SOL Weekly News Review — September 21–27, 2026

2026-09-28 · Weekly News Review · SolDataLab Research Desk

A follow-through week, and the kind that tells you more than the breakout did. SOL opened at $111.15, never traded back below it on a weekly-close basis, and finished at $121.38 - its highest weekly close in the seven weeks this site tracks - up 9.2% against a Bitcoin that gained 4.0% and an Ethereum that gained 1.9%. The story underneath the price was structural: the consensus layer crossed from roadmap to running code, the institutional demand channel printed its largest week ever, and the dollar layer set a record while changing its composition. What did not happen is the more useful part - no mainnet date for Alpenglow, no evidence that the DEX decline was real usage, and no second opinion on whether one record ETF week is a trend. Below, the week's stories, each with the numbers and the take.

The week in one paragraph

SOL closed at $121.38, up 9.2%, its highest weekly close of the seven weeks this site tracks, outperforming Bitcoin by 5.2 percentage points and Ethereum by 7.3. TVL rose 7.4% to $6.64 billion, network fees rose 15.3% to $112.88 million - the highest weekly fee print of the series - and Jito MEV tips rose 26.4% to $2.87 million, also a series high. DEX volume fell 11.9% to $17.48 billion with a measurement story attached (a restated prior week and a reporting change at one aggregator, not a usage collapse). Stablecoin supply rose 6.3% to a record $16.74 billion on this site's DefiLlama basis, with the composition shifting toward USDT and a record third-bucket share. Solana ETFs took in $188.21 million - the largest week since launch, and the thirteenth straight of net inflows. Alpenglow completed its testnet activation. Every figure is from DefiLlama's September 28 vintage or the sources attributed below.

The week's data snapshot

Metric	Value	Source
SOL weekly close	$121.38 (+9.2%, highest weekly close of the 7-week series)	DefiLlama
Weekly high close	$122.08 (Sat, Sep 26)	DefiLlama
TVL	$6.64B (+7.4%)	DefiLlama
DEX volume (7d)	$17.48B (-11.9%, vs a restated $19.84B base)	DefiLlama
Network fees (7d)	$112.88M (+15.3%, series high)	DefiLlama
Jito MEV tips (7d)	$2.87M (+26.4%, series high)	DefiLlama
Stablecoins total	$16.74B (+6.3%, record on this site's DefiLlama basis)	DefiLlama
USDC / USDT on Solana	$7.30B (+4.9%) / $2.67B (+25.9%)	DefiLlama
Stablecoin third bucket	$6.77B, 40.4% of the base (both series highs)	DefiLlama
Solana ETF net inflow (week to Sep 25)	+$188.21M, largest week since launch; 13th straight positive week	SoSoValue
Record ETF day	$86.67M (Sep 25; BSOL $55.73M, GSOL $18.47M)	SoSoValue via CoinGlass
Alpenglow (SIMD-0326)	testnet switch completed Sep 24 (slot 444,625,255); devnet Sep 25; no mainnet date	Solana Compass / Anza
RWA on Solana	$4.6B (Sep 23, +11.5% over 30 days)	RWA.xyz

The through-line: this was a week where the things that usually confirm a price move - fees, tips, TVL, stablecoins, institutional flow - all pointed the same direction, and the one series that fell (DEX volume) fell for reasons that have nothing to do with whether anyone is using Solana. That is the rare configuration this column looks for, and it is why the honest read is bullish without being loud about it.

1. SOL closed at $121.38 - its highest weekly close in seven weeks

The daily-close series (DefiLlama, each point the price at 00:00 UTC of the labelled date) walked from $111.15 on Monday to $118.87 on Tuesday, held the $118s through Wednesday, pulled back to $114.99 on Thursday, recovered to $117.01 on Friday, stepped to a $122.08 high close on Saturday, and faded 0.6% to $121.38 on Sunday. That is a 9.2% week on the rounded figures, a market cap of roughly $71.3 billion on 587.7 million circulating SOL, and a position 58.8% below the $294.33 all-time high of January 19, 2025. The week's shape matters more than the percentage: every midweek test was bought, the low close was the Monday open, and the high close was the Saturday print - an uptrend that intends to continue looks exactly like this on a daily-close series.

My take: Two consecutive 9-12% weeks, the second one holding the first's breakout rather than giving it back, is what a trend week looks like in the relative-performance panel - not a squeeze, because the structure is orderly, and not a single-session spike, because the move distributed across six sessions. The caveat this site keeps repeating still holds: at $121.38, SOL is barely halfway back from its September low and 59% below its peak. This is a recovery with room, not a rally that has run its course, and the data does not support calling it either more or less than that.

2. Alpenglow reached the stage where it can actually fail

Alpenglow, Solana's consensus overhaul delivered as SIMD-0326, went live on the Solana testnet on September 23 with epoch 1042, and the switch completed on September 24 at slot 444,625,255, after 82% of testnet stake had acknowledged the network's final TowerBFT vote. Devnet completed its own switch on September 25. The new protocol replaces TowerBFT with Votor, a direct-voting design whose target is transaction finality in roughly 150 milliseconds instead of the roughly 12.8 seconds current mainnet finality takes. Source: Solana Compass and Anza's testnet updates, as carried by CoinMarketCap community roundups on September 27.

What 150 milliseconds actually buys is the removal of the revert window - the seconds during which a rational counterparty must assume your transaction might unwind. That window is where institutional risk lives: bridges wait for finality, exchange deposit-crediting keys off finality depth, oracle updates and perps settlement sequences compress against it. Sub-second finality with Solana's throughput is a different product to sell to payment and market-infrastructure buyers than 13-second finality, even if retail never notices the difference. It does not change fees, throughput, or validator economics - those are separate tracks - and it is not Firedancer.

My take: The milestone is real, but the market's job is to price mainnet, and mainnet has no date. The September 28 calendar entry some outlets carried is the tentative activation of the Agave 4.3 feature set on mainnet - a routine runtime upgrade, not Alpenglow - and conflating the two is exactly the error this column exists to catch. What changed this week is the stage of the roadmap: a testnet switch is the first point at which an upgrade can observably fail - validator clients crash, vote latency shows its tail, edge cases become incidents. That is progress you can measure, and it is also why no mainnet date exists yet. Watch validator-client readiness and governance actions specifying an activation epoch; a date from any other source is noise.

3. Solana ETFs: a record week, and the thirteenth straight

US spot Solana ETFs took in $188.21 million of net creations across the five trading days of September 21-25 - the largest weekly total since the funds launched, and the thirteenth consecutive week of net inflows. The daily prints were $26.10M (Sep 21), $28.87M (Sep 22), $13.77M (Sep 23), $32.81M (Sep 24) and a single-day record of $86.67M on September 25, of which Bitwise's BSOL contributed $55.73M and Grayscale's GSOL $18.47M. For context, the prior week's take was $13.2 million - the complex just printed more in one Friday than it did in the entire week before. Source: SoSoValue daily flow data, as reported by SolanaFloor and CoinGlass on September 25-26.

My take: Two records in one week - largest week, largest day - matter less than the streak behind them. A fund complex that has taken net inflows every week for thirteen, through a flat week and a breakout week alike, is behaving like a structural buyer rather than a momentum trade. The channel is still small against a $71 billion asset, and $188.21M cannot mechanically move that market cap - this site's discipline is to attribute correlation on the flow data, not causation. But the direction is the information: the only demand layer this dashboard tracks that has grown every week for three months chose the breakout's second week to set its record, and the record day preceded Saturday's push to the weekly high by one session. One record week is a headline; thirteen quiet ones plus a record is a trend forming in the open.

4. The dollar layer set a record and changed shape

Total stablecoin supply on Solana reached a record around September 25. The Solana Foundation's official account cited $17.3 billion; SolanaFloor's tracker put it at $17.39 billion. This site's own series, which uses DefiLlama's USD-pegged chart for Solana, printed $16.74 billion at the week-ending snapshot - the methodology differences (which issuers and peg types are counted) explain the spread, and this site states its basis rather than picking the biggest number. Source: Solana Foundation X account and SolanaFloor, September 25; DefiLlama stablecoin charts, September 28.

The composition moved as much as the total, and it is the more interesting half. USDT on Solana jumped from $2.12B to $2.67B (+25.9%) - its first sharp expansion in weeks - while USDC rose 4.9% to $7.30B, matching its six-week high. The non-USDC/USDT bucket, where issuers like PayPal USD and the smaller dollar tokens live, rose to $6.77B, its highest print of the seven weeks this site tracks, and now holds 40.4% of the base.

My take: A record total and a record third-bucket share at the same time is what resilience looks like for a dollar layer. A 40% share outside the two majors means Solana's stablecoin base no longer depends on a duopoly - a structure last cycle's chains never had. The USDT-led composition shift also fits the kind of week this was: trading-driven demand tends to show up first in USDT minting and in the long tail of tokens, not in USDC settlement. That is exactly the pattern you would expect in a 9.2% price week, and it is visible only because this site keeps the series on one stated methodology instead of cherry-picking the biggest tracker number.

5. ZetaChain holders voted to become a Solana token

ZetaChain token holders approved winding down the project's own Layer 1 blockchain and migrating ZETA to Solana as a native SPL token, citing speed, liquidity and AI-agent infrastructure. The vote confirms a migration that had been proposed earlier in the month. Source: Coingabbar's September 27 roundup on CoinMarketCap community, citing the project's governance channels.

My take: One migration is a headline; the pattern is the story, and this is the second or third time this year a dedicated chain has chosen to retire its validator set and become a Solana token rather than keep competing for the same liquidity. That is the ecosystem's quietest form of endorsement - louder than any partnership announcement, because it is a real economic decision with real opportunity cost, made by the people who hold the token. It compounds the liquidity argument that already made Solana the default destination this cycle: every chain that dissolves into an SPL token makes the Solana liquidity pool deeper and the case for launching a new L1 weaker. Watch whether the next one names Solana in its migration vote, because the precedent is now a template.

6. Bitquery put a number on the bot question: 58%

Blockchain analytics firm Bitquery published research in September estimating that roughly 58% of Solana DEX volume is bot-like - sandwich and arbitrage flows that churn size without expressing human conviction. The finding landed the same week Solana's DEX total fell 11.9% to $17.48 billion, and it is the statistical backdrop to an anomaly this site observed while fetching: Jupiter's reported volume dropped far more sharply than its fee line, which points to a reporting change at the aggregator level rather than a usage collapse. Source: Bitquery research, as reported September 22-26.

My take: The number is an estimate with a wide error bar, but its direction matches what the fee tape says - fees per dollar of DEX volume run near 0.65% precisely because so much of the volume is adversarial flow paying for priority. That is not a knock on the chain; it is the tax real users pay for fast settlement, and it is measurable. The practical consequence is the one this site has been pushing for weeks: treat DEX volume as an upper bound on activity, and treat fees ($112.88M, a series high, on a steady $15-17M daily band) and tips ($2.87M, also a series high) as the harder currency. Wash-volume disclosures are a net positive for anyone using DEX figures as an activity signal, because they turn a vanity number into something you can discount honestly.

7. The measurement story inside the DEX number

Two things sat inside this week's DEX figure that have nothing to do with usage. First, DefiLlama restated the prior week's DEX total upward to $19.84B on September 28, so this week's -11.9% is measured against a bigger base than the $18.46B published a week ago - on that older basis the week would read -5.3%, not -11.9%. Second, Jupiter's reported volume fell far more sharply week over week than its fee line did, a divergence the fee data - which held its normal range - does not support as a real usage drop. Sunday's aggregate printed $1.08B against Saturday's $2.16B, a gap that carries the Jupiter reporting change and should be read as a floor, not a collapse. Source: DefiLlama chain dashboard, September 28 vintage.

My take: This is the fourth consecutive fetch window in which DefiLlama revised a recent Solana day or week after first publication. The lesson for readers is mechanical, not theological: never treat a fresh weekly DEX or fee print as final, and never let a single week's movement drive a conclusion when the source is still settling its own books. The site's rule is to state the caveat rather than adjust a number it cannot recompute - which is why this article flags the anomaly instead of 'correcting' the DEX series downward by an amount no free source lets us verify.

8. From the institutional file: hires, a perps push, and $4.6B of RWA

Three smaller items from the institutional file, each a quiet signal rather than a headline. The Solana Foundation added two senior hires with backgrounds at rival exchanges, part of a push into tokenized finance and payments infrastructure (source: Coingabbar roundup, September 27). A trading competition with a seven-figure prize pool launched, aimed at drawing perpetual-futures activity onto the chain (same source). And RWA.xyz's dashboard put Solana real-world-asset value at $4.6 billion on September 23, up 11.5% over 30 days - a private-credit and treasury-token base that keeps grinding higher without a single headline day.

My take: None of these moves the tape on its own, but together they describe a build-out that the price week was riding rather than causing. Senior hires from exchanges signal an institutional-sales posture; a perps competition signals a bet that derivatives activity is a channel worth subsidizing; and an RWA base crossing $4.6B quietly makes Solana a top-three chain for tokenized real-world assets by this site's tracked dashboard. The through-line with the ETF and stablecoin records is that the demand-side infrastructure - regulated, institutional, dollar-denominated - is what got built this week, and the price simply noticed.

Three stories that mattered most

Eight items make a long list, so here is where I would spend attention if I could only follow three:

The confirmation configuration (items 1, 3-4, 8). Fees, tips, TVL, stablecoins and ETF flow all rose with price, and the one series that fell (DEX) fell for measurement reasons, not usage ones - the rare week where the confirmation panel is unanimous.
The record ETF week inside a thirteen-week streak (item 3). $188.21M with an $86.67M record day: the only demand layer this site tracks that has grown every week for three months chose the breakout's second week to set its record.
Alpenglow reaching the stage where it can fail (item 2). A completed testnet switch is the first observable checkpoint between roadmap and mainnet - and the reason no mainnet date exists yet.

What I'm watching next week

The week earned a clean bullish read, but every trend this site has tracked broke on a second datapoint, not the first. The items below are the ones that would convert this week's headlines into a continuation - or turn them back into spikes.

Whether the record ETF week repeats. One record week is a headline; a second would make thirteen quiet weeks plus two records a trend. A first down week after thirteen up ones would not break the trend by itself, but it would remove the only demand layer that has grown every week.
Whether stablecoin supply holds above $16.4B or fades back below it. Above, the record reads as a level change; below, it reads as mint-and-redeem churn.
Whether the DEX fee line follows volume down. Fees hit a series high this week; if next week's fee tape rolls over while volume is already soft, the activity layer stops confirming the price layer.
Any Alpenglow mainnet scheduling signal from validator-client teams. Testnet success is not a date; a governance action specifying a mainnet activation epoch would be the first real one.
Whether SOL holds $114.99 (Thursday's test) and $111.15 (the week's opening print). A weekly close back below the open hands back the entire move and restores the September band.

Source: Solana Compass / Anza - Alpenglow testnet activation (epoch 1042, slot 444,625,255)SoSoValue / SolanaFloor - Solana ETF daily net inflows, week of Sep 21-25, 2026 CoinGlass - Solana ETF flow table (BSOL/GSOL daily prints)Solana Foundation - record stablecoin supply announcement, Sep 25, 2026 DefiLlama - Solana prices, TVL, DEX, fees, stablecoins (Sep 28 vintage)Bitquery - bot-like share of Solana DEX volume, September 2026 Coingabbar via CoinMarketCap community - ZetaChain vote, Foundation hires, perps competition (Sep 27, 2026)RWA.xyz - Solana real-world-asset value, Sep 23, 2026

Did Alpenglow go live on Solana mainnet this week?

No. Alpenglow (SIMD-0326) completed its activation on testnet on September 24, 2026, and on devnet on September 25. No mainnet activation date has been announced. The September 28 date in some calendars is the tentative Agave 4.3 feature activation on mainnet, which is a routine runtime upgrade, not Alpenglow.

How much did Solana ETFs take in this week?

$188.21 million of net inflows across September 21-25, the largest weekly total since the funds launched and the thirteenth consecutive week of net inflows. The single-day record was $86.67 million on September 25 (BSOL $55.73M, GSOL $18.47M) - more than the entire prior week's $13.2M take.

Did stablecoins on Solana really set a record?

Yes, though the number depends on the tracker: the Solana Foundation cited $17.3 billion and SolanaFloor $17.39 billion around September 25, while this site's DefiLlama-based series printed $16.74 billion at the week-ending snapshot. The spread is methodology, not error - each counts a different set of issuers and peg types. The composition shift (USDT +25.9%, record third-bucket share of 40.4%) is the more informative part.

Why did DEX volume fall 11.9% while everything else rose?

Part of the decline is measurement: DefiLlama restated the prior week's DEX total upward to $19.84B on September 28, and Jupiter's reported volume dropped far more than its fee line, which suggests a reporting change rather than a usage collapse. Against the $18.46B print published a week ago, this week would read as a 5.3% decline instead of 11.9%. Bitquery's ~58% bot-volume estimate is the structural backdrop.

What is the single most important thing to watch next week?

Whether the record ETF week repeats. Twelve weeks of modest inflows followed by a record is a trend only if it continues; a single record week is a headline. The secondary watch is any Alpenglow mainnet scheduling signal from validator-client teams, and whether stablecoins hold above $16.4B.

Need on-chain energy without the price tag?

Rent TRON Energy at Tronsell →

A service we run and trust: a ~400M TRX self-operated energy pool, with 60-90% savings versus on-chain energy costs.

Tags: SOL ETF Stablecoin Network DeFi