or ~8–10% for long-term. Self-managing from abroad is a job, not a savings.
• Utilities + wifi — owner-paid in furnished strategies; tenant-paid long-term.
• Furnishing amortization & wear — the furniture package is capital that depreciates fastest under nightly churn.
• Platform fees, cleaning, RNT compliance — short-term only, and not small.
• The 4×1000 and money-movement costs — 0.4% skimmed as revenue moves through Colombian accounts, plus FX costs repatriating income.

Strategy by strategy, honestly

Notice what's not in this table: a promised percentage. Any specific "Medellín yields X%" claim is answering a question with too many variables — neighborhood, building fees, purchase price paid, strategy, operator skill, occupancy achieved — to have one answer. Which is exactly why the useful skill is underwriting, not quoting.

How to underwrite a real unit in six lines

1. Revenue, evidenced: actual rents for comparable units in the same building or block — pulled from live listings and, better, from what a broker knows actually gets paid. For STR, comparable listings' calendars, not their nightly prices.
2. Occupancy, haircut: whatever the optimistic number is, model the market-average case too. If the deal only works in the optimistic column, it doesn't work.
3. Expenses, itemized: the building's real administración fee (ask for the actual number — it's not a secret), predial estimate, management at honest rates, utilities, a furnishing reserve.
4. All-in denominator: price plus the ~3–5% transaction stack (closing-costs breakdown) plus the furniture. Yields computed on the sticker price flatter themselves.
5. Purchase price pressure: the cheapest yield improvement in this market is paying less — every point negotiated off an inflated ask (the gringo-price problem) flows straight to your return forever.
6. Compare against the boring alternative: if leverage is involved, Colombian mortgage rates near 15% EA devour rental spreads — see the mortgage reality — and even cash buyers should sanity-check the net against what the same dollars earn elsewhere. Medellín property can absolutely win that comparison; it wins it on net numbers plus the things yield doesn't capture: a home in a city you love, visa eligibility, peso diversification, and an asset you can walk through.

The bottom line

Medellín rental investment works — for owners who bought well, picked the strategy their building and temperament actually support, and modeled the expense side before wiring money. It disappoints owners who bought a gross-yield screenshot at a gringo price. The difference was never the city; it was the arithmetic. Do the subtraction first, and the number you're left with is one nobody had to promise you.