A 10.7% baseline appears in three separate frameworks inside the same document. That is not corroboration; that is repetition. When a single decimal point is reused across a monetary-supply model, a gold-substitution model, and a network-adoption model, the correct forensic response is not to celebrate consensus but to ask who supplied the input twice.
JPMorgan Asset Management's Long-Term Capital Market Assumptions, 2027 edition, assigns Bitcoin a baseline annualized return of roughly 10%. The number circulated widely after a tweet from Bitwise's Matt Hougan on October 20, and by the time it reached retail timelines it had hardened into three price targets: $215,000 base, $386,000 bull, $25,000 bear, all compounded off a $83,000 spot price on September 30. The arithmetic is clean. The arithmetic is also the problem.
The compounding is internally consistent and externally meaningless. BeInCrypto generated those three prices, not JPMorgan. The asset manager published return assumptions. The media outlet supplied the starting price and the exponent. Multiplying one party's rate by another party's spot quote produces a number that belongs to neither, yet carries the authority of both. This is the mechanism by which false precision enters institutional crypto discourse: not through fabrication, but through unauthorized composition.
I have spent eleven years watching this pattern. During the FTX collapse I reconstructed a fragmented internal ledger against on-chain deposits and found a $2.4 billion discrepancy that no media outlet had reported, because the press was compounding narratives instead of reconciling accounts. The lesson generalizes. When you see a precisely stated terminal value, ask who computed it and under what authority. A number with four significant figures implies four significant figures of confidence. Here, only two are earned.
The more interesting artifact sits one layer down. JPMorgan's model is not a single forecast but a weighted composite of three frameworks: monetary supply at 65% weight, gold substitution at 30%, network adoption at 5%. The baseline returns attributed to each are 10.7%, 10.7%, and -3.2% respectively. Weighted:
0.65 x 10.7 + 0.30 x 10.7 + 0.05 x (-3.2) = 6.955 + 3.210 - 0.160 = 10.005
That resolves to 10.0%, matching the headline. The two dominant frameworks carry identical baselines. This is not triangulation. It is a single assumption wearing two labels.
The monetary-supply framework treats Bitcoin as a claim on a debasing fiat base. The gold-substitution framework treats it as a capture of a portion of above-ground gold's store-of-value premium. These are conceptually distinct mechanisms. If they were genuinely modeled independently, their baseline outputs would diverge at the decimal level, because the input distributions differ. Identical outputs from supposedly independent processes indicate either shared inputs upstream or calibration to a predetermined target. I have seen this before, in audit trails where two expense categories reconcile perfectly because the same receipt was entered twice.
The network-adoption framework, at 5% weight, is the only one producing a negative baseline. This is the tell. The framework that actually models Bitcoin as a technology โ as a network whose value derives from usage, transaction volume, and developer activity โ projects decline. JPMorgan assigned it the smallest weight in the composite. That allocation decision, not the 10% headline, is the substantive editorial act. A 5% weight on the only framework that models the asset as a technology is a statement about priors, not about data.
Consider what happens under reweighting. If network adoption carried 30% weight and gold substitution 5%, the baseline collapses toward 3.1%. If adoption carried the 65% weight, the composite turns negative. The model's output is dominated by whichever narrative receives the largest coefficient. The 10% figure is therefore not a prediction about Bitcoin. It is an arithmetic consequence of a portfolio-allocation decision made by an asset manager whose business includes selling exposure to the asset class it is modeling.
This is not an accusation of bad faith. It is a structural observation. Long-term capital market assumptions exist to anchor institutional allocation decisions. They are inputs to portfolio construction, not forecasts to be tested. When a bank publishes a 10% baseline for an asset it also offers products around, the number's function is to justify a position size, not to survive falsification. Readers who treat it as a prediction have misread its genre.
Now the contrarian turn, because the bulls are not entirely wrong and pretending otherwise would be its own form of dishonesty.
The strongest case for JPMorgan's framing is that monetary debasement is genuinely the dominant driver of Bitcoin's price history. Every major cycle since 2020 correlates more tightly with global liquidity expansion than with any on-chain metric. If you accept that Bitcoin is primarily a macro asset โ a high-beta expression of fiat debasement โ then the 65% weight is correct, and the two 10.7% baselines are not suspicious but expected, because both frameworks are ultimately proxies for the same underlying variable: the expansion of the monetary base. Under that reading, the network-adoption framework's negative output is accurate and its low weight is honest. Bitcoin's user growth, transaction throughput, and fee revenue have not scaled with its market capitalization. Assigning that reality 5% weight is a defensible modeling choice.
I find this argument serious. I also find it incomplete, because it concedes the bull case's weakest point while framing the concession as strength. If Bitcoin's price is 65% determined by fiat monetary policy, then Bitcoin is not an independent asset; it is a leveraged derivative of central bank behavior. That proposition should disturb holders far more than a $25,000 bear case. The bear case is a price outcome. The weighting is a structural verdict about what the asset actually is.
My own audit work on rollup bridges and autonomous agent exploits taught me to distrust composites whose components are never independently stress-tested. In 2024 I found a re-entrancy condition in a $150 million TVL bridge that permitted infinite minting under specific race conditions. The developers' initial response was that the exploit was theoretically possible but practically irrelevant. They were modeling the system as three weighted components; the attacker was modeling it as a single state machine. The composite view concealed the failure mode. Weighted frameworks smooth precisely the discontinuities that destroy systems.
Apply this to the JPMorgan model. A 10% baseline derived from a 65/30/5 composite describes a world of continuous, moderate outcomes. Crypto does not produce continuous, moderate outcomes. It produces regime shifts: December 2017, March 2020, November 2022, and whatever the AI-agent oracle manipulations of 2026 eventually resolve into. A weighted-average model calibrated on long horizons cannot price a regime shift, because regime shifts are not samples from the distribution โ they are changes to the distribution. This is not a flaw unique to JPMorgan. It is a flaw in the entire genre of long-term capital market assumptions applied to an asset with fat-tailed, path-dependent, reflexivity-driven dynamics.
The September 30 spot price of $83,000 sits in the middle of this analysis like an unexamined constant. BeInCrypto compounded from it without asking whether $83,000 represents fair value, distressed value, or a local maximum. The entire three-scenario structure inherits whatever error that starting point contains. Compound a 10% rate from a wrong base and you get a wrong terminal value with impeccable internal logic. Proof exists; it is merely waiting to be verified โ and here, the base case is the unverified premise.
The 2026 family office report referenced alongside these figures adds a fourth element to the timeline, completing a set of documents that are internally self-consistent and mutually reinforcing. That consistency is the product of a shared information environment, not independent confirmation. Four sources agreeing on a number they all sourced from the same place is one source.
The forward-looking question is not whether Bitcoin returns 10% annualized over the next decade. The question is what happens to institutional allocators when the composite framework they used to justify position sizes encounters a regime the model cannot represent. Long-term capital market assumptions are not forecasts. They are rationalization instruments, and their function is to make a position size defensible in a committee meeting. When the regime shifts, the committee will discover that the number it approved was never a prediction at all. It was a permission slip.
The algorithm remembers what the witness forgets. The model output will be cited for a decade. The weighting decision that produced it will be forgotten within a quarter. And the ledger, as always, will balance โ but the ethics of the allocation will remain uncalculated, because no line item in any of these three frameworks was ever designed to capture them.