MKTMarketsExplainer NO. ---

JPMorgan bitcoin 10% return assumption tops stocks, bonds and gold

J.P. Morgan Asset Management has put a range on bitcoin returns for the first time, with a 10% annual base case over the next decade. Here is how the estimate was built, how it compares with stocks, bonds and gold, and why its authors remain cautious.

By Rowan Ashby Published 21.10.26 12:57 UTC 8 min read

JPMorgan bitcoin 10% return assumption tops stocks, bonds and gold — Unknown /

Key takeaways

J.P. Morgan Asset Management's 2027 Long-Term Capital Market Assumptions (LTCMA), released on October 19-20, 2026, put a range on bitcoin returns for the first time, with a base case of about 10% a year over a 10- to 15-year horizon.
That base case sits above the firm's figures for U.S. large-cap stocks (7.5%), global aggregate bonds (5.3%) and gold (5.5%), according to the firm's release and media coverage of the report.
Media reports differ on the scenario range: CoinDesk cites a bull case of 17% and a bear case of -9.7%, while Unchained cites 16.6% and -11.2%. JPMorgan's public summary pages do not show the scenario figures.
The report's authors assume annualized volatility of about 55% and keep bitcoin out of the firm's main covariance matrices, which feed its portfolio models.
The authors describe bitcoin as a speculative asset and a hedge against monetary easing, and say they do not make a case for it as a portfolio diversifier.

What happened

J.P. Morgan Asset Management's 2027 Long-Term Capital Market Assumptions give bitcoin a base case of about 10% a year over a 10- to 15-year horizon. That JPMorgan bitcoin 10% return assumption sits above its figures for U.S. large-cap stocks (7.5%) and gold (5.5%). For a shorter-term pattern, see how the 2026 low compares with past bitcoin midterm cycles.

The bitcoin figures spread quickly once crypto media read the full report. CoinDesk reported on October 20 that the base-case assumption is a 10% annualized return over the next decade, a number that tops the firm's projections for stocks, bonds and gold. Unchained, in a report republished by Yahoo Finance on October 21, gave the same base case and added details on the method behind it.

The headline comparison is striking because it comes from a bank whose chief executive, Jamie Dimon, called bitcoin a "fraud" in 2017 and has remained publicly skeptical since, as CoinDesk noted. The report itself, however, is far more cautious than the headline suggests.

What long-term capital market assumptions are

Capital market assumptions are long-horizon estimates of the return, volatility and correlation of major asset classes. Large asset managers publish them so that pension funds, endowments, insurers and financial advisers have a consistent starting point for strategic asset allocation: deciding, for example, what share of a portfolio should sit in equities, bonds or alternatives over many years.

J.P. Morgan's version covers more than 200 asset classes in more than 20 base currencies over a 10- to 15-year horizon, and draws on input from more than 90 portfolio managers, economists and analysts, according to the firm. Several features matter when reading any number in the report:

They are passive returns. The figures describe what a broad market exposure might earn, not what an active manager might add or lose.
They exclude costs. The firm says the assumptions do not account for fees, expenses, taxes, trading costs or liquidity constraints.
They are not tactical. The LTCMA page states the assumptions are not designed for short-term decisions. A decade-long estimate says little about the next month or year.
They are not guarantees. The firm describes the figures as illustrative and scenario-dependent.

In other words, an LTCMA figure is a modelling input, not a forecast of where any price will be on a given date.

How the bitcoin estimate was built

The bitcoin section was written by Sheena Shah, a portfolio manager, and Dennis Cristallo, head of digital asset strategy at J.P. Morgan Global Private Bank, according to Unchained and CoinDesk. Unlike stocks or bonds, bitcoin has no earnings, dividends or coupons, so the usual building blocks (yield, growth and valuation change) do not apply directly. The authors instead blended three separate models and weighted each by their confidence in it, Unchained reported:

Monetary model (65% weight): compares bitcoin's roughly four-year supply cycle, tied to the network's scheduled issuance cuts, with growth in U.S. dollar and global money supply.
Gold-complement model (30% weight): assumes some investors treat bitcoin as a complement to gold, a "digital gold" framing.
Network-usage model (5% weight): based on activity on the Bitcoin network. Unchained reported that this is the only one of the three models with a negative base case, at -3.2%.

Unchained also reported that the inputs start from a spot price of about $83,000. The weighting itself tells a story: nearly two-thirds of the estimate rests on the idea that bitcoin behaves like a bet on monetary expansion, and only a small slice rests on its use as a payment or settlement network.

Because bitcoin's relationship with other assets has been unstable, JPMorgan has not yet placed it in the main covariance matrices that drive its portfolio models, the firm's LTCMA page says. That means the bitcoin range sits alongside the core assumptions rather than inside the framework the firm uses to build model portfolios.

How bitcoin compares with other asset classes

The table below sets the bitcoin base case beside selected figures from the same report. All are annualized U.S. dollar assumptions over 10 to 15 years.

Asset class	2027 LTCMA assumption (annualized)	Source
Bitcoin, base case	10%	CoinDesk; Unchained
Bitcoin, bull / bear case	17% / -9.7% (CoinDesk) or 16.6% / -11.2% (Unchained)	Media reports; not on JPMorgan public pages
U.S. small-cap equities	7.7%	CoinDesk; Unchained
U.S. large-cap equities	7.5% (up from 6.7%)	J.P. Morgan AM release
Global equities	7.5%	J.P. Morgan AM release
60/40 stock-bond portfolio (USD)	7.0%	J.P. Morgan AM release
U.S. high-yield credit	6.7%	J.P. Morgan AM release
U.S. investment-grade credit	6.1%	J.P. Morgan AM release
Gold	5.5% (unchanged)	J.P. Morgan AM release
Global aggregate bonds	5.3%	J.P. Morgan AM release
U.S. intermediate Treasuries	4.6%	J.P. Morgan AM release

Report data as of September 30, 2026, according to Unchained; the firm's LTCMA page cites data as of September 2026. For context, bitcoin traded at about $83,435, with a market value of roughly $1.68 trillion, as of 12:54 UTC on October 21, 2026, according to CoinGecko.

On raw return alone, the bitcoin base case is the highest figure in the table. But a return assumption cannot be read in isolation. The firm's own approach to portfolio construction weighs expected return against expected volatility and correlation, and on those measures bitcoin stands far apart from the other assets listed.

The caveats: volatility, correlation and method

Volatility of about 55%

The authors assume annualized volatility of roughly 55% for bitcoin, according to Unchained. That is several times the volatility typically assumed for broad equity indices. High volatility has a mechanical effect on long-run compounding: large losses require even larger gains to recover, so the compound (geometric) return an investor actually experiences is lower than the simple average of yearly returns. The wider the swings, the bigger that gap. A 10% compound assumption paired with 55% volatility implies a path with very large drawdowns along the way.

Unstable correlations

The authors wrote that they "do not make a case for bitcoin (BTC) as a diversifier in portfolios due to its unstable market correlations," as quoted by Unchained. CoinDesk reported that they see bitcoin mostly trading like a high-beta risk asset that tracks global money supply growth, while at times being treated as digital gold. That shifting behaviour is the reason the firm keeps bitcoin out of its main covariance matrices.

A wide and disputed range

The two published versions of the scenario range span from a bear case near -10% or -11% a year to a bull case near 17% a year. Either way, the spread between outcomes is far wider than for any traditional asset in the report. Bullquill could not confirm which pair of figures appears in the full report, because JPMorgan's public summary pages do not reproduce them. The difference may reflect rounding, a revised draft or a different measure, but that is not established.

The authors' own doubts

Perhaps the most notable caveat comes from the authors themselves. Unchained quoted them writing that "it is premature to create a 10-year forecast for bitcoin" and that bitcoin "lacks a strategic role in portfolios" beyond bets on fiat currency expansion or future trading of blockchain-based assets. They also said portfolios with any bitcoin exposure should be stress tested against both bull and bear cases, and that bitcoin's usefulness in portfolios may improve if correlations stabilise and volatility falls.

Model risk

The estimate leans heavily on one relationship: bitcoin's supply schedule versus money supply growth. If that link weakens, or if regulation, technology or market structure change the way bitcoin trades, the 65% weight on the monetary model could prove misplaced. Ten-year assumptions for traditional assets have long histories to test against; bitcoin's trading history is shorter than a typical LTCMA horizon is long.

Why the report is getting attention

Three things explain the reach of the story. First, JPMorgan is one of the largest asset managers in the world, and its LTCMA figures are widely used by advisers and institutions as a reference point. Second, the inclusion of bitcoin at all signals that client interest is now too large to ignore: Unchained reported that the report puts the crypto market above $2.5 trillion and that about a tenth of clients surveyed for J.P. Morgan Private Bank's 2026 Global Family Office Report held crypto. Third, the contrast with the bank's past public scepticism makes for an easy headline.

What has not changed is just as important. The firm has not added bitcoin to its core portfolio models, has not presented it as a diversifier, and frames the range as a modelling exercise rather than a recommendation.

What it means

The 2027 LTCMA marks a shift in how a major traditional asset manager treats bitcoin: from an asset it largely declined to model to one with a formal, if heavily qualified, long-run return range. For readers, the useful lesson is about how to read such numbers. A base-case return is only one point in a wide distribution; it excludes costs and taxes; it assumes a horizon of a decade or more; and in bitcoin's case it comes with volatility of around 55% and correlations the authors themselves call unstable. The headline that bitcoin "tops stocks, bonds and gold" is accurate on the single measure of expected return, but it leaves out the risk side of the comparison that the report's authors put at the centre of their analysis.

Open questions remain: which scenario figures appear in the final report, whether bitcoin will eventually enter the firm's main covariance matrices, and how the estimate will change in future editions as more data accumulates. J.P. Morgan Asset Management has scheduled a webinar on the 2027 assumptions for October 22, 2026, which may add detail.

Frequently asked questions

What are long-term capital market assumptions?

They are long-horizon estimates of the return, volatility and correlation of major asset classes. J.P. Morgan Asset Management's version covers more than 200 assets over 10 to 15 years and is used as an input for strategic asset allocation, not for short-term decisions.

What return does JPMorgan assume for bitcoin?

Media coverage of the 2027 report says the base case is about 10% a year over the next decade. Reported bull and bear cases differ by outlet: 17% and -9.7% according to CoinDesk, and 16.6% and -11.2% according to Unchained.

How does that compare with stocks, bonds and gold?

The same report assumes 7.5% a year for U.S. large-cap stocks, 7.0% for a U.S. dollar 60/40 portfolio, 5.3% for global aggregate bonds and 5.5% for gold, according to the firm's press release.

Does JPMorgan include bitcoin in its portfolio models?

No. The firm says bitcoin is not yet included in its main covariance matrices, and the report's authors say they do not make a case for bitcoin as a diversifier because of its unstable correlations.

Why is the estimate considered uncertain?

The authors assume volatility of about 55% a year, rely mainly on a model linking bitcoin's supply cycle to money supply growth, and wrote that it is premature to create a 10-year forecast for bitcoin.