What Is a Crypto Factor Stack? Combining Funding, OI, Volume, Regime, and Breadth

A factor is one measurable dimension of a market that carries information about it. On its own, a factor is rarely a decision; it is one reading among several. Price momentum is a factor. Funding is a factor. So are open interest, traded volume, the broad market regime, and breadth across the universe. A factor stack is simply the practice of measuring several of these at once and combining them into a single score, so no single dimension gets to decide alone.

The reason this matters is that any one indicator can be loud and wrong. A market can be ripping on momentum while funding screams that the move is already crowded. Open interest can be climbing while volume quietly thins out, which is a different and more fragile picture than open interest climbing on rising participation. A single-factor view forces you to pick which lens to trust and then live with its blind spots. A stack lets the factors check each other.

On perpetual futures, a small set of factors does most of the work, and each answers a distinct question.

Read alone, each of these is a half-truth. Read together, they describe a fuller picture: not just that a market is moving, but whether the move is clean, crowded, broad, supported by volume, and consistent with the regime.

• Momentum and relative strength ask whether the market is moving and whether it is leading or lagging its peers. This is the part a chart screener handles well, and it is the entry point, not the conclusion.
• Volume asks whether the move has participation behind it. A move on thin volume and a move on heavy volume can look identical on price and mean opposite things about how durable it is.
• Funding asks who is paying to hold the position. Persistently rich funding tells you the crowd is already leaning one way and is paying for the privilege, which is information about how crowded, not how strong, the move is.
• Open interest asks how much positioning has actually accumulated, because open interest is the stock of contracts that eventually has to unwind. The interaction between a price move and the open interest behind it is one of the richest readings in the whole stack.
• Regime asks what kind of market you are in at all: trending or chopping, expanding or compressing, risk-on or risk-off at the index level. The same single-factor reading means very different things in different regimes.
• Breadth asks whether the move is broad or isolated, because a single coin running while the universe is flat is a different story than a universe lifting together.

The case for a stack is robustness, not magic. Independent factors fail at different times and for different reasons, so combining them tends to be steadier than leaning on the best-looking single signal, which is usually the one that happened to work most recently. A stack also resists the trap of optimizing one indicator until it looks perfect on history, because the score has to satisfy several dimensions at once rather than being tuned to flatter one.

Stacking is not a guarantee, and it is important to say so. Factors can be correlated, so adding more of them does not always add proportionally more information; five readings that all really measure the same thing are closer to one factor wearing five hats. Factors also drift as market structure changes, so a stack that described the market well in one regime can degrade in another. And a stack is only as honest as the record kept on it. The whole exercise collapses if the factors are weighted by hindsight to fit the trades that already worked.

This is where methodology has to meet accountability. A factor stack produces a score, the score produces a short list of observations, and the only way to know whether the stack is any good is to record what those observations did over time, in public, with wins and losses kept at equal weight and the sample size visible. A stack you can only judge by its winning examples is indistinguishable from a stack that got lucky.

SENTINEL is built around exactly this factor stack — momentum, volume, funding, open interest, regime, and breadth combined into one score per Bybit perpetual — and every CORE observation it surfaces resolves to a public receipt. That standard cuts both ways on purpose. At the time of writing the recent CORE cohort is negative, and the performance page reports that plainly instead of quietly reweighting the factors until the page turns green. A factor stack is a process, not a promise, and the receipts are how you tell the difference. Nothing here is financial advice, and a published record is not a forecast; it is the evidence that lets you decide whether the process deserves your attention.