This post covers the completed daily session of 2026-08-16 across the twenty USD-quoted crypto pairs tracked by Skia Paper from Coinbase Exchange candles. The table rendered alongside this piece shows each pair's close, session change, 7-day and 30-day change, and where the close sat inside the pair's own 30-day range.

The session was classified mixed — a label that describes what the closed session produced, not a characterisation of any session that has not yet opened.

What the Session Did

Breadth was narrow: only 2 of 20 pairs closed higher on 2026-08-16. ICP led at +1.22% and STX followed at +0.33%; every other pair was flat or negative, with DOGE printing exactly +0.00%. The median session change across all twenty pairs was -0.82%, meaning half the board moved less than that and half moved more.

The weakest pair was FET at -4.84%, and the distance between the strongest and weakest — dispersion — was 6.06 percentage points. That figure is moderate: wide enough to separate pairs meaningfully, but not the kind of extreme spread that appears in high-volatility sessions.

Range positions, which show where each close sat between the pair's 30-day low (0%) and 30-day high (100%), skewed heavily toward the low end of recent history. FET closed at 1% of its 30-day range, XLM at 1%, HBAR at 2%, XRP at 3%, and APT at 3%. LINK was the notable exception, closing at 81% of its 30-day range despite a session decline of -0.82%. ICP closed at 65% and ATOM at 62%, the only other pairs in the upper half of their ranges. The majority of the board huddled in the lower third of their respective 30-day windows.

The 7-day and 30-day columns in the table show the wider context. LINK carried a 7-day gain of +14.64% and a 30-day gain of +13.71% into this session. STX, by contrast, sat -25.97% over 30 days despite a small positive session. Those divergences are visible in the table and are not reconciled by a single day's movement.

Regime: What Mixed Means Here

The mixed label is assigned when the session does not fit cleanly into broad-advance, broad-decline, compressed, or dispersed. On 2026-08-16, the breadth figure of 2/20 would ordinarily suggest a broad-decline reading, but the dispersion of 6.06 percentage points and the uneven positioning of pairs within their 30-day ranges produced enough internal variation to push the classification away from a clean broad-decline.

A broad-decline session typically features tight clustering of losses across the board alongside low dispersion. Here, the losses were widespread but the magnitude varied considerably — from near-zero moves in AVAX (-0.06%) and BTC (-0.29%) to the sharp decline in FET (-4.84%) and APT (-3.54%). That spread of outcomes within a predominantly negative session is the defining feature of the mixed label as applied to 2026-08-16.

Compressed sessions, by contrast, show low dispersion regardless of direction. Dispersed sessions show high dispersion with less directional consensus. This session sat between those poles: direction was mostly negative, but the spread of outcomes was not trivial.

Cost of Participation in Sessions Like This

When the median session move is -0.82% and most pairs are clustered in the lower portion of their 30-day ranges, the relationship between round-trip transaction costs and session-level price movement becomes relevant as an observation about execution, not as guidance.

A standard maker/taker fee schedule on a major spot exchange runs roughly 0.05% to 0.60% per side depending on tier and order type, producing round-trip costs in the range of 0.10% to 1.20% of notional. On a session where the median move was -0.82% and many pairs moved less than 0.50% in absolute terms — BTC at -0.29%, AVAX at -0.06%, ETH at -0.36% — a round-trip fee consumes a material fraction of the session's total price movement for those pairs. Only the outliers at either end of the dispersion range (ICP at +1.22%, FET at -4.84%, APT at -3.54%) produced moves large enough to dwarf typical round-trip costs.

Slippage compounds this. In a session where most pairs are sitting near their 30-day lows, bid-ask spreads on thinner pairs tend to widen relative to calmer periods, raising the effective cost of execution beyond the headline fee rate. The range-position data in the table — with XLM at 1%, XRP at 3%, HBAR at 2% — reflects pairs that have been under sustained selling pressure over the 30-day window, a condition historically associated with wider spreads on spot venues.

The only strategy structures that clear a round-trip fee when session moves are this small are those operating at high frequency, large size, or extended holding periods. A single-session entry-and-exit on a pair that moved less than 0.40% on 2026-08-16 does not generate enough gross movement to absorb even the lower end of the fee range.

What One Session Does Not Establish

A single daily session is a sample of one. The mixed classification describes what the session that closed on 2026-08-16 looked like; it carries no predictive content about the session that follows. The label is a post-hoc description applied after candles close, not a signal generated before they open.

Breadth of 2/20 is a notable figure, but it has appeared in sessions that were followed by recoveries, by further declines, and by lateral movement in roughly equal historical frequency. The figure describes participation, not momentum. Nothing in the breadth count, the dispersion measure, or the range-position data establishes what any pair will do in the next session.

The 30-day range positions are particularly prone to misreading. A pair closing at 1% of its 30-day range — as FET and XLM did on 2026-08-16 — is near the bottom of recent history, which is a factual statement about price location. It is not evidence of a floor, a support level, or a reversal condition. Range position measures where a close sat; it does not measure why it sat there or what comes next.

Similarly, LINK's 30-day gain of +13.71% and its range position of 81% describe a pair that has been stronger than most of the board over the past month. Whether that relative strength persists, reverses, or mean-reverts is not answerable from the data in this session read. The session desk publishes what closed; the interpretation of what follows is outside the scope of what a single day's candles can support.

The 2026-08-16 session closed with 18 of 20 pairs in the red, a median move of -0.82%, and most of the board sitting in the lower reaches of its 30-day range — a configuration that is accurately described by the data and accurately limited in what it implies about anything beyond itself.

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.