Session Read 2026-09-28: 3 of 20 Pairs Higher
This post covers the completed daily session of 2026-09-28 on Coinbase Exchange, drawn from the desk's standard 20-pair USD spot candle set. The session closed before this piece was written; no prices here are live.
The regime label for the session is broad-decline. That label describes what the closed candles showed — it is not a characterisation of any session that has not yet ended.
Explore independent investment research, ratings, portfolio tools and market insights to help you make more informed investing decisions.
What the Session Did
Breadth on 2026-09-28 was 3 of 20 pairs closing higher. The three advancers — HBAR, LINK, and XLM — were not marginal: HBAR posted a session gain of +27.32%, LINK +10.16%, and XLM +7.69%. The remaining 17 pairs declined, with the median session change landing at -3.09% (ADA, which closed at exactly that figure, sits at the midpoint of the distribution). The weakest pair, NEAR, shed -10.46% on the session.
Dispersion between the strongest and weakest was 37.78 percentage points — a wide spread for a single daily candle. That figure reflects the gap between HBAR's outlier advance and NEAR's steep decline, with most of the field clustered in a narrower band of moderate losses. The table shows the full distribution; the notable feature is how isolated the three advancers were from the broader field.
Range position — where each pair's close sat inside its own 30-day high-to-low range — was elevated for the advancers: LINK closed at 98% of its 30-day range and XLM at 97%, meaning both closed near the top of the range they established over the prior month. HBAR sat at 85%. Among the decliners, most closed in the 49%–80% band of their own 30-day ranges, indicating that the session's losses, while real, did not push the majority of pairs to new 30-day lows. STX and ATOM were the exceptions, closing at 50% and 49% of their respective 30-day ranges.
Several of the harder-hit decliners on the session carry substantial 30-day gains in the table: NEAR's 30-day change stands at +158.14% despite the -10.46% session loss; OP shows +51.59% over 30 days against a -10.11% session; FET shows +49.57% over 30 days against -9.21% on the session. Those figures describe the context of where pairs arrived entering the session, not a causal explanation for what happened within it.
| Pair | Close (USD) | Session | 7d | 30d | Range pos. |
|---|---|---|---|---|---|
| HBAR | 0.122000 | +27.32% | +30.66% | +61.57% | 85% |
| LINK | 15.44 | +10.16% | +17.16% | +34.76% | 98% |
| XLM | 0.232352 | +7.69% | +7.92% | +29.26% | 97% |
| ETH | 2,687.41 | -0.03% | -3.18% | +9.35% | 74% |
| ICP | 3.13 | -0.47% | +3.37% | +26.08% | 78% |
| BTC | 83,456.74 | -1.19% | -3.62% | +6.68% | 69% |
| XRP | 1.50 | -1.34% | -2.59% | +7.12% | 61% |
| SOL | 118.81 | -2.60% | -0.08% | +12.50% | 79% |
| AVAX | 10.61 | -2.63% | -5.52% | +44.78% | 75% |
| ADA | 0.247150 | -3.09% | +0.82% | +22.48% | 76% |
| DOGE | 0.093860 | -3.10% | -5.95% | +10.09% | 56% |
| APT | 0.828900 | -4.00% | +4.32% | +52.60% | 84% |
| SHIB | 0.000006 | -4.05% | -5.80% | +10.29% | 61% |
| DOT | 1.18 | -6.75% | -1.94% | +39.71% | 75% |
| STX | 0.314100 | -7.18% | -8.00% | +28.84% | 50% |
| ATOM | 1.73 | -8.03% | -4.20% | +15.45% | 49% |
| IMX | 0.175410 | -8.06% | +16.10% | +36.33% | 75% |
| FET | 0.224650 | -9.21% | +9.59% | +49.57% | 69% |
| OP | 0.134280 | -10.11% | +5.32% | +51.59% | 74% |
| NEAR | 4.82 | -10.46% | +12.93% | +158.14% | 80% |
Regime and Its Boundaries
The broad-decline label is assigned when the majority of the 20 pairs close lower and breadth falls well below the ten-pair midpoint. At 3 of 20 closing higher, this session sits firmly in that classification rather than near its boundary with mixed, which would require something closer to an even split.
The session is distinguished from compressed by its dispersion: 37.78 percentage points between the top and bottom pair is not a tight cluster. It is distinguished from dispersed — which implies wide moves in both directions across the field — by the directional tilt: the wide spread here was driven primarily by a single outlier advance (HBAR) against a field of decliners, rather than by roughly equal volatility on both sides. A dispersed session typically shows more advancers contributing to the spread; here, the three advancers were the minority driving most of the upside dispersion while the 17 decliners provided the downside mass.
What this label does not encode is magnitude at the pair level. A broad-decline session where the median is -0.30% and one where the median is -3.09% carry the same label; the figures in the table are what distinguish them.
What This Session Meant for Execution Cost
On a session where the median move was -3.09% and the majority of pairs moved between roughly -1% and -10%, the round-trip cost of a spot trade is a meaningful fraction of the available move for anything near the median. Coinbase Exchange publishes a maker fee of 0.40% and a taker fee of 0.60% for retail volume tiers at the base level; a round trip at taker rates therefore costs approximately 1.20% in fees alone before spread or slippage is considered. Against a median session move of -3.09%, that fee floor consumes a substantial share of the gross move.
For the outlier pairs — HBAR at +27.32% or LINK at +10.16% — the same 1.20% round-trip fee floor is a far smaller fraction of the gross move. Size and holding period are the two levers that change this ratio; on a session with this dispersion profile, the cost burden falls most heavily on any strategy that traded near the median rather than the extremes.
Spread and slippage are not captured in daily candle data, but sessions with large single-pair outlier moves — particularly in smaller-cap pairs — are typically associated with wider quoted spreads and higher slippage on larger orders during the period of rapid price movement. The HBAR +27.32% figure describes what the candle recorded; it does not describe the cost of executing at any point during that move.
What One Session Does Not Establish
A single daily session is a sample of one. The broad-decline label describes what 2026-09-28 produced; it does not characterise the sessions before it or after it, and the desk does not use it to imply either. Regime labels are assigned after the close, from the data the session generated, and carry no predictive content.
The dispersion figure of 37.78 percentage points is large relative to a typical session, but one wide session does not establish that dispersion is structurally elevated. A single outlier — here, HBAR — can inflate this figure substantially; remove it and the spread narrows considerably. The figure belongs to this session and is not evidence about the distribution of future sessions.
The range-position column shows where pairs closed inside their own 30-day ranges on this specific date. A pair closing at 97% or 98% of its 30-day range has closed near the top of the range it built over the prior month; that is a description of where price was, not a statement about where it is going. Equally, a pair closing at 49% or 50% of its 30-day range has given back roughly half its 30-day move as of this session's close — again, a description of position, not a forecast of continuation or reversal.
The 30-day figures for several decliners are large — NEAR at +158.14%, OP at +51.59%, FET at +49.57% over 30 days — but a single session loss does not establish whether those 30-day moves are intact, reversing, or pausing. The data describes what closed; it does not resolve what those multi-week trends will do.
The 2026-09-28 session closed with 17 of 20 pairs lower, a median move of -3.09%, and a dispersion figure shaped largely by a single outlier advance; those are the facts the candles recorded.
Sources
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.