A daily-reset leveraged token resets its exposure relative to its current value, so its objective applies to each stated reset period. Its return across several periods comes from compounding those period results. It does not promise a fixed multiple of the underlying asset's total move since purchase. Provider rules can also include intraday threshold rebalances, costs and tracking differences.
A token labelled 3x therefore needs a clock as well as a direction. Three times which return, measured between which observations, and with what happening to exposure in between? Answering those questions explains why two underlying price paths can finish at the same level while a daily-reset product finishes at different values.
Identify the rule behind the token
A reset target is a rule for the position behind the token. If an idealized long product targets 3x exposure and has $1,000 of net value, it starts the next period with $3,000 of underlying exposure. After a price move, both that exposure and the net value change. Rebalancing adjusts the position toward the target again.
Actual implementations need their own evidence. BYDFi's rebalancing documentation describes a scheduled daily adjustment at 00:00 UTC+8 and a separate threshold mechanism. That is a provider-specific schedule, not a definition of every leveraged token. A calendar-day chart can also cut across the provider's reset interval.
The word “token” tells you little about the exposure rule by itself. The distinction between a transferable holding and its underlying position also matters when comparing an LP token with a liquidity-position NFT. For a leveraged product, establish what positions or claims the issuer manages, how it values them, and what causes it to resize them.
Keep three exposure instructions separate:
| Instruction | What stays fixed? | What changes? |
|---|---|---|
| Target 3x at each reset | Exposure divided by current net value at reset | Position quantity and exposure |
| Hold an unchanged position | Underlying-equivalent quantity | Market notional and effective leverage |
| Maintain a fixed dollar notional | Dollar exposure at each adjustment | Quantity and leverage relative to equity |
The second and third instructions are different. Calling both “fixed exposure” conceals a trading decision. The comparison below uses unchanged quantity, with fixed borrowing, as its clearly defined alternative to resetting leverage.
Build a simple position ledger
All numbers that follow are hypothetical teaching inputs. They represent an idealized long 3x daily target, not a particular token or executable strategy. There are no fees, financing costs, funding payments, taxes, execution losses, external flows, threshold resets or tracking deviations. Each day's return is measured exactly between the stipulated daily reset observations. The investor's modeled token-unit holding stays constant; the exposure represented by it is resized.
Start the underlying at $100 and net value at $1,000. A simple cash-and-borrowing illustration buys 30 underlying units for $3,000, funded by $1,000 of equity and $2,000 of debt. Its initial exposure divided by equity is 3x.
For the unchanged-position comparison, keep all 30 units and the $2,000 debt unchanged. Its later equity equals the value of those units minus the debt. The leverage versus position-size guide explains why a fixed quantity's dollar sensitivity comes from that quantity, even when a displayed leverage ratio changes.
The daily-reset model follows a different rule: multiply each day's opening value by one plus three times that day's decimal underlying return. A +5% underlying return therefore gives the stipulated model a +15% return for that period. The next period uses the resulting value as its base.
Two paths, one underlying endpoint
Path A rises 5% on each of two days. Path B first rises 22.5%, then falls 10%. Both take the underlying from $100 to $110.25, a cumulative gain of 10.25%.
| Observation | Path A underlying | Path A reset value | Path B underlying | Path B reset value |
|---|---|---|---|---|
| Start | $100.00 | $1,000.00 | $100.00 | $1,000.00 |
| End of day 1 | $105.00 | $1,150.00 | $122.50 | $1,675.00 |
| End of day 2 | $110.25 | $1,322.50 | $110.25 | $1,172.50 |
For Path A, the daily-reset value is $1,000 multiplied by 1.15 and then by 1.15, producing $1,322.50. Its cumulative return is 32.25%.
For Path B, multiply $1,000 by 1.675 and then by 0.70, producing $1,172.50. Its cumulative return is 17.25%. Both products delivered exactly the assumed 3x daily returns, yet their ending values differ by $150.
Multiplying the underlying's total 10.25% gain by three gives 30.75%. Neither daily-reset path produces that number. The unchanged 30-unit position does: 30 units at $110.25 are worth $3,307.50; subtracting the fixed $2,000 debt leaves $1,307.50. Under these cost-free assumptions, that position has the same endpoint for either path because its quantity and debt never change.
This is the distinction between a daily objective and a holding-period calculation. The SEC staff's investor bulletin on leveraged and inverse ETFs explains the daily-compounding issue for ETFs. The arithmetic is relevant here, but that does not make an exchange-issued token an ETF or give it the same legal structure and protections.
Follow the exposure change that creates the difference
The return multipliers explain the result. A position ledger explains how the reset changes the next day's exposure.
At the end of Path A's first day, the original 30 units are worth $3,150. With $2,000 of debt, equity is $1,150. Restoring 3x requires $3,450 of exposure, so the model adds $300 of exposure at that observation.
Path B has a larger first-day rise. The same 30 units become worth $3,675, leaving $1,675 of equity. Restoring 3x now requires $5,025 of exposure, an addition of $1,350. The larger position then experiences the second day's 10% decline.
| First reset calculation | Path A | Path B |
|---|---|---|
| Net value before reset | $1,150 | $1,675 |
| Exposure before reset | $3,150 | $3,675 |
| Target exposure after reset | $3,450 | $5,025 |
| Exposure added | $300 | $1,350 |
| Underlying units after reset | 32.857143 | 41.020408 |
| Debt after reset | $2,300 | $3,350 |
The model finances the added exposure with more borrowing. Debt rises by $300 on Path A and $1,350 on Path B; it stays at $2,000 only in the unchanged-position comparison. Buying at the reset price changes quantity and debt together without changing equity at that instant.
Quantities in the table are rounded for display; the calculations use unrounded values. At the common $110.25 endpoint, the reset positions have gross holdings worth $3,622.50 on Path A and $4,522.50 on Path B. Subtract their respective $2,300 and $3,350 debts to recover the $1,322.50 and $1,172.50 ending values.
On Path B, the reset position loses $502.50 on day two, while the unchanged position loses $367.50. Their $135 difference reconciles the daily-reset endpoint of $1,172.50 with the unchanged-position endpoint of $1,307.50.
This ledger uses borrowing to make exposure changes visible. A provider using derivatives may implement the exposure differently. The important invariant in this illustration is exposure divided by net value at the reset, not a claim about an issuer's financing arrangements.
Path dependence also needs precise wording. The two paths above have different sets of reset-period returns. Merely reversing the same two daily returns would leave this frictionless model's final compounded value unchanged, because multiplication commutes. Intermediate values would change. Actual trigger rules, costs or other state-dependent events can require a more detailed reconstruction.
Separate compounding from fees and execution
The daily-reset return on Path A exceeds 30.75%; Path B's return falls below it. Compounding is therefore not an automatic daily fee or a rule that value must always decay. Its result depends on the returns observed between resets. The arithmetic versus geometric return comparison explains why adding percentages and linking growth factors answer different questions.
Our $150 difference contains no fees. A real product can have additional costs and discrepancies. KuCoin's leveraged-token risk disclosure identifies fees, slippage and the rebalancing algorithm among the risks, describes intraday threshold rebalancing, and warns that token value can reach zero. Its published fee figures are dated; use the applicable current terms when reconstructing a specific holding.
Do not read a smooth chart or an automatic reset as protection against losing the purchase value. Likewise, a final modeled net value is a valuation, not proof that a sale occurred at that amount. Record issuer-reported net asset value (NAV) alongside your actual purchase and sale prices, trading fees and execution timestamps.
Availability is another evidence check. Binance's 2024 notice ending its BLVT services scheduled the end of support by 3 April 2024. An old tutorial or a retained chart does not establish that the same product remains tradable. Confirm the actual listing, regional eligibility and redemption terms before using historical documentation to interpret a present holding.
Reconstruct a holding with a reset worksheet
Start with the claim the token represents. A liquid staking or liquid restaking token needs a different claim analysis from a managed leveraged position. A familiar ticker suffix is not enough to select a return model.
For the leveraged-token case, save these fields together:
| Field | What the reconstruction needs |
|---|---|
| Benchmark | Exact asset, index or contract and valuation source |
| Target | Long or inverse direction, leverage target or range |
| Reset clock | Time zone, interval and actual reset timestamps |
| Extra triggers | Threshold conditions and observed trigger events |
| Value series | Consistent observations before and after each reset |
| Costs | Applicable charges and whether value already includes them |
| Execution | Your quantities, fills, fees and any corporate-action adjustments |
Then work through each actual interval, rather than multiply one chart's cumulative return by the ticker's leverage number. If a threshold reset occurred between two daily closes, split the reconstruction at that event. If cost deductions are already in the issuer's value series, avoid subtracting them twice. A missing reset record is a missing input, not evidence that exposure stayed unchanged.
If the simplified multiplier produces zero or a negative value, stop using the continuing positive-value model and examine the issuer's extreme-market rules. That algebra does not establish an executable token price or how the product would be handled.
Finally compare three quantities: the underlying's compounded return, the value change implied by the documented exposure rule, and your return from actual fills. Retain any unexplained difference as unresolved rather than label it all “decay.” The useful output is an exposure-and-value ledger that explains each reset and reconciles to the ending holding.