Samsung's Stock Perp Just Printed -1,821% Annualized Funding

I pull our own live funding feed every morning before I look at anything else — 866 Binance USDT perpetuals, refreshed every 15 minutes. Most mornings it's boring: a pile of coins near 0.01%, a few hot ones near 0.05-0.1%. This morning one line read -1.6635%. That's not a typo and it's not a display bug. It's SAMSUNGUSDT, Binance's tokenized Samsung stock perpetual, and it's the most extreme number on the entire sheet by a wide margin.

Run it through the same simple-APR formula our own funding rate APR calculator uses — rate per window × 3 windows a day × 365 days — and -1.6635% becomes -1,821.5% annualized. Nobody actually holds this for a year. But the sign and the size both matter, so let's take them one at a time.

Negative means shorts are paying, not receiving

On Binance, a positive funding rate means longs pay shorts; negative means shorts pay longs. -1.6635% isn't a small credit for shorts — it's a large bill for them, paid straight to whoever's long, three times a day, every day the rate holds. Scale that to a real holding period instead of a full year: over just 10 days (30 funding prints at this rate) a short position hands over roughly 49.9% of its notional in funding alone — before the stock price moves a cent. I checked this against our own funding rate pain calculator using the live rate and a 10-day window and it lines up with the sheet's own 10-day column.

That's the part I keep coming back to. A trader who shorted Samsung's perp on a bearish thesis and is right about direction can still lose money, because the funding bill outruns a normal stock move. I've written before about funding alone eating a leveraged position's entire liquidation buffer before price does anything — this is the same mechanic, just on the extreme end of the distribution instead of the median.

It's not just Samsung — the tail is stock perps

I sorted all 866 pairs by APR to see if this was a one-off glitch. It isn't. Of 866 pairs, 553 read positive, 264 read exactly 0.0000%, and only 49 are negative — but nearly every extreme on the negative side is a tokenized stock, not a crypto coin: NMR -274.8%, SAGA -173.2%, ONE -118.8%, COTI -99.9%, AMC -77.1%, EBAY -67.8%. On the positive tail it's the same story in reverse — CYPH +141.5%, TTWO +102.8%, WMT +85.9%, ASML +65.5% — stock perps stacked at both extremes while BTC and ETH sit within a point or two of zero.

The reason is liquidity, not the underlying stock's volatility. Major crypto perps have deep, roughly balanced open interest, so the funding mechanism only has to nudge the rate a little to keep the perp pinned near spot. A stock perp with a fraction of that open interest, sitting almost entirely on one side (in Samsung's case, apparently almost entirely short), needs a much bigger rate to force any rebalancing at all. I covered the same thin-liquidity effect from the leverage side in why exchanges cap stock perp leverage so much lower than crypto majors — this is the funding-side version of the identical problem.

What I do with this

Before I'd touch a "story" short on a stock perp — betting a company's overpriced, a product flopped, whatever the thesis — I now check the live funding rate first, not last. A -1,821% APR number means the market has already made that trade extremely expensive to hold, independent of whether the thesis is right. I run the actual rate through the funding rate APR calculator for the annualized cost and the funding rate pain calculator for what a realistic 5-10 day hold actually costs in dollars, before I size anything. On a normal BTC or ETH perp that check takes five seconds and rarely changes my plan. On a stock perp, this morning proved it can flip the entire trade.