This piece covers the completed daily session of 2026-08-05 across 20 USD-quoted crypto pairs sourced from Coinbase Exchange daily candles. The session closed and is no longer active; every figure here is a historical record of that single day.

The post describes breadth, dispersion, and where pairs settled inside their own 30-day ranges. A separate table renders the full per-pair data; the text below draws on that table rather than restating every row.

What the Session Did

On 2026-08-05, only 6 of the 20 tracked pairs closed higher, while 14 closed lower. The median session change was -0.72%, indicating that the typical pair lost ground on the day even though a handful of names posted modest gains. APT led with a session change of +2.42%; FET sat at the other end at -3.62%. That gap — the dispersion figure — came to 6.04 percentage points.

The range-position column shows where each pair's close sat relative to its own 30-day high and low, with 0% at the low and 100% at the high. The distribution on 2026-08-05 was notably skewed toward the lower end of individual ranges. Several pairs — STX at 6%, XLM at 2%, XRP at 14%, and IMX at 15% — closed very near their 30-day lows. ADA was the clearest outlier in the other direction, closing at 82% of its 30-day range despite a session decline of -0.94%, reflecting how far it had run over the prior seven days (+17.94% on the 7d column). ETH closed at 73% of its range on a session gain of +2.08%.

The 7-day and 30-day columns in the table show that some of the session's losers had accumulated meaningful short-term gains heading in — DOT carried a 7d figure of +10.29% into the day before slipping -1.69% on the session. That contrast between recent trajectory and single-day outcome is visible across several pairs and is part of what the table is designed to surface.

Regime: What "Mixed" Describes

The session was classified as mixed. That label applies when breadth is neither strongly positive nor strongly negative, and when dispersion is wide enough that the session cannot be summarised by a single directional statement. On 2026-08-05, breadth of 6/20 sits clearly on the declining side — a broad-decline label would typically require more pairs closing lower with tighter clustering — but the 6.04-percentage-point spread between the strongest and weakest pair, and the presence of meaningful gainers alongside meaningful losers, pulled the classification away from a clean broad-decline reading.

A compressed session would show narrow dispersion with most pairs moving in a tight band. A dispersed session would show wide spread but without a dominant directional lean. Mixed sits between those: the spread is real, the losses outnumber the gains, but the cross-pair behaviour was too varied to resolve into a single narrative. The label describes what closed on 2026-08-05; it carries no information about the session that followed.

Cost Conditions on a Mixed Session

The practical cost question for any session is whether the moves on show are large enough to absorb the round-trip expense of being involved. On 2026-08-05, the median session change was -0.72%. A standard maker-taker fee schedule on a major spot exchange typically charges a taker fee in the range of 0.05% to 0.60% per side depending on volume tier, implying a round-trip cost of roughly 0.10% to 1.20% at the extremes. At the median move of -0.72%, a round-trip at even a mid-tier taker rate consumes a substantial fraction of the day's typical displacement — and that is before accounting for bid-ask spread or slippage on less liquid pairs.

The pairs that moved most on the session — APT at +2.42% and FET at -3.62% — offered enough gross displacement that round-trip fees would represent a smaller share of the move. Pairs clustered near the median, however, operated in territory where execution costs are not a rounding error. Spread and slippage tend to widen on pairs sitting near multi-week lows, where liquidity is thinner; several of the range-position readings on 2026-08-05 were in the single digits or low teens, which is a structural observation about where those books were positioned, not a comment on any individual trade.

What One Session Does Not Establish

A single daily session is a sample of one. The mixed classification on 2026-08-05 describes the internal structure of that day's closes; it does not indicate that adjacent sessions will carry the same character, nor that the pairs near their 30-day lows will remain there or recover from them. Range position at the close of one session is a description of where price ended relative to a trailing window — it is not a signal about where price goes next.

The dispersion figure of 6.04 percentage points is wide enough to matter for cost analysis, but dispersion itself is not stable from session to session. A wide-dispersion day can be followed immediately by a compressed one, or by another dispersed session in the opposite direction. The regime label is assigned after the session closes and reflects only what that session did.

The 7-day and 30-day figures in the table are trailing summaries. ADA's 7d reading of +17.94% and DOT's of +10.29% describe what those pairs did in a window that has already elapsed. Nothing in this post's data establishes whether those moves were complete, partial, or coincidental with the session-level behaviour observed on 2026-08-05. The structural failure of reading too much into a single session is not a matter of bad data — it is a matter of sample size.

The 2026-08-05 session closed with 14 of 20 pairs in the red, a median decline of -0.72%, and enough cross-pair variation to resist a single directional label — a day that the data describes clearly and that the data alone cannot explain.

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.