Solana Weekly Intelligence — August 10–16, 2026

2026-08-16 · Weekly Intelligence · SolDataLab Research Desk

Every week I answer the same five questions with the same rule: real data or no data. This week is the series baseline — my daily window begins August 10 — so the answers describe levels rather than changes: a 94-cent price range, series-floor DEX and fee prints, a $4.82B TVL plateau, and a stablecoin mix at its starting point ($2.90B USDT against $6.70B USDC). Baselines are where the next five weeks' trends become visible in hindsight; this report reads them before they move.

Question 1 — What did the market actually do?

Almost nothing, visibly: $76.21 → $75.92 → $76.22 → $75.56 → $76.20 → $75.34 → $75.28 — a 94-cent total range, the narrowest of the series, closing down 1.2%. But the relative print carries the signal: BTC fell 2.8% and ETH 1.5% the same week, so SOL's −1.2% is +1.6pp versus Bitcoin and +0.2pp versus Ethereum. In a down market, falling least is a form of strength, and as the opening entry of a series that would see +3.3pp and +10.0pp beats in the following two weeks, this is where the relative-strength trend begins.

Question 2 — Is DeFi still growing?

The honest answer at baseline: unknowable from this data alone — and said plainly rather than papered over. What the levels show: $10.88B of weekly DEX volume and $4.82B of TVL, both the lowest prints of the five-week series, both stable through the flat tape. What the following weeks would show: both climbing immediately on the breakout (+81.5% and +22.8% over two weeks, no down week). A base that holds its levels through a down tape and expands on the first trend week is a healthy base — that judgment is next week's to make, on next week's data.

Question 3 — What is happening to the stablecoin base?

The starting mix, with one detail worth flagging in hindsight: USDT ended the baseline week at $2.90B — its highest print of the five weeks I track — against USDC at $6.70B. The following weeks would rotate the mix one direction throughout: USDT sliding to $2.54B by mid-September while USDC climbed to $7.30B, roughly $360M out and $600M in. The baseline is where that trend is easiest to state: whatever drives it, it had not started yet. The mix this week is the counterfactual the rotation defines itself against.

Question 4 — Is fee revenue telling the truth?

At baseline, the fee series opens at $63.45M with Jito tips at $1.53M — the floor of the series. The annualized pace ($3.3B) is the quiet headline: even in the narrowest, quietest week of the dataset, the chain's fee revenue ran at a scale that pays for its security budget with room to spare. What would make the fee question interesting is movement — and the following weeks would deliver it: +30.9%, then to a $98.8M peak, a $108.3M series high by mid-September. The truth-telling starts here, at the floor: a baseline that admits it is a baseline.

Question 5 — What would change my mind?

Three falsifiable things, framed for a baseline week: a break below $75.28 — the Sunday close — would question the mid-$70s base itself; a DEX print next week below $10B would mark this week's level as a floor being lost rather than held; and a USDT print above $2.90B would defer the rotation story. Conversely, the bull setup strengthens if the range resolves upward on expanding volume — compression this narrow (94 cents) usually precedes expansion, and the fee floor at $63.5M gives the tape a level to step up from. The structural backdrop is unchanged: four US spot SOL ETFs trading, BSOL the largest fund, cumulative inflows compounding — with the anchor fund's $1B milestone (per Arkham) landing within two weeks of this baseline.

Health check

The ambers are labeled honestly: at a series baseline, most activity metrics are simultaneously the floor and the starting point, and calling them 'weak' would be as misleading as calling them 'strong.' The greens are structural: a full stablecoin base, a live ETF complex, a clean regulatory slate. The week's job was to hold the base — and the base held. The trend judgment belongs to the weeks that followed, which is exactly how a baseline should work.

Reading a week with no changes: the levels-only method

Most weeks, the five questions are answered with changes - WoW percentages, step-ups, divergences. This week has none: it is the series baseline, so the method changes, and saying how is more useful than pretending it does not. What a single week of levels can support: the levels themselves ($10.9B DEX, $63.5M fees, $4.82B TVL, $15.76B stablecoins, all machine-fetched); the relative print (+1.6pp versus Bitcoin in a down week - the one directional signal on offer); and structural facts that do not depend on time series (a live ETF complex, a full stablecoin base).

What it cannot support, stated plainly: trend (no prior week to compare), momentum (same reason), and confirmation (there is nothing yet to confirm). The health check's all-amber rows for activity metrics are the method made visible - each metric is simultaneously the series floor and its starting point, and grading it either 'weak' or 'strong' would both be claims the data cannot cash.

The method's payoff arrives later. Because this week's levels are fixed from real fetched data, every WoW figure the site ever computes - the +81.5% DEX spike, the 71% fee climb, the stablecoin rotation endpoints - traces back to rows that were never estimated or reconstructed. A levels-only week is not a wasted week; it is the week that makes every later percentage honest. The one genuine signal it did carry - relative strength in a down market - turned out to be the trend's opening entry, but the report says that with the series in hand, not from the baseline alone.

The baseline's prints in annualized terms

Even a week with no changes can be scaled, and the floor's scale is the more durable fact the baseline produces. The standing rule applies: annualizing a floor understates a run-rate just as annualizing a peak flatters one - but the floor is the number that persists, so this is the direction the arithmetic works in the series' favor:

The fee line is the one worth sitting with. A $3.3B annualized pace was paid in the narrowest, quietest week of the dataset - a 94-cent range on a down-market week - which means the chain's security-cost coverage and its fee economics do not depend on excitement. Every later week in the series, including the $108.3M peak, is measured against this floor; the floor is what the chain is, and the peaks are what it does.

The volume line carries the baseline's forward value differently: ~$566B annualized was the pre-breakout pace, and the same arithmetic two weeks later ran near $1.0 trillion on the spike week. Both numbers are method-flattered in opposite directions - floor understates, peak flatters - which is precisely why the site's standing rule quotes weekly sums in reports and reserves the annualized frame for scale checks like this one. The range between the two annualized figures is the honest description of how far the market moved; the weekly sums are how it is tracked.

Sources

Source: DefiLlama — TVL, DEX, fees, stablecoins, pricesArkham IntelligenceCoinMarketCap — Solana

Tags: NetworkDeFiStablecoinSOL ETF