This post covers the daily session that closed on 2026-09-03 across 20 USD-quoted crypto pairs sourced from Coinbase Exchange candles. The session has ended; the figures below are final closes, not live prices.

The regime label for this session is broad-advance: every one of the 20 tracked pairs recorded a positive open-to-close change. What that label describes, what distinguished this session from adjacent regime types, and what a single session cannot establish are each addressed in turn below.

What the Session Did

On 2026-09-03, breadth reached its maximum: all 20 of 20 pairs closed higher. The median session change was +4.62%. The strongest mover was ADA at +9.99%; the weakest was IMX at +1.59%. The gap between them — dispersion — was 8.39 percentage points, a figure wide enough to matter for any strategy that holds a fixed basket rather than concentrating in a single name.

Range-position readings in the table show where each pair's close sat inside its own 30-day high-to-low range, with 0% at the 30-day low and 100% at the 30-day high. BTC closed at a range position of 95% and ETH at 92%, meaning both settled very near the top of their respective 30-day windows on this session. By contrast, XLM, ATOM, and APT each closed at range positions of 46%, sitting roughly in the middle of their 30-day spans despite posting session gains of +5.12%, +3.62%, and +2.99% respectively. STX and LINK closed at 84% and 83% of their 30-day ranges, consistent with their 30-day changes of +104.08% and +45.37% — the largest on that horizon in the tracked set.

The distribution of session gains was not uniform. A cluster of pairs — ADA, DOGE, XRP, LINK, and HBAR — posted moves between +6.33% and +9.99%, while a second cluster from DOT through IMX posted moves between +1.59% and +2.99%. The median of +4.62% fell between those two groups, with AVAX landing exactly on it.

Regime Context: What Broad-Advance Describes

A broad-advance label is assigned when the majority of tracked pairs close higher in a single session. On 2026-09-03, that condition was satisfied at its extreme: the count was 20 of 20. This distinguishes the session from a mixed regime, where gains and losses are roughly split, and from a compressed regime, where the range of moves is narrow regardless of direction. It also differs from a dispersed session, where a wide spread of outcomes occurs without a clear directional lean.

The 8.39-percentage-point dispersion on this session is notable. A broad-advance with high dispersion means the advance was uneven: participation was universal, but magnitude varied substantially. Pairs with weaker 7-day readings — FET at -4.24% over seven days, SOL at -4.80%, ATOM at -3.67% — still posted positive session changes on the 3rd, suggesting the session-level lift was broad rather than concentrated in pairs that had already been trending higher on the weekly horizon.

Execution Cost in Context

Coinbase Exchange publishes a tiered maker-taker fee schedule. At standard retail taker rates, a round-trip (entry plus exit) on a spot pair carries a cost in the range of roughly 0.50% to 1.00% of notional, depending on the fee tier and whether limit or market orders are used. Against the session moves recorded on 2026-09-03 — where even the weakest pair, IMX, gained +1.59% open-to-close — the round-trip fee represented a meaningful fraction of the available gross move for the smallest movers, and a comparatively small fraction for pairs like ADA (+9.99%) or DOGE (+7.49%).

Spread and slippage are separate from the posted fee. On a session with universal positive closes and above-average moves, bid-ask spreads on liquid majors such as BTC and ETH tend to be tighter than on lower-liquidity pairs, but that is a structural observation about microstructure, not a claim about conditions on any specific session. Slippage scales with order size relative to resting liquidity; nothing in the daily candle data speaks to intraday depth.

The relevant point for cost analysis is that a strategy clearing its round-trip fee on a day like this does so through the size of the move rather than through frequency or an extended holding period — the session itself provided the gross margin. On sessions with smaller median moves, that same fee becomes a larger hurdle relative to gross price change.

What One Session Does Not Establish

A single broad-advance session is a sample of one. The regime label describes what happened on 2026-09-03; it carries no predictive content about the session that follows. Broad advances have historically been succeeded by sessions of every regime type, including broad declines, and the classification itself is constructed only from the session that has closed.

The range-position data shows where pairs closed within their 30-day windows, but a high range position — BTC at 95%, ETH at 92% — does not indicate whether those pairs are near a reversal or continuing further. The 30-day range is itself a trailing window that shifts daily; a pair at 95% of its range is near its recent high by definition, but the range will reprice as new sessions close.

Dispersion of 8.39 percentage points means the basket was not homogeneous. A strategy that held the full set of 20 pairs equally weighted would have experienced a different outcome than one concentrated in the top or bottom performers — but which pairs will lead or lag on any subsequent session is not inferable from this data. The structural failure of any regime-following approach is that the regime is labeled after it has occurred, not before.

The 2026-09-03 session produced the widest possible breadth reading in this 20-pair set, with a median gain of +4.62% and a dispersion wide enough that the experience of holding any individual pair differed materially from the median.

Note: This explains how a mechanism works. It is educational and is not financial, investment or trading advice, not a recommendation, and not a signal service. Any figure describes a session that has already closed and says nothing about what happens next. Trading cryptoassets risks losing money, including all of it.