Nine years ago, Jamie Dimon called bitcoin a fraud. According to Journal du Coin, his bank's asset management division has just assigned it an expected return of 10% a year over 10 to 15 years, more than Wall Street's large caps and more than gold. On the same day, US bitcoin ETFs recorded their best day of inflows in nearly a month. Two institutional signals that deserve a close reading, because neither of them says what bitcoin will actually return.

Bitcoin enters JPMorgan's long-term assumptions

Every autumn, J.P. Morgan Asset Management publishes its Long-Term Capital Market Assumptions, a document that puts a figure on the expected return of more than 200 asset classes over 10 to 15 years. Pension funds, insurers and asset managers use it to build their strategic allocations. The 31st edition came out on October 19, 2026, with data as of September 30, Journal du Coin reports.

Bitcoin appears in it for the first time. "For the first time, we are setting a return range for bitcoin, even though we are not yet including it in our main covariance matrices," the bank writes in the presentation of its study, quoted by the site.

The figures, as relayed by Journal du Coin: 10% a year in compound return for bitcoin, 7.5% for US large-cap equities (a forecast raised by 0.8 percentage points, thanks to margins above 15%), 8.2% for US listed real estate, 5.5% for gold and 7% for a classic portfolio made up of 60% equities and 40% bonds. The site translates these rates into multipliers: at 10% a year, an investment would be multiplied by 2.59 in ten years and by 4.18 in fifteen years, compared with ×2.06 over ten years for US large caps and ×1.71 for gold.

A manager's assumption, not a promise of returns

A capital market assumption is a portfolio construction tool: it is used to compare asset classes with one another over a long horizon, not to forecast the price in any given year. And the 10% figure cannot be read without its counterpart. JPMorgan assumes annual volatility of 55% for bitcoin, compared with around 18% for US large caps and 17.2% for gold, according to Journal du Coin. At that level, the site points out, seeing an investment lose half its value in twelve months is nothing exceptional.

This turbulence explains the most telling gap in the table: bitcoin's arithmetic mean return reaches 20.9% a year, but falls to 10% once compounded. More than ten points vanish through the effect of price swings, and it is the compound return that measures what an investor actually keeps.

The bank also remains cautious about how this figure is used. Bitcoin only enters a small separate matrix, alongside a few listed assets, and no specific allocation is proposed; the cryptocurrency stays out of the main portfolio models. According to crypto.news, quoted by Journal du Coin, the report also warns against the idea that distrust of the dollar would automatically push investors towards cryptocurrencies.

Interpretations are already diverging. Matt Hougan, chief investment officer of Bitwise, shared the table on X: in his view, these figures make bitcoin the most profitable asset of the coming decade, and JPMorgan is even too cautious, with overestimated volatility and underestimated returns. This is a statement from an issuer of crypto products, not a data point. Journal du Coin, for its part, notes that Grayscale measures realised volatility of 43% over the past twelve months, far from the 55% used by the bank.

The market, for its part, did not follow: bitcoin was down 1.5% over 24 hours and trading at around $81,340 on Tuesday morning, according to CoinGecko as cited by the site. On the private banking side, JPMorgan has allowed its clients to buy bitcoin since May 2025, without providing custody.

Bitcoin ETFs: inflows driven by a single fund

On Monday, October 19, US spot bitcoin ETFs recorded $526.6 million in net inflows, according to Sosovalue data reported by Journal du Coin. That is their best day since the $714.7 million of September 22, after weeks in which we were mostly following bitcoin ETFs seeing outflows.

The details qualify the rebound. IBIT, BlackRock's fund, took in $665.2 million on its own, its biggest daily inflow since October 7, 2025. These subscriptions offset withdrawals at two competitors: $83.6 million for Fidelity's FBTC and $54.9 million for Grayscale's GBTC. The other funds show zero net flows.

The session wipes out the roughly $228.5 million of outflows from the previous week, but is not enough to turn October positive: cumulative withdrawals for the month still amount to around $90.5 million. Journal du Coin takes care to point out that these data show sustained demand for BlackRock's product, without signalling a return of subscriptions across all issuers, nor proving a transfer of clients from Fidelity or Grayscale to IBIT. Daily flows describe what goes into and out of each fund, not the path taken by investors.

Ethereum: a fourth session of withdrawals

Spot Ethereum ETFs, for their part, lost $103.3 million on the same Monday, including $95.4 million for BlackRock's ETHA, $6 million for Fidelity's FETH and $1.9 million for Grayscale's ETHE. The session extends the withdrawals of October 14, 15 and 16. Over the last five trading days, including the inflows of October 13, the balance comes to around $338 million in net outflows.

Since the start of October, outflows are approaching 900 million to one billion dollars, according to estimates cited by Journal du Coin. The segment nevertheless retains around $14.8 billion in net assets and nearly $12.8 billion in cumulative net subscriptions since its launch in July 2024. Here again, the opposite movements of IBIT and ETHA do not prove that investors are selling Ethereum to buy bitcoin.

What these two signals say, and what they don't

Our reading: JPMorgan's publication matters less for its 10% figure than for what it normalises. That one of the world's major asset managers gives bitcoin a line in a reference document for pension funds and insurers is something new. But the bank pairs it with 55% volatility, keeps it out of its main models and recommends no allocation. It is a working assumption, revised every year, not a commitment.

Monday's flows call for the same caution. One strong day, concentrated on a single product, in a month that is still negative, does not make a trend. To measure how far things have come, you can reread the assessment drawn up one year after the October 10 crash. The rest of October will show whether IBIT remains an isolated case or whether the other issuers win back subscriptions.

Sources

  • Journal du Coin: « Bitcoin : JPMorgan lui prédit 10 % par an, mieux que les actions américaines et l'or »
  • Journal du Coin: « ETF Bitcoin : Collecte record depuis près d'un mois, Ethereum reste dans le rouge »