The mistake I see over and over is that people size a leveraged position on the price move they expect and never price the cost of just sitting in it. They treat a perp long like it is free to hold, the way spot is free to hold. It is not. A leveraged position is a rented position, and the rent shows up in three places that most people only add together after the trade has already gone sideways on them.
None of this is exotic. It is arithmetic. But the numbers compound quietly enough that a position which looks like a clean directional bet is actually a bet that a move arrives before the carry eats it. Once you see the carry as its own line item, a lot of trades that felt like conviction start to look like you paying a toll and hoping the move covers it.
The three costs that never show up as a single number
The first is funding. On a perpetual future, longs and shorts exchange a payment on a fixed schedule, typically every eight hours, so three times a day. When the market is skewed long, which is most of the time in a bull tape, longs pay shorts. The rate looks tiny per interval, something like a fraction of a percent, and that smallness is exactly what fools people. A rate that looks harmless per eight hours annualizes into a number that would make a credit card blush. When funding is running hot, it is not unusual to see it annualize north of 30 percent, and in genuinely frothy conditions much higher than that.
The second is borrow. If you are on margin rather than a perp, you are borrowing the asset or the quote currency to open the position, and that loan carries an interest rate that also accrues continuously. On a perp the funding rate absorbs most of this, which is why I mostly talk in funding terms, but if you are running isolated margin on a spot exchange or holding a dated contract on borrowed collateral, the borrow line is separate and real.
The third is the one nobody budgets for: fee drag from re-entries. Leverage forces you to be twitchy. A position that would never threaten a spot holder can wick into your liquidation zone, and either you get stopped, or you close early to avoid it, or you get liquidated outright. Every time you rebuild the position you pay taker fees on the full notional, not on your margin. Do that three or four times across a choppy month and the round-trip fees alone can rival a chunk of your funding bill.
A 30-day long, worked out loud
Take a long you intend to hold for a month. Assume funding is annualizing around 36 percent, which is roughly a tenth of a percent per day, split across three payments. That is not a stress scenario. That is an ordinary hot-but-not-crazy tape.
Over 30 days, a tenth of a percent per day compounds to roughly three percent of the notional in funding alone. Notice the word notional. This is the trap. Funding is charged on the size of your position, not on the margin you posted. If you are running five times leverage, that three percent of notional is fifteen percent of your actual capital, gone to funding, before the price has done anything at all.
Now layer the fees. Say the tape is choppy and you rebuild the position twice after getting shaken out. At a taker fee of a few hundredths of a percent per side on full notional, a couple of round trips adds another meaningful slice, again measured against notional. Add it up and the price has to move a few percent in your favor just to get you back to flat on the month. At five times leverage, the underlying only needs to drift against you by a fraction of that few percent, plus your buffer to liquidation, and you are out before the thesis even had a chance to play.
The uncomfortable version of this: a leveraged long in a high-funding regime is short time. You are not just betting on direction. You are betting the move shows up fast. The longer you are right slowly, the more of your edge you hand to the people on the other side of funding.
A quick way to decide how to hold the exposure
Before I open anything leveraged, I run the same short checklist. It takes two minutes and it has talked me out of more bad holds than any chart ever has.
- Annualize the current funding rate. Multiply the eight-hour rate by three, then by 365. If that number is above roughly 20 to 30 percent, treat the perp as expensive rent and ask whether you actually need the leverage.
- Estimate your hold horizon honestly. A three-day tactical long and a six-week thesis are completely different animals. Carry is close to irrelevant on the former and dominant on the latter.
- Multiply funding cost by your leverage to see it as a percentage of your capital, not your notional. This is the number that actually hurts, and it is the one the exchange interface hides from you.
- Add a fee budget for the re-entries you will realistically need. If the setup is choppy, assume two or three rebuilds, not zero.
Once you have those four numbers, the alternatives sort themselves out. If you want the exposure and you plan to hold for weeks, spot is almost always cheaper, because spot pays no funding and cannot be liquidated. The cost of spot is opportunity cost on the capital you tie up, nothing more. You give up the leverage, but in a high-funding regime the leverage was quietly costing you more than it was worth.
Dated futures are the middle path. A quarterly contract prices its carry into the basis up front, so instead of paying funding every eight hours and guessing at the running total, you pay a known premium baked into the price when you buy. If that annualized basis is cheaper than where perp funding is running, the dated contract is the better vehicle for a multi-week hold, and you lose the twitchiness of a rate that can spike on you mid-position.
Options change the shape of the problem entirely. You pay a premium once, and that premium is the maximum you can lose. There is no funding accruing against you and no liquidation to defend. The cost is theta, the decay of the option as expiry approaches, and in a very high-funding environment the premium on a call can actually work out cheaper than a month of funding on the equivalent leveraged long. Options are not free and they are not simple, but for a defined-horizon directional view when carry is punishing, they are worth pricing out rather than dismissing.
What this changes in practice
The habit worth building is small. Before you size a leveraged position, price the carry as its own line and compare it against spot, a dated contract, and a call for the same view. Most of the time you will still take the perp, because the hold is short and the rate is calm. But the times you catch yourself about to pay 30 percent annualized to hold a slow thesis are the times this saves you real money, and those are exactly the times the position felt like conviction rather than a toll you forgot you were paying.
Leverage is a tool for expressing urgency, not for expressing patience. If your view needs weeks to be right, hold it in something that does not charge you rent by the hour.