Solana Company's Flywheel Turned — Just Not Very Far Yet
The validator finally has third-party stake and the treasury finally has a revenue line, but 93% of Q3's staking rewards still came from the company's own SOL — and the stock has given back its premium-to-NAV window.
HSDT · Earnings Call · 2026-11-11
The flywheel's first turn — and how small the turn actually was
For three quarters, Solana Company management has sold a "self-reinforcing flywheel": advisory wins bring institutions on-chain, the Pacific Backbone validator monetizes them, the net revenue recycles into SOL, and a bigger treasury makes the company a more credible counterparty for the next institution. Through Q2 that was a slide. This quarter it became a P&L line. “The third quarter is when the flywheel started to turn.” — Choon Wee Chee, Management · 2026-11-11
Mechanically: third-party stake on the Tokyo-anchored validator went from a few hundred thousand SOL in late July to 1 million at September 30, and roughly 1.5 million by November 3. Total revenue was $3.4 million, of which $3.3 million was staking, at an ~82% gross margin. That is a genuine change of state — the company now books recurring fee revenue on other people's tokens, not just its own.
But the CFO's answer to an analyst question about the yield breakdown punctures the scale of it. “For the $3.4 million of revenue, we have 31,000 of staking rewards, of which approximately 29,000 units pertain to our individual SOL and the rest from third party.” — Agustina Tjandrasuwita, Management · 2026-11-11 In other words, roughly 2,000 of 31,000 SOL earned came from third parties — perhaps 7% of the quarter's revenue. The external delegation only started on July 31, so this is a run-rate story rather than a Q3 revenue story. Investors should model it as one.
Joseph Chee all but said so: “Going from 300,000 to over 1 million of SOL is probably going to slow down a bit in the fourth quarter. I don't think it's going to go from like here on to $3 million by the end of the quarter.” — Choon Wee Chee, Management · 2026-11-11 Note what this is: a CEO voluntarily decelerating his own most celebrated metric, one quarter after guiding that the infrastructure would be “actually going to be profitable for this year” — Choon Wee Chee, Management (likely CEO or senior executive) · 2026-08-14. The promise was kept. The slope is being haircut in the same breath.
The 100 basis points nobody should underwrite
Staking yield is the whole investment case in miniature. Solana Company delivered 5.9% APY against a ~4.9% network average in Q3 — 100bps of outperformance built from validator selection, MEV capture, and rebalancing. Fedor Shabalin of B. Riley asked the only question that matters: is that spread durable? Chee dodged it ("we're not providing a forward-looking statement here"). Cosmo Jiang gave the honest answer, and it is the most valuable passage on the call:
There have been major upgrades across one of the large validator clients recently that had been lagging for a while... it's possible that the average is getting pulled up because some people are catching up in their techniques that other people had. And when there's less dispersion amongst validators, then inherently — if we're picking the best, but the best are not... others are catching up to the best and the gap may also close a little bit.
That is a director of the company telling you the moat is a function of competitor sloppiness plus relationships, and that Alpenglow changed the reward distribution enough to add randomness. A 100bp edge on a 4.9% base is a 20% revenue premium on the staking line. If it halves, the "operating businesses generate revenue independent of SOL price" leg of the pitch gets thinner.
The GAAP gulf: $105 million of net income that isn't
Headline net income of $105.2 million against a $352.8 million loss a year ago looks transformational. It isn't operating income; it is SOL marking itself to market. The company recorded a $102.7 million unrealized gain on digital assets and receivables as SOL rose ~60% in the quarter from roughly $74 to roughly $118. The fundamentals history shows how violently this single line swings reported results: operating income ran −$353M in 2025Q4, +$123M in 2026Q1, −$100M in 2026Q2, then +$103.9M in Q3. Nothing operational changed across those four prints except the token price and one impairment.
The real operating story is duller and better: G&A fell to $3.4 million from $11.1 million in Q2 (which carried ~$6.8 million of PoNS-divestiture severance), below even the Q1 level. Chee's "lowest operating cost yet" claim checks out. But the CFO pre-empted the celebration — costs will normalize between what our current rate is and what our historic rate as Hong Kong opens. Meanwhile the top line is still small enough that valuation multiples are noise: price to revenue sits at 7.8x, down from 637.7x a year ago, and free cash flow was −$12M in the last reported quarter. The company is financed by its balance sheet, not its income statement.
From buyer to seller: the capital-allocation flip, and the tape that closed it
This was the quarter the two-sided toolkit got used in both directions. With shares below NAV, Solana Company repurchased 1.4 million shares at an average of $1.96 (~$2.8 million; $8.7 million year-to-date). Then, as SOL rallied and the discount to NAV narrowed, it priced a registered direct offering on September 30 — 4.36 million shares at $3.433, a 5% premium to NAV, plus equal warrants at $3.776 — for ~$15 million gross, and put the proceeds into ~125,000 SOL.
This was telegraphed, not sprung. In August, Jiang had said: “At these levels that we're looking to raise capital accretively as opposed to buying back aggressively.” — Cosmo Jiang, Director at Solana Company and General Partner at Pantera Capital · 2026-05-15 Credit where due: management said what it would do and did it. The framing — “we aim to give you more SOL per share over time” — Cosmo Jiang, Management · 2026-11-11 — is coherent, and 0.0274 SOL per diluted share is the metric to hold them to.
The problem is the tape. The stock peaked at $3.27 on September 22 and is down 42% from there; the last 90 days read +3.3% only because of a +43% eleven-week run followed by a −28% five-week slide. SOL itself is roughly 30% off its late-September high, which the analyst flagged on the call and Jiang conceded by noting blockchain activity is reflexive to price. The shares now trade below the $1.96 average at which the company bought back 1.4 million of them last quarter — which, if it persists, reopens the buyback side of the toolkit and closes the issuance side. The premium-NAV window that funded the SOL purchase was a two-week phenomenon.
Hong Kong, tokenized equity, and a theme the market is fighting
The genuinely new vocabulary this quarter is real world asset tokenization and institutional plumbing. Chee spent real airtime on Solana's on-chain RWA market climbing from ~$3.7 billion in July to above $4.6 billion by late September, on tokenized SK hynix shares, on SBI's yen stablecoin partnership and Shinhan's won-denominated fund MOU. He also disclosed selection by the Hong Kong SAR government's strategic-enterprise office — the only digital asset infrastructure company so designated in 2026 — and a planned Hong Kong headquarters, plus a targeted Q4 ISO 27001 certification. The company also voted its first on-chain governance ballot, publishing positions in advance.
That is a real institutional-adoption narrative, and it is not company-unique: tokenized equities, stablecoin adoption and the GENIUS Act all surfaced in the global keyword set this year. But the market's voting machine is running the other way right now — the global 30-day tape shows real world asset exposure among the decliners, alongside bitcoin price and AI-cloud-service baskets. The story and the tape are in open disagreement, which is precisely why a levered SOL proxy with $110 million of market cap is a volatile way to express a correct long-term view.
What to watch
The flywheel is real; it is just not yet big enough to detach the P&L from SOL. The Q4 scorecard is narrow: does third-party stake keep compounding from 1.5 million SOL, does the 100bp staking spread hold as validator dispersion compresses, does the Hong Kong build-out raise the cost base before it raises revenue, and does the discount to NAV stay wide enough to make buybacks the accretive trade again? Answer those and the "more SOL per share" promise has a denominator. Miss them and Q3's $105 million of net income becomes a reminder that this is a token position with an advisory business attached — not the other way around.