Why aggregator headlines drift from the actual filings

Look up a company’s Bitcoin holdings on two different trackers and you’ll often get two different answers — and both may disagree with the company’s own report. That’s not necessarily anyone lying. It’s the predictable result of how aggregated figures are assembled versus how a company discloses. Once you know the handful of reasons the numbers split, you’ll know instantly which one to trust.

The most common cause is simple timing. A company files a purchase today; a tracker might not pick it up for days, or it may still be showing the figure from the last quarterly report. The aggregator isn’t wrong so much as behind.

This cuts both ways. A tracker can lag a fresh purchase and show too little, or it can carry an old announced target the company never fully hit and show too much. The fix is always the same: check the date attached to each figure, and trust the one tied to the most recent filing.

Press releases and trackers love a round number. “Over 50,000 Bitcoin” makes a cleaner headline than 49,447. Aggregators often inherit that rounding, or apply their own, so a tracker total can be a tidy approximation rather than the exact figure on the books.

For a quick sense of scale, rounding is harmless. For anything that gets cited — a per-share calculation, a premium, a ranking — the exact filed count is what you need, because small differences compound once you start dividing by share counts or comparing two companies head to head.

This is the drift that does the most damage. A company announces it intends to hold a large amount of crypto — often through a merger or a planned raise — and a tracker books the full figure immediately, as if the coin were already on the balance sheet. Until the deal actually closes and the asset shows up in a filing, that holding is a plan, not a position.

• An announcement is not a holding. A company can announce a target and never reach it, or close at a different size, or not close at all.
• A merger is not done until it’s done.A blank-check company that says it will merge with a crypto holder doesn’t hold that crypto yet — its filings still show the cash it’s sitting on until the merger consummates.

The subtler errors come from the structure of who owns what. A holding can get counted twice — once for a parent company and again for a subsidiary that’s already inside it — or a coin pile a company manages on behalf of others can get mixed in with what it owns for itself. An aggregator stitching together dozens of sources has many seams where a figure can be duplicated or double-booked.

These are exactly the errors that don’t announce themselves: the total simply looks a little too high, with no obvious reason why. The only reliable defence is to rebuild the holding from the underlying filings rather than trusting the assembled total.

When the figures disagree, the order of trust is clear: the company’s own most recent filing first, its investor-relations page next, and an aggregator last. A filing is dated, exact, and legally binding; an aggregated total is a convenience that’s only as fresh and as careful as its weakest source.

That’s the whole reason we reconcile every holding back to its primary source — a filing, a fund sponsor’s report, or an on-chain record — and label anything that’s only an aggregate estimate as exactly that, instead of letting it pose as a confirmed position. For the full method, see how we verify holdings.