DFDV — Q2 2026 Financial Report Analysis

A Solana holding company that lost $27 million in Q2, mostly on its tokens

DeFi Development Corp. used to be Janover, a small commercial-real-estate lending marketplace. In 2025 it turned itself into a "digital asset treasury": a listed company whose main job is to hold Solana's token, SOL, stake it to earn rewards, and raise money to buy more. Its Q2 2026 10-Q shows what that model looks like in a falling crypto market. The net loss was $27.3 million, against a $15.4 million profit a year earlier. Almost all of the swing came from the tokens: a $21.5 million net loss on digital assets, mostly from SOL it sold or converted into staking tokens below its purchase price. Revenue rose 67% to $3.3 million, but that is staking income, not a business with customers. By June 30, liabilities exceeded assets and shareholders' equity had turned negative.

At a glance

• Net loss of $27.3 million (–$1.00 per share), of which $21.5 million was losses on the SOL treasury. The first-half loss was $110.7 million, against a $14.7 million profit in H1 2025.
• Stockholders' equity of –$12.0 million at June 30, down from +$99.3 million at December 31, 2025. Total liabilities ($215.4 million) now exceed total assets ($203.3 million).
• Staking revenue of $3.3 million in the quarter did not cover running costs: operating expenses excluding token losses were $4.6 million and interest was $3.7 million. Management now funds the gap by selling new shares.

The quarter in numbers

Operating margin (the share of revenue left after running costs) is meaningless here: the token losses and gains flow through operating expenses, so a $3 million revenue line is set against a $21 million mark on the treasury. We show it only because it is the company's reported figure.

What the company actually owns

SOL appears in three places on the June 30 balance sheet:

The fair-value SOL alone tells the price story. It was worth about $124 per token at December 31 and about $73 at June 30 (dividing the reported fair value by units), a drop of roughly 41%. The company paid $135.2 million for the SOL in that line, so it sits about $70 million below cost.

The pledged $101.6 million is not free to sell. It backs $89.8 million of "digital asset financing arrangements": SOL and USDC borrowed at 10–13% annual fees, which lenders can liquidate if collateral coverage falls to around 150% and is not topped up. The MD&A says that of the $97.3 million fair value of its unpledged holdings, only $47.4 million is unencumbered.

The company's own headline figure is broader. Its August 12 shareholder letter counts 2,311,523 "SOL and SOL equivalents" against 35.3 million fully converted shares (which assumes its convertible notes turn into stock). That gives 0.066 SOL per share, up 24% from 0.053 a year earlier. SOL per share is management's own non-GAAP metric, not an audited figure.

Why the loss happened

The 10-Q breaks down Q2's $21.5 million net loss on digital assets:

• $17.6 million of realized losses from "converting a portion of our SOL holdings into liquid staking tokens as well as from selling digital assets". Liquid staking tokens are receipts for staked SOL that can still be traded. Converting to them counts as a sale for accounting purposes, which locks in the drop below cost.
• $10.0 million of impairments on those liquid staking tokens.
• Partly offset by $6.0 million of unrealized gains as SOL's price moved up within the quarter.

Below the operating line: interest expense rose to $3.7 million from $0.8 million, because the company now carries $126.1 million of convertible notes (principal due 2030) plus the SOL loans. A $3.2 million commitment fee on its equity line of credit sat in "investment and other expense". A $2.8 million gain came from buying back July 2030 notes for $5.0 million against a $7.9 million carrying value.

What the headline numbers hide

• The loss is mostly paper, but the cash burn is real. H1 net loss was $110.7 million, while operating cash outflow was $14.1 million. The gap is token marks and derivative losses, which are non-cash. But the $14.1 million outflow is more than twice the $5.98 million of H1 revenue. Per the shareholder letter, the company entered Q2 with about $27 million a year of cash costs: about $17 million of operating spending and $10 million of interest. Staking income does not cover that at current SOL prices and staking yields.
• Shareholders now fund running costs. Since late June management has sent all staking rewards into the treasury and paid expenses through its at-the-market share program (selling new shares into the market). It sold about 478,000 shares for $1.4 million. Management says this reduced SOL per share by about 1.4%, and it intends to continue until yield recovers. The stock traded at 0.8x net asset value (mNAV) on August 12. At that discount, every share sold gives new buyers more SOL per dollar than existing holders own.
• Share count nearly doubled year over year. Weighted basic shares rose from 14.1 million to 27.4 million, from the 2025 issuance that built the treasury. In H1 2026 the company went the other way: it bought back 1.6 million shares for $10.5 million and issued 1.16 million.
• Accounting cuts one way on part of the treasury. SOL is marked to market. The $20.5 million of liquid staking tokens is held at cost less impairment, so declines hit earnings but recoveries do not. That is how $10.0 million of impairments appeared in a quarter when SOL itself produced $6.0 million of gains.
• Leverage is high for a single-asset vehicle. Management's own August 12 figures were net debt at 98% of assets and 104% of SOL holdings, with total debt at 216% of market capitalisation. With negative book equity, a further SOL decline falls straight on common shareholders, after lenders and noteholders.
• One-offs both ways. Q2 2025's profit came almost entirely from a $21.2 million token gain in the treasury's first quarter. Q2 2026 includes a $2.8 million gain on debt buybacks and a $3.2 million equity-line fee, neither of which should recur at that size.
• The legacy business is gone. The board approved winding down Janover Capital Markets and Janover Insurance on March 31. Real estate revenue was $58,000 in Q2, so there is no operating business to cushion the treasury.

Takeaway: DFDV is now close to a leveraged bet on SOL with negative book equity. Staking revenue grew 170%, but it no longer covers interest and overhead. Shareholders are funding the gap through new share issuance at a discount to asset value, so per-share SOL ownership depends on the SOL price recovering, not on anything the company's operations can produce.

Outlook

Management withdrew its plan to give a June 2027 SOL-per-share forecast. The letter says the company missed its original June 2026 target and that past growth "came from select windows of opportunity". It kept a long-term target of 1.0 SOL per share by December 2028, about 15 times the August level, and says that assumes a "constructive crypto environment" in 2027 and 2028. It also expects lower operating costs from Q3 as one-time legal and accounting spend rolls off.

Things have moved since quarter-end. An October 5 8-K reported about 2,564,212 SOL and SOL equivalents worth $302 million, 11% more than on August 12. It gave preliminary estimates of double-digit SOL-per-share growth and more than a doubling of NAV per share from August 12 to September 30, as SOL recovered. The company also started paying dividends on a new variable-rate preferred stock ("CHAD", currently 13% a year on a $10 stated value) and set up a $300 million ATM program for it. That shifts funding from convertible debt to preferred equity, but a 13% dividend is a heavy fixed cost on a treasury yielding single-digit staking returns. In the Q3 10-Q, expected around mid-November, we'll check three things: whether stockholders' equity is positive again, whether operating costs actually stepped down, and how much preferred was issued against the SOL it bought.