ETFIQ
RankingsCompareHoldingsResearch
My portfolioWeekly note

SOXS Direxion Daily Semiconductor Bear 3X ETF

Aims to return three times the opposite of the daily move of semiconductors (SOXX)Direxion1.00% fee$1.8bnSince Mar 11, 2010SEC filings ↗Every Direxion fund ETFIQ covers
Share on XDownload as PDFCompareAdd to portfolioAlerts

Three months to Sep 30, 2026: SOXS returned −13.2% where its own daily promise gave −16.8%, 3.7 points over.

−5.1%
What SOXX did
−16.8%
What −3 times a day gives
−13.2%
What the fund returned
+3.7 pts
The fund itself added
What it aimed at, and what it did
2.4 pts ahead of its stated multipleSOXS returned −31.6% while −3 times SOXX's move would have been −34.0%SOXX +11.3% ×−3 implies−34.0%SOXS returned−31.6%2.4 pts ahead of its stated multipleSOXS returned −31.6% while −3 times SOXX's move would have been −34.0%SOXX +11.3% ×−3 implies−34.0%SOXS returned−31.6%

Over the 1 month to Sep 30, 2026, SOXS returned −31.6%, where −3 times that move implies −34.0%. The difference is +2.4 pts.

28.5 pts short of its stated multipleSOXS returned −13.2% while −3 times SOXX's move would have been +15.4%SOXX −5.1% ×−3 implies+15.4%SOXS returned−13.2%28.5 pts short of its stated multipleSOXS returned −13.2% while −3 times SOXX's move would have been +15.4%SOXX −5.1% ×−3 implies+15.4%SOXS returned−13.2%

Over the 3 months to Sep 30, 2026, SOXS returned −13.2%, where −3 times that move implies +15.4%. The difference is −28.5 pts.

113.8 pts ahead of its stated multipleSOXS returned −90.7% while −3 times SOXX's move would have been −204.4%SOXX +68.2% ×−3 implies−204.4%SOXS returned−90.7%113.8 pts ahead of its stated multipleSOXS returned −90.7% while −3 times SOXX's move would have been −204.4%SOXX +68.2% ×−3 implies−204.4%SOXS returned−90.7%

Over the 6 months to Sep 30, 2026, SOXS returned −90.7%, where −3 times that move implies −204.4%. The difference is +113.8 pts.

ETFIQ Decay Resistance Score
91.611th of 125

SOXS finished 3.7 points ahead of what -3 times SOXX’s daily move would have given, compounded through the same days.

points against its own daily promise, compounded+3.7 pts · 92nd pct
All 125 ranked by this score →How it is computed →

Did it keep up with its own daily multiple, compounded day by day?

Points against its own daily promise, compounded
QBTZ −24.9 ptsRKLZ −21.5 ptsIONZ −14.0 ptsRGTZ −10.3 ptsASTN −9.3 ptsSPCQ −6.0 ptsSMZ −5.6 ptsMUZ −3.1 ptsAPLZ −2.0 ptsSMCZ −1.1 ptsCONI −0.9 ptsSVIX −0.9 ptsSMST −0.8 ptsBMNZ −0.7 ptsSNK −0.5 ptsCORD −0.3 ptsCRCD −0.3 ptsSSPC −0.2 ptsIREZ −0.1 ptsMSTZ −0.1 ptsPLTZ −0.1 ptsSTSM −0.1 ptsNBIZ 0.0 ptsGDXD +0.1 ptsBEZ +0.2 ptsNVDQ +0.2 ptsDAMD +0.3 ptsETHD +0.3 ptsSNDQ +0.3 ptsTSLZ +0.3 ptsBTCZ +0.5 ptsCBRZ +0.5 ptsLITZ +0.5 ptsSBIT +0.8 ptsMSFD +1.0 ptsSETH +1.0 ptsEFZ +1.1 ptsSPCG +1.1 ptsBITI +1.2 ptsDULL +1.2 ptsGGLS +1.2 ptsMETD +1.2 ptsPALD +1.2 ptsAMZD +1.3 ptsBZQ +1.3 ptsYXI +1.3 ptsAAPD +1.4 ptsAMZO +1.4 ptsCSCS +1.4 ptsEUM +1.4 ptsJDST +1.4 ptsNVD +1.4 ptsNVDS +1.4 ptsORCS +1.4 ptsQCMD +1.4 ptsREK +1.4 ptsEFU +1.5 ptsNFXS +1.5 ptsSCO +1.5 ptsTSDD +1.5 ptsTSMZ +1.5 ptsDOG +1.6 ptsDRIP +1.6 ptsEWV +1.6 ptsMUD +1.6 ptsMYY +1.6 ptsNVDD +1.6 ptsRWM +1.6 ptsSBB +1.6 ptsSKRE +1.6 ptsDUG +1.7 ptsDUST +1.7 ptsPSQ +1.7 ptsSEF +1.7 ptsSH +1.7 ptsSPDN +1.7 ptsTBX +1.7 ptsAVS +1.8 ptsEPV +1.8 ptsERY +1.8 ptsTBF +1.8 ptsEEV +1.9 ptsRXD +2.0 ptsREW +2.1 ptsTSLS +2.1 ptsSMN +2.2 ptsSZK +2.2 ptsPST +2.3 ptsTSLQ +2.4 ptsWEBS +2.4 ptsFNGD +2.5 ptsSDS +2.5 ptsSKF +2.5 ptsSRS +2.5 ptsDXD +2.6 ptsQID +2.6 ptsSDP +2.6 ptsTBT +2.7 ptsFXP +2.8 ptsSIJ +2.8 ptsTECS +2.8 ptsMZZ +2.9 ptsEDZ +3.0 ptsGLL +3.0 ptsLABD +3.0 ptsSCC +3.0 ptsSDD +3.0 ptsTWM +3.0 ptsHIBS +3.2 ptsYANG +3.2 ptsSPXS +3.4 ptsSPXU +3.4 ptsSQQQ +3.4 ptsDRV +3.5 ptsFAZ +4.0 ptsSDOW +4.0 ptsTTT +4.0 ptsZSL +4.1 ptsTZA +4.2 ptsSMDD +4.3 ptsSRTY +4.4 ptsTMV +4.6 ptsTYO +4.7 ptsBIS +5.1 ptsSOXS +3.7 pts+3.7 pts−24.9 pts+5.1 pts
−24.9 pts to +5.1 pts across 125

That is the share of the 125 inverse ETFs over three months sitting at or below SOXS.

A position in this set, not a rating: it moves when the set moves, and it is not a figure to average against another desk's.

WindowFundSOXX-3x the moveDifferenceSOXX moved about
1M Aug 31 – Sep 30, 2026−31.6%+11.3%−34.0%+2.4 pts37%
3M Jul 1 – Sep 30, 2026−13.2%−5.1%+15.4%−28.5 pts47%
6M Apr 1 – Sep 30, 2026−90.7%+68.2%−204.4%+113.8 pts54%
1Y Sep 30, 2025 – Sep 30, 2026−96.7%+110.4%not meaningful over this windownot meaningful over this window46%
3Y Sep 29, 2023 – Sep 30, 2026−99.8%+266.3%not meaningful over this windownot meaningful over this window40%
Since launch Mar 11, 2010 – Sep 30, 2026−100.0%+4123.2%not meaningful over this windownot meaningful over this window31%

SOXS over each window to Sep 30, 2026. Every figure is an ETFIQ calculation. Source: ETFIQ.

Head to head
SOXS vs SOXL

In plain words

Read the multiple against the whole window instead and −3 times SOXX's −5.1% implies +15.4%, which makes SOXS look 28.5 points short. 32.2 of that is daily compounding, which happens to any −3 times fund over the same path, and the rest is the fund. SOXS aims to return -3 times SOXX's move each day, then resets. Over one day it does that. Over longer, the daily results compound, so the total is not -3 times the period's move: in a market that falls and comes back it is reliably less, and in a steady run it can be more. SOXX moved at 47% annualized over that window. That is what decides how far the two figures separate: the same fund on a calm underlying and a violent one is a different product.
Sets out to return-3 times the daily move of SOXX, reset daily
Multiple a holder actually got over the window (ETFIQ)+2.57 times, against -3 stated
UnderlyingSOXX
Segmentsector or country, sector
Net assets (issuer page, as of Sep 30, 2026)$1.8bn
Expense ratio (485BPOS XBRL, Feb 26, 2026)1.00%
LaunchedMar 11, 2010
Underlying volatility over the window (ETFIQ)47% annualized
SOXS (Direxion Daily Semiconductor Bear 3X ETF), leveraged ETFs fields as of Sep 30, 2026. Source: ETFIQ.
Returns are ETFIQ calculations from Tiingo end-of-day prices, distributions reinvested, measured against the underlying over exactly the same days. How these figures are computed

Questions people ask about SOXS

Did SOXS return -3 times SOXX?
Over the window to Sep 30, 2026, SOXX moved −5.1% and SOXS returned −13.2%. −3 times that move is +15.4%, so the fund came out 28.5 points short of it.
Why does SOXS not return -3 times over a year?
Because it resets daily. SOXS aims at -3 times each day's move, and daily results compound. Over a run where the underlying falls and comes back, compounding a levered daily return leaves a holder behind what -3 times the period move would suggest; over a steady run in one direction it can leave them ahead. The longer the holding, and the more the underlying moves about, the further apart the two figures get.
Sources and dates
Prices, SOXS and SOXXTiingo end-of-day prices ↗
Fee, 1.00%SEC 485BPOS XBRL · Feb 26, 2026
Net assets, $1.8bnIssuer page · Sep 30, 2026
IssuerDirexion website ↗
All fund filingsFund filings on SEC EDGAR ↗
SourceSEC investment company series and class file, which lists every registered fund ↗

ETFIQ links to the documents behind every figure. It is not affiliated with any issuer, and a link is not an endorsement.

How these figures are computed →Standards and sources →Open data →
Cite this page

ETFIQ, SOXS, leveraged ETFs, data as of Sep 30, 2026. https://etfiq.com/funds/soxs

Open SOXS live on ETFIQ

The label bar for SOXS, redrawn every trading night. Free to use with the credit link.

Free to use with attribution; the underlying files are at Open data.

Stated multiple
What the fund sets out to return against its underlying, each day. A fund at 2x aims to return twice the underlying’s daily move, and an inverse fund at -2x aims to return twice it in the opposite direction.
Daily reset
The fund starts each day aiming at the multiple again, from wherever it now stands. It is the mechanism that makes these funds behave as they do, and the reason the stated multiple applies to a day rather than to a holding period.
Compounding
Daily results multiply together rather than adding up. Two days of the multiple is not twice the multiple, and over a run of days the difference between the two grows with how far the underlying moves about.
Decay
What compounding costs a holder when the underlying falls and rises back to where it started. The underlying is level and the fund is down; the more violently it moved, the further down.
Difference against the stated multiple
The fund’s return minus the stated multiple times the underlying’s return, in percentage points. The one figure that says whether a holder got what the fund’s own multiple describes.
Every term used here, defined in full on the leveraged ETFs vocabulary page. →
What is an inverse ETF?→Direxion: every fund ETFIQ covers→Leveraged ETFs against their own promise, 2026-Q2→Leveraged ETFs against their own promise, August 2026→