← 블로그← All posts

2026-09-22

The −100% Club: Every Leveraged and Inverse ETF in Our Universe, Measured

By FinScope Research · Data as of 2026-09-18

Most warnings about leveraged ETFs are qualitative. You are told about “volatility decay” and “daily reset,” you nod, and you buy the fund anyway because the mechanism sounds like a rounding error. This article replaces the adjective with a census.

We classified all 815 US stocks and ETFs in the FinScope universe with the same rule the site’s own screener uses, which reads the fund’s legal name for leverage and inverse markers (2x, 3x, Bull, Bear, Ultra, UltraPro, Inverse, Short). That isolates 52 leveraged or inverse products, leaving 763 ordinary stocks and funds as the control group. Then we measured both groups on the same two questions: how far did it fall from its all-time high, and did it ever get back?

The two groups do not belong on the same chart

The median ordinary security in this universe has, at some point, lost about 71% of its value. That is already a brutal number, and it is the honest baseline: owning equities means surviving deep holes. The median leveraged product lost 95.1%.

The row that matters most is the fourth one. Among ordinary securities, roughly one in five never reclaimed the high it fell from. Among leveraged and inverse products, it is nearly two in three. These are not instruments that fall hard and bounce back. In this dataset they are mostly instruments that fall hard and stay down.

The fifteen at −100.0%

Fifteen products in the universe show a maximum drawdown that rounds to −100.0% from their split-adjusted all-time high.

Two things in that table deserve more attention than the headline number.

First, the trough dates. Fourteen of the fifteen bottomed within the last twelve months, and none of the bear funds bottomed in a market crash. Their all-time low is the present. A stock at −100% from a 2000 peak is a historical artifact; these funds are setting new lows now, because the loss is not an event, it is the operating characteristic of the product over a long market uptrend.

Second, fourteen of the fifteen are inverse funds. That is the point people miss when they read “leveraged ETFs decay.” The single long 2x fund on the list, NUGT, got there through the gold-miner collapse. The rest got there by being short a market that went up for fifteen years while resetting their exposure every single day.

What “−100.0%” actually means here

We round to one decimal place, so −100.0% is not literally zero. SOXS is the clearest case. Its split-adjusted price series peaks on 2010-08-31 and ends at $41.49 on 2026-09-18, a ratio of roughly 1.2 × 10¹⁰ — the adjusted peak is about twelve billion times the current price, which is a decline of about −99.999999%.

That enormous adjusted peak is a consequence of repeated reverse splits. When a fund does a 1-for-10 reverse split, every price before it is divided by ten for comparability. Do that often enough over fifteen years and the back-adjusted 2010 price becomes astronomically large. UVXY shows a peak-to-current ratio near 2.9 × 10¹⁰; SQQQ and TZA sit around 2.5 × 10⁵.

It is worth being precise about what this does and does not mean:

• It is economically real. An investor who bought at inception and held, adjusting for every reverse split, would have essentially nothing left. The reverse splits are how the fund keeps a tradeable share price while the underlying value collapses; they do not restore value.
• It is not a statement that a share bought today will go to zero tomorrow. These funds track their daily objective reasonably well over a single day. The destruction happens through compounding across many days.
• It does make a lifetime max-drawdown figure nearly useless for comparing one of these funds to another. Once several products all round to −100.0%, the metric has no resolution left. For these instruments, look at the common-window comparison over a period you would actually hold, not the lifetime number.

This is the same class of measurement trap we documented when a stock split corrupted a ticker’s drawdown history in how a stock split poisoned our drawdown data. The arithmetic is correct; the interpretation is where people get hurt.

Why the survivors look deceptively healthy

Nineteen of the 52 leveraged products did recover their worst drawdown, and their median recovery took less than a year, which looks fast next to the 3.7-year median for single stocks we measured in recovery times for 800+ stocks and ETFs.

That speed is a selection effect, not a feature. A leveraged fund amplifies the rebound as well as the fall, so when it does recover it recovers quickly. But the 33 funds that never came back are excluded from that median by construction — you cannot measure recovery time for a recovery that has not happened. Read the “fast recovery” number only alongside the 63% that are not in it.

What to do with this

• Check the holding period the number describes. A lifetime drawdown on a daily-reset product describes an investor nobody actually is. Use FinScope’s period control to measure the window you plan to hold.
• Treat inverse funds as trades with a clock. The census here is not “inverse funds are badly run.” It is that a product resetting daily short exposure against a long-rising index accumulates loss structurally, and fourteen of our fifteen worst cases are exactly that.
• Do not read −100.0% as the floor. It is a rounded display of a number that keeps getting smaller. Several of these funds set their all-time low within the last three months of this dataset.

The per-fund pages carry the full history and the underwater chart: SQQQ, SOXS, UVXY, TQQQ, SOXL. For the mechanism rather than the census, see the leveraged ETF guide and what TQQQ’s −82% actually says.

FAQ

How many leveraged or inverse ETFs are down 100%? Fifteen of the 52 leveraged and inverse products in this 815-security universe show a maximum drawdown that rounds to −100.0% from their split-adjusted all-time high. Fourteen of those fifteen are inverse (bear) funds.

Is a −100% drawdown a data error caused by reverse splits? No, though reverse splits are why the number is so extreme. Back-adjusting for repeated reverse splits makes the historical price very large, and the decline from it is genuine for a buy-and-hold investor. What it does mean is that the lifetime figure has lost the resolution to compare these funds with each other.

Do leveraged ETFs recover faster than normal funds? The ones that recover do, with a median under a year in this data. But 63% of leveraged and inverse products here never recovered their worst drawdown at all, versus 21% of ordinary securities, and those failures are invisible in any recovery-time average.

Methodology: leverage classification comes from the security’s name using the same rule as the FinScope screener (classifyAsset). Drawdown is measured on split- and dividend-adjusted daily closes; maximum drawdown is the largest peak-to-trough decline in the security’s full history, and “never recovered” means the close has not yet reclaimed that prior peak. Figures cover 815 securities as of 2026-09-18. See the methodology page and the editorial policy.