Leveraged & inverse ETFs

Definition

Leveraged and inverse ETFs deliver a MULTIPLE (2x, 3x, −1x, −2x, −3x) of an index’s DAILY return, reset every day. The daily reset is the entire story: over any period longer than one day, the compounded product of daily multiples is not the multiple of the period return — a path-dependent gap (“volatility drag”) that grows with volatility and holding period. These are daily instruments; multi-week holding is a different, usually worse, position than the label implies.

How it works / structure

• Mechanics: the fund holds swaps/futures sized to the target multiple of NAV and REBALANCES DAILY to restore the ratio — buying exposure after up days, cutting after down days (momentum-chasing by construction; in size, a documented end-of-day flow).
• The compounding arithmetic: index goes +10% then −10% (net −1%); the 2x fund goes +20% then −20% (net −4%); the gap is ≈ −(k² − k) × σ²/2 per period scaled by variance — the higher the volatility, the faster the decay (Cheng-Madhavan derive the exact expression).
• Engine-relevant parameters: multiple k, underlying index, expense ratio (high for the category), and — for replay — the daily-reset compounding modeled exactly, never approximated by k × period return.
• Inverse funds: same reset arithmetic in reverse; −1x held long is NOT equivalent to a short position (strategy-short-selling), which compounds differently and has different costs.

When it applies

Single-day to few-day tactical expressions where the borrow/ margin alternative is unavailable or slower; intraday instruments in strategy-day-trading-styles universes (their liquidity is genuinely deep in the majors). The platform treats any multi-week leveraged-ETF holding thesis as requiring the decay arithmetic acknowledged in the thesis itself.

Risk profile & failure modes

• Volatility decay: in choppy markets both the 2x and the −2x fund on the same index can LOSE money over a month — the signature result; realized examples are abundant in every high-vol episode.
• Buy-and-hold mismatch: the 2009 SEC/FINRA alert exists because retail held daily instruments for months and received outcomes unrelated to the label.
• Extreme-day termination risk: −3x funds facing a +34% underlying day face NAV wipeout mechanics; several volatility-linked products terminated in the February 2018 episode (instrument-vix-futures).
• Rebalance-flow signaling: known end-of-day rebalance flows in size invite front-running costs paid by the fund.

Evidence & limits

Cheng-Madhavan (2009) is the canonical treatment: the exact compounding arithmetic, the variance-drag expression, and the end-of-day rebalance flow estimates. The SEC/FINRA alert documents the regulatory view. The decay is mathematics, not opinion; its SIZE per period is a function of realized variance — small in quiet trends (leveraged funds can beat k× in smooth trends; the same arithmetic cuts both ways), large in chop.

Falsifiable-thesis examples

Illustrations only, not signals:

• “Held for the next 60 sessions, the 2x fund on index I will return less than 2× the index’s period return, given realized daily vol above 1.2%” — falsified by the recorded pair.
• “The −1x fund on I will underperform a modeled short of I (borrow at quoted rates) over the next quarter” — falsified by the paired accounting.

Cross-references

• Plain wrapper: instrument-etf; the honest short alternative: strategy-short-selling
• The decay driver: regime-volatility (variance regimes)
• The catastrophic cousin: instrument-vix-futures (2018 termination episode)
• Sizing honesty: risk-scenario-analysis