UVXY vs VXX vs SVXY: Do VIX ETFs Always Lose Money?

On 5 November 2026 SPY, the big S&P 500 ETF, fell 4.4%. VIX, the market's fear gauge, jumped from 16.61 to 28.18. That's a 70% rise in one day, the third biggest since 1990.

If you held UVXY, you made 45.4%. VXX made 30.1%. SVXY, the fund that bets the other way, lost 15.4%. Days like that are why people buy these funds. So we tested what happens on all the other days.

UVXY vs VXX vs SVXY at a glance

All of these follow one index, the S&P 500 VIX Short-Term Futures Index. They differ in how much of it they give you, and in which direction. We've added VIXY because it's the 1x fund with the longest record.

The fees barely matter here. A fund that loses half its value in a typical year isn't losing it to a 0.95% charge. What matters is what these funds own, and it isn't VIX.

What $10,000 became in each

We put $10,000 in each fund at the first close of UVXY and SVXY, on 4 October 2011, and followed it to 5 November 2026.

VXX's current note started on 25 January 2018. From that day, $10,000 in VXX became $125 and $10,000 in VIXY became $115. They're close to the same thing. UVXY became $1.88.

The share prices hide all this, because these funds keep doing reverse splits. UVXY has done 13 of them since 2011. Together they turn 15 billion old shares into one new one. Its first close was $34.30, which on today's share count is about $514.5 billion a share. It closed yesterday at $23.61.

Here's the part that surprises people. VIX didn't fall anywhere near that much. It went from 40.82 to 28.18, a 31% drop, and October 2011 was a nervous time to start. So try a calm start instead. VIX's lowest close in the study was 9.14, on 3 November 2017. It has since risen 208%. VIXY has fallen 99% over the same days.

Why they fall: VIX futures usually cost more than VIX

VIX is an index. It's worked out from S&P 500 option prices, and it shows how big a move the market expects over the next 30 days. You can't buy it. What you can buy is a VIX future, a bet on where VIX will be on a set date.

So that's what these funds hold. Their index keeps about one month to expiry on average, by selling some of the nearest future every day and buying the next one out.

Most of the time, later futures cost more than nearer ones, and all of them cost more than VIX. That's called contango. It's normal, because when VIX is calm, traders expect it to rise from there, and they'll pay for protection.

Here's a typical day. On 19 May 2026 VIX closed at 18.06. The June future, with 29 days left, settled at 20.42. The July future settled at 21.72, 6.4% higher.

A future ends at VIX's level on its expiry date. So if VIX just sat at 18.06 for those 29 days, the June future would fall 11.6%. Nothing happened and you'd still lose. Then the fund sells it and buys the pricier July future, and the same thing starts again.

We checked every day of Cboe's futures data since January 2013. On 84.9% of 3,486 days, the second future cost more than the first. The median gap was 6.4%. Paying something like that every month is how a fund loses half its value in a year.

The opposite is called backwardation, and it's what panic looks like. Yesterday the first future settled at 24.54 and the second at 22.19, 9.6% lower. The day before, the second was 1.4% higher. While that lasts, the roll works in the funds' favour. It rarely lasts. The bars above show how few days it gets.

UVXY has a second problem. It resets to 1.5x every day, so it also pays the same decay as other leveraged funds. Our TQQQ article walks through that maths. Since the switch to 1.5x, $10,000 in VIXY became $72 and $10,000 in UVXY became $1.10.

Do they work in a spike?

Yes, but less than the headline number. VIX rose 20% or more on 83 days in the study. On a median day like that, VIX rose 26.2%, VIXY 13.8% and UVXY 21.6%. Futures already sit above VIX, and traders expect a spike to fade, so they move less. Yesterday was the same: VIX rose 70% and VXX 30%.

To be fair to these funds, we gave them perfect timing. For every spike, we bought UVXY at its lowest close in the 15 days before VIX reached 30, and sold at its best close. Then we counted how long it took UVXY to fall back to the price we paid.

The best one was early 2020. With perfect timing UVXY rose 945%. It then gave all of that back in 184 trading days. The median spike took 44 trading days to give back, and three of them took about a week.

So the gains are real, and they're short. Nobody gets the perfect dates in that table. You'd have to buy before the panic and sell near the top, while the news is at its worst.

What if you only buy when VIX is low?

That's the obvious fix. VIX can't stay low forever, so wait for it to be cheap and then buy. We tested it on UVXY, on every trading day since 2011.

It helps a bit. Buying with VIX at 13 or lower lost money 71% of the time after three months, against 84% on any day. The average result improved from −23.4% to −1.9%. But that's still a small loss, and it only gets that close to zero thanks to a few huge wins like early 2020. The typical result was still a loss of about a quarter.

Are they a good hedge?

ProShares pitches UVXY and VIXY as a way to reduce stock market risk, and on a day like yesterday they do. Over a whole bad year, it's less clear. In 2022, SPY lost 18.2%. VIXY lost 25.0% and UVXY lost 44.8%. Stocks fell slowly, VIX stayed under 40, and the roll kept charging.

VIXY rose in two of fourteen full years, 2018 and 2020. UVXY rose in one. Over one-year holds, VIXY lost money 90% of the time and UVXY 96%.

What about going the other way with SVXY?

If the long funds lose so reliably, the fund that bets against them should win. And it does, most of the time. SVXY rose in eight of fourteen full years and on 59% of days. This year it was up 15.6% before yesterday. One day took it to −2.2% for the year.

The risk is one day. By 11 January 2018 SVXY had turned $10,000 into $262,632. SPY had turned it into about $27,950. Then VIX more than doubled on 5 February 2018, its biggest rise since 1990. SVXY's close on 6 February 2018 was 83% lower than the day before. It lost 90% in three trading days.

A similar note, Credit Suisse's XIV, lost about 96% and was shut down. SVXY survived. At the close on 27 February 2018, ProShares cut it from −1x to −0.5x, and cut UVXY from 2x to 1.5x.

SVXY has never got back. It's still 80% below its January 2018 high. It reached a low 95% below it in March 2020.

The gentler −0.5x version has done fine since. From 28 February 2018, $10,000 became $22,146, or 9.6% a year. But SPY made 13.8% a year over the same days and turned $10,000 into $30,812. SVXY has lost to plain stocks over the full study too, and it fell when stocks fell. So it's stock market risk with a worse crash built in.

VXX vs VIXY vs UVXY: which one, if you must?

• VXX is a note, not a fund. Barclays owes you the value, so you take on the risk of Barclays. It also has a history: on 14 March 2022 Barclays stopped issuing new VXX notes and warned they could trade away from their real value. It restarted on 26 September 2022.
• VIXY is the same exposure as a fund, for a slightly lower fee of 0.85%. ProShares says it generates a K-1 tax form, which is a US form that can make tax filing harder. Tax rules vary, so check yours.
• UVXY is the same index at 1.5x. It moves more each day and decays faster. It's the one for a trade measured in days.
• SVXY is the opposite bet. It's the only one here that has made money over the long run, and the only one that has lost 83% in a day.

Outside the US

Many brokers outside the US won't sell you US-listed ETFs like these. In the UK and Europe, retail investors usually can't buy them at all. WisdomTree lists a similar product in London, the WisdomTree S&P 500 VIX Short-Term Futures 2.25x Daily Leveraged (VIXL in dollars, VILX in pence), with a 0.99% fee. At 2.25x it follows the same index with more leverage than UVXY, so everything here applies to it, only harder. Check what your broker offers and read the product's fact sheet.

The risks

• Time works against the long funds. VIXY lost money in 72% of all one-month holds and 80% of three-month holds. UVXY did worse.
• A big VIX move isn't the same size in the fund. VIXY caught about half of VIX's rise on its biggest days.
• SVXY can lose most of its value overnight. Its worst fall was mostly after the stock market had closed, so a stop-loss wouldn't have saved you.
• Reverse splits don't change your money, but they hide the record. A price chart that looks normal can sit on top of a 99.99% loss.
• Issuers can change the terms. UVXY and SVXY changed their multiples in 2018, XIV closed and VXX stopped issuing in 2022.

So do VIX ETFs always lose money?

Over days, no. Yesterday shows that. Over months and years, the long funds nearly always do, and it isn't bad luck. It's built in, because they keep buying futures that cost more than VIX.

If you trade UVXY or VXX, think in days, and plan your exit before you buy. After a spike, the clock runs fast: a median of 44 trading days, and sometimes about a week.

If you're thinking of SVXY as a steady earner, look at February 2018 first. For most people, plain SPY has paid better, with a worst fall of about a third instead of 95%.