Everyone is watching the price. No one is watching the plumbing.
This month The Economist walked into a crowded bull market and pointed at the broken pipe underneath the floorboards: perpetual futures, it warned, quietly drain about 10% per year from long positions. Not through bad luck. Not through liquidation events. Through the design of the product itself.
In a bull market that has taught a generation of retail traders to 'long until infinity,' that claim should have triggered a cascade of rethinking. Instead, the reaction was a shrug. Ten percent felt too small against 40% quarterly pumps.
It isn't.
I spent four months in 2017 modeling how liquidity recycled through ICO token sales, and a recurring pattern keeps haunting me: the most dangerous cost in finance is the one built into the instrument rather than printed on an invoice. Perpetual futures are the cleanest example of that truth.
The Hidden Mechanism
BitMEX invented the perpetual contract in 2016. No expiry date. No settlement. To keep the market price from running away from spot, it introduced funding rate — a periodic payment between longs and shorts. Exchanges recalculate it every eight hours, or in some cases every hour. The formula sounds harmless: a fixed base rate usually 0.01%, plus a premium/discount coefficient tied to how far the perp price sits from its index.
That 0.01% per period is the quiet vampire. Three payments per day, 365 days per year: 0.01% x 3 x 365 = 10.95%. If the market were completely balanced in the long run, holding a perpetual long would cost roughly 10% per year just to maintain the position. The Economist did not find a bug. It found the engine.
Tracing the liquidity ghosts through the ICO fog, I remember how quickly we learned that recycled volume looks identical to organic demand. The same illusion is at work here. A perp chart with heavy open interest looks like conviction. But every minute of that conviction carries a rental fee, and the renter is almost always the long side.
To see why this is a structural feature, compare it to traditional futures. A quarterly futures contract has a built-in roll: when the futures price converges to spot at expiration, the long either gains or loses depending on the term structure. Perpetuals simply turn that basis into a continuous funding payment. The difference is that in traditional markets, the basis is normally shared between the two sides of the equation; the curve can be in backwardation, and longs get paid. In crypto, the overwhelming historical bias has been contango, especially in bull markets. That means the curve is your landlord.
The 10% Figure Is the Floor, Not the Ceiling
This is the critical part that mainstream coverage tends to muffle. The Economist's estimate is a long-run equilibrium assumption. It assumes the funding rate settles at 0.01% per period, with longs and shorts taking turns paying depending on the premium. In reality, bull markets are asymmetrical. Sentiment pushes the perp price above spot for weeks at a time. Funding rates rise to 0.05%, 0.1%, even more. That turns the annualized cost into 60%, 120%, or higher — a tax that scales with euphoria.
The real number is worse when you add the full cost stack. Taker fees on each side, bid-ask slippage on the way in and out, and the occasional forced liquidation that is almost guaranteed when you combine leverage with a coin that moves 10% in a day. Put it all together and the all-in carry for a leveraged long is closer to 15% to 50% per year. The Economist's 10% is the lower bound of a structural wealth transfer, not an upper ceiling of acceptable pain.
The Math That Kills Accounts
Let me show you why this is not a small detail. Suppose you deposit $10,000 on a derivatives exchange and open a 10x long on Bitcoin at $100,000. Your notionally controlled position is $100,000. Funding is paid on the notionally controlled amount, not your margin. If annualized funding is 10%, the contract will charge $10,000 per year in funding payments. That is exactly equal to your entire deposit. One year of flat prices with a 10x long and a 10% funding cost wipes you out completely — even if Bitcoin doesn't move.
This is the non-obvious point that makes the warning so serious. Leverage multiplies the leak by the same factor that multiplies your P&L. The typical retail trader thinks in terms of 'funding is 0.01%, no big deal.' They rarely translate it back to their margin balance. I remember watching this pattern during the 2020 DeFi summer, when I was modeling arbitrage mechanics against traditional FX forward markets. The same arithmetic kept appearing: every mechanism that imposes a carrying cost on one side eventually becomes a transfer to the other side. In cross-border payments, the cost is hidden in FX spreads. Here, it's hidden in the funding rate.
The liquidation dimension deserves a separate paragraph. Most retail traders don't plan for it, but they should. A 10x long can be liquidated by a 10% adverse move, plus fees. If you are holding across days or weeks, the probability of hitting that threshold at least once is far higher than most traders assume. Each liquidation pays not only the exchange and the insurance fund, but also the opposite side of the book, which typically re-enters lower. The combined effect is a repeated tax on directional conviction. This is not captured by The Economist's 10% calculation.
Bull Markets Tax the Bull
Here is the punchline for the current cycle. In a bull market, funding rates are usually positive. It is not unusual to see funding on BTC perps sitting at 0.03% per period for weeks on end. That is 32% annualized. On altcoins, it gets absurd. The market has designed a system where the direction that is most optimistic is charged the most rent.
This is the opposite of the 'free leverage' story that crypto-native marketing departments love to sell. The Economist's 10% per year is not an eternal truth. It is a bull-market estimate. In fact, the more convincing the bull narrative becomes, the higher the funding tax goes. Even a modest bull market with funding at 0.02% per period gives an annualized charge of 24%. A euphoric market at 0.06% per period burns 72% per year. The price of conviction is built into the instrument.
Who Is Collecting the Rent?
It is worth asking who sits at the other end of the trade. The answer is the market's institutional plumbing: market makers, basis-trading desks, quant funds. They deploy a simple strategy: short the perp, buy the spot asset, collect funding while delta-neutral. In a structural world where retail longs pay 10% or more annually, these funds are the silent beneficiaries. That's why the warning should be read as a class conflict, not a market forecast.
The exchange itself is not innocent either. It earns trading fees, not funding fees. So it has no direct incentive to reduce funding. If open interest stays high, the fee engine runs. Even if funding eats through retail accounts, the exchange is mildly protected by its own fee model. This is a principal-agent problem dressed up in Greek letters.
On-chain perpetual protocols such as dYdX, GMX and Hyperliquid have tried to differentiate by making funding transparent or even zero on certain pools. But transparency does not eliminate the carrying cost; it only makes it visible. A long is still a long, and paying 20 cents per day for every $10,000 of notional exposure is still a leak. The more useful question is not which platform has the lowest fee, but which market structure ends up paying the trader for risk. That is rare.
And then there is the regulatory elephant. The Economist is not just a media outlet; it is an agenda-setting mechanism for central banks and finance ministries. Once '10% annual drain' enters the policy lexicon, it becomes evidence for restrictions. The UK FCA already banned crypto derivatives for retail. The EU's ESMA has repeatedly restricted CFDs. The CFTC has gone after derivative platforms. The mainstream warning will be quoted in consumer protection submissions, and it will be used to justify position limits or product bans. The market hasn't priced the probability that this warning becomes a rule.
The Bear Case Most People Miss
Now the contrarian argument. If you are long crypto, should you be worried? Maybe not in the way you think. The conventional bear case is that retail gets educated, withdraws from perps, and the market deflates. But the more precise bear case is that the warning works too well. Leverage is a fuel. Take it away and the crypto market may become a calmer, deeper, but far more boring ocean. Retail will migrate toward spot, regulated futures, and structured products where costs are disclosed. Liquidity in the current system may drop. Spreads may widen. Perp prices could decouple from spot in episodes of stress.
That could be the most harmful outcome of all: not a sudden crash, but a slow seductive normalcy that makes crypto behave like every other asset class. The 10% funding tax would be replaced by annual management fees, custody fees, and a thousand invisible basis points from an intermediary. In that world, the innovation of decentralized settlement becomes a footnote. The bear case isn't that The Economist is right; it's that regulators will use the article to make The Economist irrelevant by killing the instrument.
Takeaway
The 10% annual drain is not a typo, a temporary market condition, or the product of a few greedy exchanges. It's the constitutional cost of a no-maturity derivative. For retail, this means a leveraged long is a decaying asset even when the chart looks strong. The rational response in a bull market isn't to close positions; it's to choose instruments where the cost flows toward you rather than away. In an industry that is busy searching for the next primitive, the quietest alpha may be the ability to stop subsidizing the other side of the trade.
Which begs the real question: what kind of market do we want to leave after the trend exhausts itself? One that charges you to believe in it, or one that pays you to stay? The Economist just put a number on that choice.
