When a bull market arrives, almost nobody wants to just hold spot. The instinct is to add leverage, and for most traders that means a perpetual future. So the real question is not whether to use leverage—it is how to get it. We tested two ways to hold the same 2X exposure: a 2X perpetual, and RiskON, which is built to deliver roughly 2X on its own. Across the BTC and ETH bull markets of the past seven years, RiskON won every one—14 of 14. Compounded through them, one dollar in RiskON grew to $27 in BTC and $109 in ETH; the same dollar in a 2X perpetual grew to just $12 and $22. Same target exposure, very different outcomes—and the difference is a cost the perpetual pays and RiskON does not.
RiskON: Leverage Without the Borrowings, Funding, and Margin Calls
Before the comparison, it is worth knowing what RiskON actually is, because it is neither a perpetual nor an option you trade on an exchange.
Deposit one unit of an asset into The Risk Protocol and it splits into two tokens, RiskON and RiskOFF. Each starts with half the economic value of the deposit, and the two always sum back to the underlying. RiskOFF is the cautious half: it gives up its gains above a +8% cap in return for protection against losses past −5%. RiskON is the other half, and it inherits exactly what RiskOFF gives up.
That inheritance is the leverage. Above +8%, RiskON earns its own gain plus the gain RiskOFF surrendered—roughly 2X leverage on the part of the move beyond the cap. A +30% rally becomes +52%; a +100% rally becomes +192%. Between −5% and +8% it tracks the asset one for one, and below −5% it takes the downside RiskOFF was shielded from, again at 2X. Drag the slider to see the whole payoff:
Where does the extra upside come from? Not from a loan. RiskON and RiskOFF are the two sides of a single fully collateralized, costless collar, struck against each other the moment you mint them: RiskOFF is long a protective put and short a call, and RiskON is long that same call and short that same put. In plain terms, RiskOFF buys insurance against deep falls and pays for it by capping its own upside, and RiskON is the counterparty on both sides of that trade. The leverage is simply the two tokens swapping risk with one another—an internal risk swap, in the protocol's own words—not capital borrowed from an exchange. Because nothing is borrowed, three costs a perpetual cannot escape never arise.
- No borrowings. The pair is 100% collateralized at mint. The 2X is an internal swap between the two tokens, not external debt.
- No funding. This is not a perp, so there is no borrowed leverage and no ongoing financing bleed. Whatever the structure costs is priced once into the collar at mint—not streamed out every hour, and not inflated by how long the crowd happens to be.
- No margin calls. With no borrowed funds to repay, there is nothing to margin-call or force-liquidate. RiskON's value can fall in a deep crash, but it cannot be closed out from under you.
The collar re-strikes every 30 days, resetting the band around the prevailing price and rolling the prior month's gain into the next. And because the payoff is fixed for the length of that epoch rather than rebalanced day by day, there is no churn for volatility to grind on—a property that matters far more than it sounds, as the next two sections show.
Rallies are amplified. Drawdowns are contained. Nothing is borrowed.
How We Measured This
The result lives or dies by the method, so start there. We used seven years of daily BTC and ETH prices, together with the funding data for their perpetual markets, from 2019 to 2026.
A bull market is defined mechanically. A day counts as bullish when the price closes above its 200-day moving average and that average is itself rising. Both tests read only past prices, so every entry and exit is one a trader could have acted on that morning, with no hindsight. The 200-day average is the standard long-term trend line: shorter windows whipsaw on noise, longer ones lag the move, and 200 days is long enough to mark the primary trend yet quick enough to turn within a cycle. Runs of bullish days form a regime, and the filter returns 8 of them in BTC and 6 in ETH, matching the cycles anyone who traded them would recognise.
In each bull market we put one dollar into each of two strategies on the same day. One is a 2X perpetual: constant 2X exposure, rebalanced daily, paying funding on the full 2X notional. The other is RiskON. We hold both to the end of the regime, sit in cash in between, and compound the results across every bull market. The figures throughout this piece are what came out.
The Tax Nobody Prices In
A perpetual future has no expiry date, and that single fact is the source of its cost. A dated future is dragged toward spot because it must settle on a fixed day. A perpetual never settles, so something else has to keep its price tethered to the asset it tracks. That something is the funding rate—a payment exchanged between longs and shorts at a regular interval. It is the contract's cost of carry, and in a market that leans long, the longs pay it.
It is tempting to think funding is zero when longs and shorts are balanced. It is not. Hyperliquid makes the mechanics explicit; its funding rate is
funding = premium + clamp(interest − premium, −0.05%, +0.05%)
where the interest term is fixed at 0.01% every 8 hours—about 11.6% a year—and is charged every hour. With the book perfectly balanced and the premium at zero, that interest term is exactly what a long pays. It is the floor, not the ceiling: holding a 2X long is holding a financed position, and the financing is never free.
The premium is where bullishness enters. It is the gap between the perpetual's price and spot, premium = (perp − oracle) / oracle. When longs are aggressive they bid the perpetual above spot, the premium turns positive, and funding climbs with it—once the premium clears the small clamp, funding tracks it almost one for one, up to a ceiling of 4% per hour. Funding is, in effect, a thermometer for how badly the market wants to be long: a balanced book reads near the 11.6% floor, and a crowded long tilt reads far hotter.
A bull market is a crowded long tilt, by definition. So its funding is not occasionally high. It is structurally, persistently high.
The data bears it out. Across the bull markets in this study, a long paid funding of 17% a year in BTC and 22% in ETH, and a 2X position pays all of it on twice the notional. RiskON, holding no borrowed position, pays none of it.
The Second Tax: Volatility Decay
Funding is the cost you can see. The second one never appears as a payment, and it is the larger of the two.
A constant 2X position has to be rebalanced to stay constant. As the price moves, it is topped up after a gain and trimmed after a loss so that exposure stays at exactly twice the equity. On a straight-line climb that would cost nothing. Prices do not move in straight lines.
Two days show the mechanism. The price rises 10%, then falls 10%. The asset travels $100 to $110 to $99 and lands down 1%. The 2X position travels $100 to $120 to $96 and lands down 4%. No fee was charged; the loss is pure path. The 10% recovery is figured on a smaller base than the 10% drop destroyed, and doubling every move widens that gap—the same reason a 50% loss needs a 100% gain to undo it.
The effect has a closed form. A position rebalanced to a constant k times the underlying compounds the leveraged daily return, and over a volatile path its growth falls short of k times the asset's by about ½·k·(k−1)·σ²·T, where σ is volatility and T is time. For 2X that penalty is σ²·T—a full unit of the asset's variance, deducted from the leveraged return for every unit of time. It grows with the square of the moves: a 10% round trip costs about 2%, a 20% round trip about 8%, a 40% round trip about 32%.
Put numbers on it. BTC and ETH have run at 60% to 80% annualized volatility; at those levels σ² is 0.36 to 0.64, so a daily-rebalanced 2X bleeds roughly 35% to 65% a year to decay alone, before a cent of funding. It is the same force—volatility drag, or beta slippage—that quietly erodes every leveraged ETF and leveraged token.
Funding is the rent on borrowed exposure. Decay is the toll on rebalancing it. A 2X perpetual pays both, every day it is open.
RiskON is re-struck once a month and never rebalanced in between, so the daily churn that creates decay simply never happens. It pays neither tax.
The Findings
In 8 of 8 BTC regimes and all 6 in ETH, RiskON beat the 2X perpetual. Not on average. In every one.
Compounded through the BTC regimes, one dollar in RiskON grew to $27. The 2X perpetual reached $12—barely past the $11 earned by holding the coin with no leverage at all. Both aimed at the same 2X; the perpetual surrendered the difference to funding and decay.
ETH is the cleaner verdict. RiskON became $109; the 2X perpetual became $22; unleveraged spot became $27. The 2X perpetual finished below holding the coin with no leverage at all, while RiskON multiplied it many times over. Same target exposure, opposite outcomes.
And it is not the gift of one lucky run. RiskON wins regime after regime, in the long trends and the short ones alike.
| Start | End | Days | Underlying | RiskON | 2X Perp | RiskON DD | 2X Perp DD | Funding/yr | Winner |
|---|---|---|---|---|---|---|---|---|---|
| 2019-09-10 | 2019-09-25 | 15 | -16% | -28% | -32% | -30% | -36% | +11% | RiskON |
| 2020-05-08 | 2021-05-20 | 377 | +314% | +624% | +431% | -68% | -71% | +33% | RiskON |
| 2021-08-09 | 2021-09-09 | 31 | +0% | +0% | -6% | -16% | -24% | +18% | RiskON |
| 2021-10-21 | 2021-12-27 | 67 | -19% | -35% | -43% | -47% | -57% | +21% | RiskON |
| 2023-01-26 | 2023-08-16 | 202 | +25% | +41% | +30% | -25% | -35% | +7% | RiskON |
| 2023-10-16 | 2024-08-08 | 297 | +117% | +201% | +198% | -39% | -55% | +14% | RiskON |
| 2024-11-02 | 2025-03-27 | 145 | +26% | +45% | +32% | -40% | -50% | +9% | RiskON |
| 2025-04-22 | 2025-11-02 | 194 | +18% | +30% | +25% | -17% | -29% | +5% | RiskON |
BTC, every bull market. The regimes that end below zero are stretches where the trend filter held into the first leg of a reversal; RiskON loses less in each, because it pays no funding and takes the early decline at 1X.
| Start | End | Days | Underlying | RiskON | 2X Perp | RiskON DD | 2X Perp DD | Funding/yr | Winner |
|---|---|---|---|---|---|---|---|---|---|
| 2020-04-30 | 2022-01-06 | 616 | +1550% | +6057% | +1901% | -78% | -88% | +33% | RiskON |
| 2023-01-14 | 2023-02-23 | 40 | +6% | +8% | +7% | -10% | -20% | +10% | RiskON |
| 2023-03-23 | 2023-08-16 | 146 | -1% | -1% | -12% | -31% | -43% | +6% | RiskON |
| 2023-11-19 | 2024-07-31 | 255 | +61% | +88% | +63% | -49% | -59% | +16% | RiskON |
| 2024-11-24 | 2024-12-23 | 29 | +2% | +2% | -3% | -25% | -35% | +17% | RiskON |
| 2025-07-28 | 2025-11-10 | 105 | -6% | -14% | -26% | -52% | -59% | +5% | RiskON |
ETH, every bull market. RiskON wins all 6.
The Downside
A leveraged-upside story usually hides a leveraged-downside cost. RiskON does not, and this is where the structure earns its keep.
Between its strikes, RiskON is a one-times claim. A 5% pullback costs it 5%. The 2X perpetual loses 10% on the same pullback, twice as much, and RiskON only steepens to 2X once a decline pushes past its floor—so it takes the early, ordinary part of every dip at half the perpetual's rate.
This is visible in realized risk, not just on paper. Averaged across the regimes, RiskON's worst drawdown inside a window was 35% in BTC against the perpetual's 45%, and 41% against 51% in ETH.
Higher returns and shallower drawdowns, in the very same windows.
Does It Survive Scrutiny?
A result is only as strong as its weakest assumption, so the comparison was re-run under every reasonable change. The bull-market definition was loosened. The perpetual was pushed to 3X. The funding was sourced from a second venue. Every version a trader could act on in real time still favors RiskON, in both direction and scale.
| Variant | Type | BTC wins | BTC ratio | ETH wins | ETH ratio |
|---|---|---|---|---|---|
| 30-day epoch, +8%/-5%, 2X Binance, MA-slope | causal | 8/8 | 2.2× | 6/6 | 5.0× |
| Bull = price > 200d MA | causal | 9/9 | 2.7× | 6/6 | 6.5× |
| 3X perp leverage | causal | 7/8 | 3.4× | 6/6 | 29.3× |
| Deribit funding venue | causal | 7/8 | 1.9× | 6/7 | 2.9× |
| Bull = trough-to-peak (20%) | hindsight | 0/12 | 0.2× | 1/16 | 0.1× |
There is exactly one way to make the perpetual win, and it is worth naming. Define a bull market by its top and its bottom—after both have already happened—and the windows become clean, uninterrupted climbs that never dip more than 20%. On those, the perpetual's two costs nearly disappear: there is no chop for decay to feed on, and a short clean run accrues little funding. That is the lone row marked hindsight above. But that definition needs a time machine. The 200-day filter has none; it holds through the chop, the pullbacks, and the rich funding that define a real bull market, which is the exact terrain where RiskON wins. The perpetual's edge lives only in a chart you can draw after the fact. RiskON's lives in the markets you can actually trade.
The Better Way to Be Bullish
The reflex to add leverage in a bull market is sound. The reflex to source it from a perpetual is the costly part. A perpetual finances your leverage and bills you for the financing—funding that climbs precisely when you reach for the exposure, and decay that grinds on every swing. RiskON manufactures the same 2X from structure rather than borrowing, so there is nothing to finance, nothing to stream, and nothing to liquidate. Across every bull market a trader could have identified in real time, RiskON beat the 2X perpetual: more upside, shallower drawdowns, the same conviction.
The lesson is not that leverage is wrong. It is that how you source your leverage matters more than the leverage itself.