Search Everything in One Place

Explore the web, images, videos, news, and more – all in one place.

Finance

These 15 funds cost investors billions over the past decade

These 15 Funds Cost Investors Billions Over the Past Decade
These 15 Funds Cost Investors Billions Over the Past Decade

If you want to know how a fund has performed, total returns are the first place to look. But performance is less meaningful if few shareholders are around to benefit from it. To get a better sense of which funds have created or destroyed the most value in dollar terms, value creation is a better measure. To home in on this, we ranked Morningstar’s ...

If you want to know how a fund has performed, total returns are the first place to look. But performance is less meaningful if few shareholders are around to benefit from it.

To get a better sense of which funds have created or destroyed the most value in dollar terms, value creation is a better measure. To home in on this, we ranked Morningstar’s database of US-based mutual funds and exchange-traded funds, focusing on those that had the biggest increase in asset size over the 10 years ended in 2025 after subtracting total inflows and outflows over the same period. The resulting number reflects how much growth a fund has created (or destroyed) from market appreciation in dollars.

In previous studies, we’ve looked at the biggest “value creators” in the fund industry, which mainly consist of large, mainstream funds from fund families such as Vanguard, Fidelity, and American Funds.

Today, we’re looking at the same issue from the opposite perspective, focusing on funds that have lost value for shareholders over the past 10 years. (For further details, see the “Methodology” discussion at the bottom of this article.)

Two important notes: This study doesn’t include bond ETFs or ETFs in the energy limited-partnership Morningstar Category because their income payouts can complicate how we estimate funds’ cumulative losses. It also focuses on results for long-only shareholders invested directly in each fund; it doesn’t capture changes in wealth for investors who might have invested in derivatives tied to these funds, such as options on ETFs, or through short sales of ETFs. In other words, we focus on the assets that these funds charged fees on.

The Results: Speculation Comes at a High Cost

The wealth destroyers are mainly a motley crew of more-specialized fund categories.

Seven of the top 15 funds on our list are in the trading-inverse equity category, which are designed to increase in value when a specific benchmark or asset type falls. These funds occasionally rise to the top by betting against the market; for example, many funds that bet against US stocks were up 15% or more in 2022. But because market returns are positive more often than not, the long-term results haven’t been pretty.

And while shareholders could theoretically use these vehicles for short-term trading purposes, the fund flows data we examined suggests that doesn’t usually work. ProShares UltraProShort QQQ SQQQ, for example, aims to deliver 3 times the opposite return of the Nasdaq 100 Index and generated an estimated $12.5 billion in total shareholder losses. The fund experienced net outflows in years when its benchmark was down (2018 and 2022) and inflows every other year, when the index surged ahead. Such funds are also subject to volatility decay, a gradual loss of value that stems from the basic math of daily compounding, especially in volatile markets.

However, leveraged bets on upside returns also led to poor results for two of the other funds on our list, Direxion Daily 20+ Yr Treasury Bull 3X ETF TMF and ProShares Ultra Bloomberg Natural Gas BOIL. The two funds suffered sharp losses thanks to their leveraged bets on long-duration Treasuries and natural gas, respectively.

Even without the toxic effects of leverage, KraneShares CSI China Internet ETF KWEB, Fidelity Series Long-Term Treasury Bond Index FTLTX, and iShares Ethereum Trust ETHA generated significant losses during the period covered in our study. An extremely narrow focus on internet-related companies based in China was especially damaging for the KraneShares fund, which racked up an estimated $5.2 billion in shareholder losses. All of these funds incurred large losses in dollar terms owing to a combination of poor performance and poorly timed fund flows.

ARK Innovation ARKK ranked fourth on the list with an estimated $5.0 billion in shareholder losses over the 10-year period (down from an estimated $7.0 billion in last year’s study). The fund posted a 35.7% total return in 2019, followed by a 152.5% runup in 2020. It garnered huge asset flows in 2020 and 2021 (totaling an estimated $14.1 billion), but returns were decimated in the 2022 bear market, when it lost 67.0% for the year. The fund recouped some of its losses after it ranked in the top 5% of Morningstar’s mid-cap growth category in both 2023 and 2025; it also earned a positive total return over the 10-year period ended in 2025. Even so, many long-term shareholders are still underwater in dollar terms because they bought in after performance had already peaked.

ARK Genomic Revolution ARKG, which ranks fifth on the list with an estimated $4.5 billion in shareholder losses, had a similar experience. Assets surged in the wake of its eye-popping performance in 2019 and 2020 (when it posted cumulative returns of more than 300%), but the fund then went on to lose about 34% in 2021 and another 54% in 2022.

The Takeaway for Investors

The biggest value destroyers in the fund industry illustrate that there’s no guarantee of success, even during a generally favorable market environment. They also provide a valuable case study in how not to invest. (As Charlie Munger was fond of saying: “Invert, always invert.”) Volatile and speculative strategies often lead to poor results, and leveraged and inverse trading vehicles can be particularly disastrous. As we’ve found in other studies, investors have been far better served by dull but mainstream funds that make straightforward investments in major areas of the market.

Methodology

Value destruction over the 10-year period reflects the sum of annual depreciation in dollars for each of the trailing 10 calendar years starting in 2016. We estimate depreciation by subtracting a fund’s cumulative flows over the 10-year period from the difference between its ending and beginning assets. With respect to ETFs, the calculation does not distinguish between flows that reflect the normal netting of supply and demand and those that stem from “create-to-lend” transactions, whereby a market maker creates shares in order to facilitate a short sale of the ETF. In these cases, there is still a long investor—that is, the buyer of the ETF shares that the short seller has borrowed and sold short.

Our approach to estimating capital depreciation is consistent with how funds account for capital gains and losses in the financial statements they file with the Securities and Exchange Commission. We compared our estimates with what the fund companies concerned have reported and found that our estimates approximate their reported figures.

For example, here is a link to the annual reports ARK filed over the 10-year period ended Dec. 31, 2025, for ARK Innovation. The relevant figure in each annual report is the ETF’s “Net increase (decrease) in net assets resulting from operations,” which is reported in its “Statement of Operations” as well as its “Statement of Changes in Net Assets.” We summed these figures from each annual report to corroborate our estimates.

Read full story on Morningstar

Related News

More stories you might be interested in.

OpenAI says AI models went rogue during testing, triggering 'unprecedented' breach at startup
Reuters·8 hours ago

OpenAI says AI models went rogue during testing, triggering 'unprecedented' breach at startup

By Raphael Satter WASHINGTON, July 21 (Reuters) - OpenAI said on Tuesday that an autonomous agent powered by its advanced AI models went rogue during a security test and triggered a hack that compromised the infrastructure of AI startup Hugging Face last week. In a blog post, OpenAI said it was testing the capabilities of some of its most advanced models in a controlled environment but that the

Top