The Chip Fund That Turned $10,000 Into $364,000 Just Bounced 59% Off the Worst Day in Its History. Here’s What History Says Comes Next

The Direxion Daily Semiconductor Bull 3X ETF (NYSEARCA: SOXL) turned $10,000 into more than $364,000 in just over 14 months. Then, on November 5, it fell 38.7%, the worst day since it launched in March 2010. It was the fund’s 7th one-day drop of 25% or more, and history says what comes next is either a fortune or a wipeout. After five of the six earlier drops, the fund was up between 228% and 678% a year later. After the sixth, this June, it is still down 64%.

The $10,000 that grew to $364,570 at the June 22 record close was worth $49,539 at the close on November 5. In pre-market trading Wednesday at 7:40 AM ET, the fund was at $65.14, up 16.2% from Tuesday’s close and 59.4% above its November 5 close. That puts the $10,000 at $78,958.

The fund is simple to describe and hard to hold. It promises three times the daily move of an index of the 30 largest U.S.-listed chip companies, before a 0.75% annual fee, and it resets that promise every single day. On November 5, when the AI selloff took the chip index down 12.1% and the Nasdaq-100 down 6.5%, the fund delivered three times the damage, on 549 million shares traded, the most since April 2025.

The history reads as an invitation to buy the dip, and that reading has a limit. Five of the six earlier drops came inside a crash that ended within weeks, in March 2020 and April 2025. The sixth came on June 5 of this year, and the index kept falling for five more months, so the fund kept tripling the losses on the way down. Nobody knows which kind of drop November 5 was, and if you own the fund, or your grandson wants you to, you do not need to know. What you can know is the arithmetic, so here it is, worked for a 68-year-old whose grandson talked him into the fund in June.

What the fund did

From the April 8, 2025 low of $8.25 to the June 22, 2026 record close of $300.77, the fund rose 36-fold while the chip index rose 323%. That is the good side of daily leverage, the side you see on the chart your grandson shows you: in a steady rally each day’s gain compounds on the last, and the fund delivers far more than three times the index.

Then the index turned. From June 22 to November 5 the chip index fell 39.6%. The fund fell 86.4%, about 2.2 times the index loss. On November 5 alone it lost 38.7%, beating its March 16, 2020 record of 38.6% by a tenth of a point. It gained 37.2% in the three sessions after that, including 22.3% on November 9, and at Tuesday’s close of $56.06 it was still up 33.4% for the year. Wednesday’s pre-market price of $65.14 put it 55.0% higher for the year and still 78.3% below the June record. A fund can be up 55.0% on the year and down 78.3% from its high at the same time. That is what three times leverage looks like on a chart.

What history says after a day like this

The fund has closed 25% or more below the previous day seven times since 2010, including November 5, so a drop this size arrives about once every two years and a half, and then usually in clusters. Here is what followed each of the first six, from the closing price on the day of the drop, adjusted for the fund’s two share splits:

Five for five with a full year of data, the smallest gain 228%. That is the pattern the dip buyers are trading on, and it is real. The sixth row is the warning. The June 5 drop looked like the others for a month, up 7%, and then the index rolled over. Three months later the fund was down 32% from the day of the drop. By November 10 it was down 69%. History never promised that the index would recover, only that the fund would triple whatever the index did each day.

Why three times the leverage became more than twice the loss

The fund’s prospectus says that returns over periods longer than one day “should not be expected to be 300%” of the index, and that the fund “is not intended to be used by, and is not appropriate for, investors who do not intend to actively monitor” their positions. The reason is arithmetic. Suppose the chip index falls 10% on Monday and rises 11.1% on Tuesday, which puts it exactly back to even. A fund delivering three times each day’s move falls 30% on Monday, to 70 cents on the dollar, then rises 33.3% on Tuesday, to 93.3 cents. The index is flat. The fund is down 6.7%. Every bounce starts from a smaller base, so a choppy market grinds the fund down even when the index goes nowhere, and you pay the 0.75% fee the whole way.

The past week is a live example. From the November 4 close to the November 10 close the chip index lost 2.4%. Three times that is 7.1%. The fund lost 15.9%, because it fell 38.7% first and then had to climb back from a much smaller base. Stretch that over five months and you get the gap between a 39.6% index loss and an 86.4% fund loss. At Tuesday’s close, the chip index needed to rise about 49% to regain its June record. The fund needed about 437%. Even after Wednesday’s pre-market jump, it still needs about 362%.

The retiree whose grandson picked it

Consider a scenario: a 68-year-old retired electrical engineer in Raleigh with a $900,000 portfolio, most of it in an IRA, and $2,900 a month in Social Security. In June his 24-year-old grandson, who works in software and trades on his phone, showed him the chart. He bought 66 shares at the June 22 record close of $300.77, for $19,851, inside the IRA. That was 2.2% of his money.

At Wednesday’s pre-market price of $65.14, his 66 shares were worth $4,299, still 78.3% below what he paid. The worst day in the fund’s history cost him $1,703, or 0.19% of his portfolio. By Tuesday’s close, the slide from June had cost him $16,151, or 1.8% of his money. The other $880,149 never noticed. Because the shares sit in an IRA, the loss gives him no tax deduction; in a taxable account it would offset $16,151 of gains, or $3,000 a year of ordinary income. That is the one thing he got wrong in the paperwork, and the one you can fix today: a bet this volatile belongs in a taxable account, where a loss is worth something.

Now change one number. Had he put 10% of the portfolio in, $90,000, the 299 shares would be worth $16,775 today, a loss of $73,225, or 8.1% of everything he has, and November 5 alone would have taken $7,723 in one session. Same fund, same grandson, same dates. The size of the position decided whether this was a story he tells at Thanksgiving or a story that changes his retirement.

The dip-buy math

His grandson now wants him to buy the dip, with another $20,000 at Wednesday’s pre-market price of $65.14. That buys 307 shares, for 373 in all, at an average cost of $106.83, and it turns a 2.2% position into a 4.4% one. To get the whole $39,849 back to even, the fund needs to rise 64% from here. If the next year looks like the year after March 16, 2020, the 373 shares would be worth about $168,935. If the next five months look like the five after June 5, they would be worth about $7,462. A repeat of November 5 on the enlarged $24,297 stake would erase $9,405 in a single session, 1.0% of his portfolio.

That spread, $7,462 to $168,935 on the same $39,849, is the honest answer if you ask what history says. The range is enormous in both directions, and the fund amplifies whichever way the index goes next, which nobody knows. Position size is the only input you control. If you hold 1% of a $900,000 portfolio in the fund, a November 5 repeat costs you $3,484. At 5% it costs $17,418, and a repeat of the June-to-November slide costs $38,885, more than a year of his Social Security.

The bottom line

The fund did exactly what its prospectus says: three times the index, every day, with no memory of yesterday. That made $10,000 into more than $364,000 in 14 months and then took 86% of it back in five. History says five of six comparable drops ended in gains of 228% or more within a year, and the sixth is still down 64%. You cannot know which this one is. You can decide how much of your retirement rides on it, and for a 68-year-old with $900,000 the difference between 2% and 10% was the difference between a $16,151 lesson and a $73,225 wound. Keep the leveraged bet at a size where its worst day ever is a rounding error, hold it in an account where a loss is at least deductible, and let the grandson keep the chart.

Contact [email protected] for any questions or corrections.