Leveraged ETF Returns: Daily Resets, Compounding and Actual P&L

A benchmark can finish exactly where it started while a daily 2× ETF loses money—even when the fund hits its daily target on both days. The multiplier applies to each reset period, not to the benchmark’s cumulative return since you bought the shares. To investigate a disappointing result, rebuild the daily path first, compare it with the fund’s net asset value, and only then reconcile your own execution prices and costs.

The opposite shortcut is also wrong: daily reset does not make underperformance inevitable. A sustained move can produce a compounded return above the benchmark’s cumulative gain multiplied by two. The useful question is which return you are measuring, over which intervals, and against which objective.

Find the reset period before doing the arithmetic

A daily leveraged ETF generally seeks a stated multiple of a specified benchmark’s one-day performance, often before fees and expenses. It may obtain exposure through derivatives and rebalance that exposure relative to changing net assets. This is an objective, not a guaranteed daily outcome. A leveraged exchange-traded note is not automatically the same legal or economic product as an ETF.

Read the benchmark definition, positive or negative multiple, reset frequency, valuation times and exceptional-adjustment provisions. Establish whether the benchmark includes distributions. A monthly-reset product requires different intervals; special intraday provisions also matter. Buying halfway through a session does not entitle you to a constant stated multiple of the benchmark’s move from your entry time.

A flat benchmark, but a lower fund value

Consider an original hypothetical calculation. Both benchmark and fund start at 100. Valuation times align, daily targets are met exactly, and fees, financing, distributions and tracking differences are excluded. The benchmark rises 8% to 108, then falls by 8/108, or about 7.4074%, back to 100. The second-day decline is not 8%: it starts from a different base.

The 2× fund loses approximately 14.8148% on day two, but that loss applies to 116: 116 × (1 − 14.8148%) ≈ 98.8148. The inverse fund gains on day two from the smaller base of 92. Both can meet every daily target and still lose over the round trip. No additional fee is needed to explain this particular shortfall.

Ideal ending value = starting value × the product of (1 + L × r) across reset periods, where L is the signed leverage multiple and r is each period’s benchmark return. Multiplying the cumulative benchmark return by L is a different calculation. Do not extend this simplified model through a zero or negative daily fund factor and present a negative theoretical NAV as an actual product outcome; extreme-event provisions must be checked.

The same 8.16% benchmark gain can deliver different returns

On path A, the benchmark gains 4% on each of two days: 100 → 104 → 108.16. The ideal 2× fund follows 100 → 108 → 116.64. Its 16.64% gain exceeds twice the cumulative benchmark gain, which is 16.32%. Compounding is not automatically a drag.

On path B, the benchmark rises 16% to 116, then falls approximately 6.7586% to the same endpoint, 108.16. The 2× fund reaches 132 before falling approximately 13.5172%, ending at 114.1572. Its gain is only 14.1572%—about 2.4828 percentage points less than on path A. Endpoint charts alone cannot explain the difference.

Be precise about “path dependence.” These paths contain different daily returns. Merely rearranging an identical set of daily returns leaves their product unchanged in this ideal model with no cash flows, changing costs or exceptional rules. It is not correct to claim that reversing the order of the same two returns must change the final value.

Break-even and exposure move with the account

After the round trip, the 2× fund is at 98.8148. Under the same exact-tracking, zero-cost assumptions, one further day would require a benchmark gain of approximately 0.5997% to restore the fund to 100: (100 / 98.8148 − 1) / 2. The benchmark would then be around 100.5997. That is a threshold for this constructed example, not a universal recovery target.

A reset adjusts exposure; it does not erase losses. In a simplified equity-equivalent illustration, NAV of 100 supports target exposure of 200. An 8% benchmark gain brings NAV to 116 and the existing exposure to 216; the new 2× target is 232, requiring 16 more exposure. This explains the mechanism, not the actual notional accounting of a particular swap portfolio.

Keeping the same number of ETF shares therefore does not keep the currency amount of risk fixed. Reassess benchmark exposure and the amount at risk after a large move. Waiting for an old index level is not a substitute for checking whether the position still fits the intended holding period and risk budget.

A 4% NAV gain can become a 1.76% account gain

Now separate fund performance from execution. Suppose observed NAV rises from 10 to 10.40, already reflecting actual fund-level expenses. You buy 100 shares at 10.20, a 2% premium, and sell at 10.40 when the premium has disappeared. Before commissions, your price return is 10.40 / 10.20 − 1 ≈ 1.9608%, not 4%.

With a hypothetical commission of 1 currency unit on each trade, no distributions, taxes or currency conversion, the initial outlay is 1,021 and net sale proceeds are 1,039. The gain is 18 / 1,021 ≈ 1.7630%. Restore the relevant tax and FX treatment for a real account. Do not deduct the entire expense ratio again from an observed NAV return that already incorporates those expenses.

Bid–ask spreads, depth and partial fills create another layer. A displayed NAV or last trade is not a promise that your whole order will execute there. The guide to limit-order queues and partial fills helps distinguish execution shortfall from fund tracking differences.

Reconcile four layers instead of calling everything decay

1. Objective: identify the benchmark, signed multiple, reset interval and valuation window. Align currency and distribution treatment.
2. Ideal path: compound each reset period separately, retaining precision. This isolates the effect of daily leverage before actual costs.
3. Fund NAV: compare the corresponding published fund series. Investigate any residual using expense, financing, replication and valuation information; do not label the entire gap a fee.
4. Your account: incorporate actual quantities, fills, commissions, taxes, distributions and FX. With external cash flows, select an appropriate, consistent return measure before comparing it with the fund.

If the ideal path explains most of the gap, the original comparison used the wrong target. If daily NAV still diverges materially, investigate the product and data. If NAV behaves as expected but the account lags, start with execution and costs. Those are three different diagnoses, and they call for different checks.