Community Trust ScoreVerified
Charles Edwards has a blunt take on the on-chain analytics industry. The founder of Capriole Investments says up to 99% of on-chain metrics are basically noise — worthless unless investors know which signals actually matter and, more importantly, how those signals get built.
It’s a striking claim, but the timing makes sense. Institutional players are piling into crypto data now in a way retail traders never did, and they’re not taking anything at face value. Julio Moreno, Head of Research at CryptoQuant, put it plainly: professional trading desks cross-verify on-chain data against the traditional data they already trust. That’s a different posture entirely from the early days, when a slick dashboard was enough to impress.
On-chain analytics started as a retail edge. Not anymore.
Institutions Won’t Take a Dashboard’s Word for It
Banks, funds, and asset managers are now demanding the same on-chain insights that once gave smaller traders an information advantage — but they’re applying a level of scrutiny those traders rarely bothered with. During a YouTube podcast, analysts laid out the shift clearly: major industry players won’t accept data just because it shows up on a screen. Moreno went further, saying institutions now actively verify on-chain data against established traditional metrics before trusting it.
That’s a meaningful change for data providers. For years, the business model was pretty much: build a dashboard, label your metrics, and watch the subscriptions roll in. Now the buyers are asking harder questions. Where does the raw data come from? How are gaps handled? What normalization choices got made along the way? Few platforms, per Edwards, bother to explain any of that clearly.
And the lack of transparency has real consequences.
FTX’s Bitcoin Reserves and the Address Grouping Problem
The FTX collapse in November 2022 turned into an accidental stress test for on-chain data providers. CryptoQuant’s numbers showed something alarming: FTX’s Bitcoin reserves dropped from 20,177 BTC to just 0.64 BTC across two days in early November 2022. A near-total wipeout, visible in the data.
But other dashboards told a different story during the same window. Some showed conflicting figures, not because the blockchain lied, but because platforms use different address grouping strategies. CryptoQuant took a conservative approach in clustering addresses. Other providers used more aggressive methods. Same underlying blockchain. Wildly different pictures.
That’s the core problem Edwards keeps coming back to. Metrics can carry the same name across platforms and still be calculated in completely different ways. The discrepancies come from data sources, normalization choices, and how each platform decides to fill gaps — and almost nobody explains the methodology out loud.
For an institution trying to make a risk call, that ambiguity isn’t acceptable.
60 Metrics, Curated — Not 600 Thrown at a Wall
Edwards isn’t just criticizing the industry. Capriole Investments built its own answer to the problem. The firm’s Macro Index pulls together more than 60 on-chain, macro, and equities metrics — a deliberately curated set, not a firehose. The point is that more numbers don’t equal more insight. Throwing every available data point into a model yields minimal benefit if you can’t separate signal from static.
That philosophy is probably the most useful takeaway for anyone evaluating data providers right now. Understanding the methodology behind a metric matters more than the metric itself. It’s a harder skill to develop than just reading a chart, but it’s the one that actually holds up when markets get chaotic.
The FTX situation made that lesson unavoidable. CryptoQuant’s conservative address clustering gave a clearer, faster picture of what was happening to those reserves. Platforms with looser methods lagged or missed it. In a crisis, the difference between those two outcomes is enormous.
Institutions know that now. The demand for methodological transparency is shaping how data providers get evaluated — and probably how they’ll compete going forward. Firms that can show their work, clearly and completely, are going to have an edge over those that just surface numbers without context.
Edwards’ 99% noise claim sounds extreme. But when you look at how differently the same metric gets calculated across the industry, and how badly some of those differences showed up during FTX, it’s hard to argue the number is too high. CryptoQuant’s Bitcoin reserve data for FTX: 20,177 BTC down to 0.64 BTC in two days.
Frequently Asked Questions
What did Charles Edwards say about on-chain metrics?
Edwards, founder of Capriole Investments, said up to 99% of on-chain metrics are noise and add little value unless investors understand the key signals and how they’re constructed.
What happened to FTX’s Bitcoin reserves according to CryptoQuant?
CryptoQuant’s data showed FTX’s Bitcoin reserves fell from 20,177 BTC to just 0.64 BTC over two days in early November 2022, a drop other platforms didn’t clearly capture due to different address grouping methods.
