Methodology
One input failed testing and stayed on screen.
The Crypto Confidence Index reads the crypto market from the inside: whether capital is rotating into risk, where the cycle sits against its own history, and whether the signals that tend to move first are improving. One of its measures did not earn a place in the composite and is published beside it anyway, labelled for what it is.
A macro read cannot see leverage building inside crypto, capital sitting idle in stablecoins, or breadth quietly narrowing while the headline price holds up. Those are crypto-internal conditions, and they have their own cycle that does not always follow the economy.
The CCI exists to cover that ground. It reads cycle position, market structure and breadth, whether stablecoin capital is being deployed or pulled back, and how derivatives positioning is leaning. It runs as the counterpart to the macro read rather than as a replacement for it.
Keeping them separate is the deliberate part. Blending macro and crypto-internal conditions into one number produces something that always has an answer and never tells you which half of the market is speaking. Two readings that can disagree are more useful than one that cannot.
On-chain profit-taking is a well-known measure and an obvious candidate for a tool like this. Testing found it added no predictive value at the composite level, so it sits outside every basket, published as its own standalone context row. Including it would have made the composite look richer while making it worse. It stays visible because it is interesting to read, not because it earns its place in the score.
An earlier version of the top-side warning fired at the wrong moments often enough that validation ruled it out. It was replaced with a different approach built on the divergence between price and the leading signals. The replacement is documented in the tool's own source as a correction rather than quietly swapped in.
The profit-taking row is judged against its own rolling recent history rather than fixed levels, because the floor it reaches at cycle lows has drifted higher every cycle. A fixed line drawn from an earlier cycle would now sit somewhere the market no longer visits. A threshold set once is a threshold that expires.
The derivatives row measures the spread between perpetual futures and spot price. That is a proxy for positioning and leverage, and it is not the funding rate, which is a different thing that gets used interchangeably everywhere. It is labelled as the spread it is, which is less familiar and more accurate.
It can miss a top that arrives while conditions look healthy
The top-side warning depends on the cycle being extended and on the leading signals diverging from price. A high that forms without those preconditions can pass without a warning. This is the counterpart to the macro tool's blind spot rather than a fix for it, and neither tool closes the gap on its own.
The bottom sequence will not skip a stage
Recovery is read as an ordered repair process, and each stage must be reached before the next can activate. A sharp recovery driven by a single catalyst, without the usual sequence underneath it, may be recognised late or missed entirely. The ordering exists to stop the tool calling a bottom on the first good week, and that protection has a cost.
One signal can be misleading in a fast rally
Stablecoin deployment is read from the share of the market that stablecoins represent. When price rises faster than stablecoin supply grows, that share falls for arithmetic reasons rather than because capital is being deployed. The reading can look constructive when nothing has actually moved.
Weekly bars only
Nothing updates until the week closes. A flash crash and a sharp reversal inside the same week may leave very little trace. It is built for regime, and regimes are not a daily question.
The measure that failed is published beside the ones that passed.
Members see the composite, the leading line, the baskets behind them, and the standalone rows kept deliberately outside the score.
Discussion