Summary

  • Greenlane’s BERA and BERA-equivalent holdings increased 4.67% from 77,712,731 units at 31 March to 81,342,311 at 30 June 2026. Period-end issued Class A shares rose 10.25%, from 629,946 to 694,544.
  • The exact ratio therefore fell 5.06%, from 123.364 to 117.116 units per issued share. The company’s approximately 37% increase is a valid year-end comparison; it is not a sequential Q2 increase.
  • Greenlane defines BERA-per-share narrowly and transparently. It excludes shares issuable under pre-funded warrants and other convertible securities, while the 10-Q reports 3,791,510 warrants outstanding and uses a different treatment for nominal-price pre-funded warrants in basic EPS.
  • Unit growth did not protect dollar value. Q2 fair value was US$16.442 million against a US$70.223 million cost basis, and 42.10% of the units were restricted at quarter end.

A treasury can acquire more of its chosen asset and still give each issued share a smaller claim on that asset.

Greenlane Holdings did exactly that in the second quarter of 2026. Its BERA and BERA-equivalent holdings rose by 3,629,580 units between March and June. Over the same interval, the number of issued and outstanding Class A shares increased by 64,598. The numerator grew; the denominator grew faster.

The resulting movement is small enough to disappear inside a year-end headline and large enough to matter. Greenlane reported approximately 123 BERA per share at the end of Q1 and approximately 117 at the end of Q2. Using the exact figures in the two 10-Qs, the ratio declined from 123.364 to 117.116, or 5.06%.

Greenlane’s Q2 release instead says BERA-per-share increased approximately 37%, from about 86 at December to about 117 in June. That statement is arithmetically compatible with the sequential decline. It chooses a longer starting point. A reader needs both comparisons because they answer different questions: the strategy remained ahead of year-end, but lost ground during the latest quarter.

Rebuild the ratio before interpreting it

At 31 March, Greenlane reported 77,712,731 units of native BERA and BERA-equivalent tokens and 629,946 issued Class A shares. Dividing one by the other produces 123.364. At 30 June, the corresponding figures were 81,342,311 and 694,544, producing 117.116.

The bridge explains the fall. Token units increased 4.67%; issued shares increased 10.25%. Greenlane’s equity roll-forward identifies the movement: 33,085 shares came from pre-funded-warrant exercises, 31,627 from strategic-advisory-warrant exercises, and 114 fractional shares were eliminated in connection with the reverse split. The arithmetic closes exactly from 629,946 to 694,544.

This is not evidence that the treasury failed in every sense. The company held more token exposure at quarter end and remained above the year-end ratio. It is evidence that accumulation alone is not the objective management says it is pursuing. If long-term BERA per share is the governing measure, every capital-structure action belongs in the same ledger as every token purchase and reward.

The distinction is especially important after reverse splits. Greenlane completed a one-for-750 reverse split in June 2025 and a one-for-eight split in April 2026. The filings retroactively adjust the presented share and per-share figures. Those splits change the unit scale, not the underlying economics, and they should not be mixed with unadjusted historical counts.

The headline denominator is deliberately narrow

Greenlane supplies an unusually useful footnote. BERA-per-share is the number of BERA and BERA-equivalent units divided by Class A shares issued and outstanding at the applicable date. It is not a measure of financial performance or fair value, it does not capture changes in BERA’s market price, and it excludes shares issuable upon exercise of pre-funded warrants or other convertible securities.

That makes 117 a valid answer to a narrow question: how many BERA-equivalent units sat behind each issued Class A share on 30 June?

It is not a diluted asset-per-share measure. At the same date, Greenlane reported 3,791,510 warrants outstanding, with a weighted-average exercise price of US$0.08 and a weighted-average remaining life of 5.16 years. The warrant count was about 5.46 times the issued-share count. Some warrants may expire, some may have conditions, and exercise can bring cash or change other parts of the balance sheet. A mechanical fully converted division would therefore be a sensitivity, not a forecast.

The more revealing comparison is inside the same 10-Q. Greenlane used 4,097,688 weighted-average Class A shares to calculate Q2 basic and diluted net loss per share. It explains that pre-funded warrants with nominal exercise prices enter basic EPS from their issuance dates because the price is non-substantive and exercise is considered virtually assured. Other potentially dilutive securities were excluded from diluted EPS because the company reported a loss.

There is no contradiction. One KPI uses period-end issued shares and expressly excludes potential shares. GAAP EPS uses a period-average denominator and includes a specified class of pre-funded warrants under its accounting rule. The two figures are built for different purposes. Their gap is the reason every “per share” chart should state which claims it counts.

More units did not mean more dollar value

The second ledger is price.

Greenlane’s Q2 units carried a historical cost basis of US$70.223 million and a fair value of US$16.442 million. Fair value was US$53.781 million below cost and represented 23.41% of that cost basis. From March to June, fair value fell 51.97%, from US$34.232 million, even as the unit count rose.

The company recognized a US$19.142 million non-cash fair-value loss during Q2 and US$32.011 million for the first half. “Non-cash” describes the accounting entry in the period; it does not make the loss economically irrelevant. “Loss” also does not mean Greenlane spent US$19.142 million of cash in Q2. The assets are remeasured at each reporting date, with the change running through earnings.

Cost basis answers what was paid or recognized when units were acquired or received. Fair value answers what the position measured at in the principal market at 11:59 p.m. UTC on the reporting date. Neither guarantees the price of a future sale. A token-per-share metric omits this price vector by design, which is why it can rise while economic value per share falls.

That is the central discipline for any digital-asset treasury. The unit ledger may show successful accumulation. The market-value ledger may show destruction. Both can be true without either number being fraudulent or useless.

The numerator is not one pile of immediately saleable tokens

Greenlane’s numerator includes native BERA and BERA-equivalent positions such as sWBERA, siBERA and iBERA. The filing describes receipt tokens, staked or wrapped positions and other protocol-native instruments economically linked to BERA. Economic linkage is not identical to immediate liquidity.

At 30 June, 34,242,942 units with fair value of US$6.922 million were restricted. Another 47,099,369 units with fair value of US$9.520 million were unrestricted. Restricted units were 42.10% of total holdings. The restrictions reflected token-allocation, treasury and protocol arrangements, including vesting after an initial cliff.

Greenlane also deploys assets into staking and validator activity. These activities can produce rewards, but they introduce protocol rules, unbonding or transfer timing, smart-contract exposure and operational dependence. The Q2 filing reports US$0.309 million of staking and yield revenue. That income was much smaller than the period’s fair-value movement and should not be annualised as though reward rates and token prices were fixed.

The control surface extends beyond the company. The board’s Digital Assets Committee oversees policy and deployment. Management chooses acquisitions, protocol instruments and operating liquidity. Berachain’s rules influence rewards and transfer paths. Market participants set the measurement price. Warrant holders control whether eligible instruments are exercised under their terms.

Liquidity has been converted into several kinds of exposure

Greenlane began the year with US$32.513 million of cash and cash equivalents. The balance fell to US$13.320 million at 31 March and US$6.065 million at 30 June. The company also held US$8.057 million of stablecoin-related protocol instruments at Q2.

Those protocol instruments are not classified as cash equivalents. Greenlane says they consist substantially of dollar-denominated stablecoins deployed into yield-generating positions and carry smart-contract, protocol-liquidity and governance risks not present in directly held stablecoins. Management includes them in its broader liquidity assessment, but the accounting separation is economically useful.

Working capital moved from about US$28.9 million at year-end to US$13.8 million at Q1 and US$8.9 million at Q2. Management said cash, protocol instruments, expected cash flows and cost reductions were sufficient for at least twelve months and that estimated operating cash needs did not rely on digital-asset holdings. That is relevant counterevidence to a near-term insolvency narrative. It is not a reason to merge cash, protocol positions and BERA into one frictionless liquidity total.

The board had also authorised up to US$2 million of share repurchases. As of the Q2 filing, Greenlane had made none. An authorisation can signal an option; it cannot be credited as a reduced denominator until shares are actually repurchased and retired or otherwise removed from the applicable count.

The right scorecard has four columns

Greenlane’s BERA-per-share measure is useful precisely because its limitation is visible. It forces management to put token accumulation beside issued-share growth. Q2 then supplies an honest negative observation: holdings rose, but not fast enough to preserve the sequential ratio.

The measure becomes misleading only when a reader asks it to answer questions it excludes. It does not say how much the holdings are worth, how liquid they are, how many potential claims sit outside issued shares, what exercise would bring to the company, or whether staking rewards compensate for operating and market risk.

A better quarterly bridge needs four columns:

  1. native BERA and each BERA-equivalent position, with rewards and purchases separated;
  2. period-end issued shares and the exact sources of change;
  3. pre-funded warrants, other warrants, options and convertibles outside the KPI, with their relevant terms;
  4. cost, fair value, restriction status and accessible liquidity.

That scorecard would not abolish the 117 figure. It would place it where it belongs.

Greenlane’s quarter therefore offers a broader rule for public digital-asset treasuries. Accumulation is not accretion. Accretion is not valuation. Valuation is not liquidity. A company can improve one line while weakening the next. The investment case turns on the bridge between them, not on the largest number in the press release.

Sources