Here's a fun question: what happens when the entire market is valuing companies using the wrong number of shares? According to JPMorgan, you get an $8 billion accounting gap that makes Bitcoin miners look far cheaper than they actually are.
The problem starts with basic arithmetic. JPMorgan analyst Reginald L. Smith flagged the issue in a November 24 research note, pointing out that Bloomberg's diluted share counts for Cipher Mining Inc. (CIFR), CleanSpark Inc. (CLSK), Riot Platforms Inc. (RIOT), and MARA Holdings Inc. (MARA) are anywhere from 20% to 33% too low. That's not a rounding error. That's the kind of gap that breaks every valuation screen on Wall Street.
The Dilution Nobody Updated
Bitcoin miners have been issuing stock like it's going out of style. They're funding land acquisitions, building out high-performance computing infrastructure, and converting facilities into long-term data centers. To pay for all of it, they've been tapping ATM programs, floating convertible bonds, and handing out options and RSUs.
JPMorgan's math shows that Cipher and CleanSpark's diluted share counts have climbed about 20% since the firm's last model update. Riot and Marathon? More than 30%. But Bloomberg's data feeds haven't caught up. If those lighter share counts are still the default reference for market cap and enterprise value calculations, investors are effectively valuing a company that no longer exists.
Wrong Inputs, Wrong Conclusions
This isn't just a data hygiene issue. The mispricing lands at a critical moment for the sector, which is splitting into two distinct camps. Cipher and CleanSpark are getting upgraded on the strength of multi-year HPC contracts and data center capacity that's becoming more valuable than Bitcoin mining itself. IREN Ltd. (IREN) is pulling in a richer cloud-style valuation.
Meanwhile, Riot and Marathon—the sector's two biggest Bitcoin holders—are seeing price targets trimmed as weaker Bitcoin (BTC) economics collide with ballooning share counts. With this much change happening at once, an $8 billion gap at the foundational data level isn't just a footnote. It changes the entire narrative around which companies are gaining ground and which are simply inflating their equity base.
Why the Numbers Actually Matter
Smith's point is simple: if your share count is wrong, every ratio built on top of it is wrong too. EV per megawatt? Wrong. Valuation comps? Wrong. That "cheap versus expensive" call your screen is telling you? Also wrong.
For a sector in the middle of a once-in-a-decade shift from pure Bitcoin mining toward AI-driven HPC revenue, accurate data isn't optional. It's the difference between spotting genuine value and chasing something that only looks like a bargain because the denominator is off by 30%.
