A company treads water on borrowed air, reaches for a drowning brand, and this afternoon told us how much air is left. Part two of two.
The bourbon is Oceans, because the rum ran out sometime this week and nobody at this end of the swamp could tell me when the next case was coming. Bourbon is better anyway, and this one got aged crossing an ocean, which makes it the only thing on this boat that improved from being out on the water. Rum is what you drink when you are pretending the water around you is blue. The water around me is the color of steeped tea and it moves when you look at it, and the things that live in it, the gators and the moccasins and the lawyers, do not pretend to be anything other than what they are. That is more than I can say for a press release.
I have six ReLiON group thirty ones dying in the belly of a twenty-six-year-old Moody, the slow lithium death that is quiet and then sudden, and I have spent this year deciding what replaces them, a hunt I ran down to the amp-hour and the mounting bracket in a separate teardown of every bank I am weighing for EOTI. Three candidates. ReLiON, who built the ones I have. Battle Born, whose parent I took apart at length in the first half of this story, The Battery Company That Bet Everything and Almost Lost It All, where I read a decade of its filings and decided it was a patient in a gown pretending to be a doctor. This is part two, and it picks up the afternoon that patient bought another one. And Dakota Lithium, the plucky Seattle outfit with the bass fishermen and the eleven-year warranty and the founder who crosses the North Atlantic on a boat called Polar Seal and posts it to the internet.
On the thirty-first of July, the third candidate stopped existing on its own. Dragonfly Energy, the company that owns Battle Born, the company I called a sinking ship in a piece I named after the temperature at which lithium iron phosphate gives up, bought Dakota Lithium. My shortlist of three now answers to two masters, and both of them are the same troubled ledger in Reno, Nevada. That ledger reported its second quarter numbers this afternoon, and the report is the reason this essay has an ending instead of a question mark. I could not have figured this one out stone cold sober, and I have been neither.
What follows I checked line by line, because the checking was the only thing that convinced me I had not dreamed it up in a blow.
Clean Liquidation
On the thirty first of July, Dragonfly Energy Holdings, ticker DFLI, completed the purchase of substantially all the operating assets of the Dakota Lithium brand. A bottle of bourbon makes a fine magnifying glass but it does not make that sentence any easier to understand, so let me pull it apart. Not the company. The assets. The name, the intellectual property, the inventory on the warehouse floor, the battery portfolio, the list of customers and dealers. An asset purchase is a particular creature. You back the truck to the loading dock, you carry out the things you chose, and you leave the corporate shell standing on the curb holding whatever debts you declined to lift. That distinction comes back to bite a Dakota Lithium owner before this is over.
Four million dollars. One million in cash, three million in Dragonfly stock counted at two dollars a share, which the release called a premium to where the stock had lately traded. Fifteen hundred thousand shares, locked up for twelve months so the recipients cannot spill them into a thin market on the first morning.
The 8-K, the filing Dragonfly made with the securities regulators that same Friday, and the contracts bolted to it, tell you more than the press release wanted to. The seller was not Clean Republic, the Seattle parent. The seller was an entity named Clean Liquidation, LLC, a California outfit acting as the assignee for the benefit of creditors of Clean Republic SODO. Clean Liquidation. They named it that on purpose. An assignment for the benefit of creditors is a formal insolvency proceeding, a cousin of bankruptcy under state law, the thing a company does when it is done and its assets need to be gathered up and sold off to pay what it owes. Dakota Lithium did not get acquired in the triumphant sense. It went under, its assets were handed to a liquidator, and Dragonfly bought them out of the wreck on an as is, where is basis, which is the language you use when you are buying a thing in whatever shape it happens to be in and asking for no promises about it.
The clock on the filing tells you how they wanted it read. The press release crossed the wire at half past seven on a Friday morning, the thirty first of July, and the filing hit the regulators five minutes later, all of it before the market opened. There is a whole grammar to when a company files things. Bad news goes out late on a Friday afternoon, after the bell, into the weekend, where it can sit and go quiet before anyone trades on it. Good news goes out early in the morning so the day has time to cheer. Dragonfly filed this one first thing Friday morning and dressed it in a release with the word strategic in the headline, timed so the trading day could greet it as a win. That is not the behavior of a company hiding a liquidation. That is the behavior of a company that bought one and wanted it seen as a purchase. The timing is the last brushstroke on the same paint job, and once you have read the contracts underneath it, you can see exactly what they were painting over.
And it was arranged before it happened. The asset purchase agreement carries the date of the thirty-first of July, and the general assignment that put the company into liquidation carries the same date, the thirty-first of July. The company assigned itself into the hands of a liquidator and sold its brand out the same door on the same day. You do not fall into liquidation and negotiate a sale of your best asset in twenty-four hours. This was a prepackaged funeral, the casket chosen and the plot dug before the patient was declared. Dragonfly, or its lenders, knew what they were buying and how, and the liquidation was the delivery vehicle, not the surprise.
And the stock did not go to the founders. The word acquisition hides this, but the contract says it in plain type. The agreement that issued the shares is titled a securities receipt agreement, and its own release clause states that in exchange for the shares, the recipient releases and discharges every claim, lien, and security interest it holds against the estate, and agrees to file the UCC termination statements that formally tear those liens up. Translate that. Dakota Lithium had a secured lender, someone who held liens on its inventory and its brand and was owed more than the corpse could cover. Dragonfly handed that lender 1,500,000 shares of its own stock, and the lender released its claims and walked away. The men who founded the company, the pilot and the chemist and the sailor, did not walk away with three million dollars in stock. Their lender did, and even the lender cannot sell it for twelve months. The founders walked away with whatever was left after the secured debt was satisfied, which in a liquidation is usually the thing that rhymes with nothing.
The twelve-month lockup has one crack in it, and the crack tells you how the lender is thinking. The shares are frozen against ordinary selling, no sales, no pledges, no hedges. But the agreement carves out a single exception. If Dragonfly itself is taken over, by tender offer or merger or any change of control, the locked shares are released to take part. The lender cannot sell into the open market for a year, but it made sure it could cash out if someone buys Dragonfly. That is a small, cold vote of confidence. The party that just took Dragonfly paper in a distressed deal wrote itself an escape hatch for the one outcome where that paper suddenly matters, and that outcome is Dragonfly being acquired.
Now the multiple, because the multiple is the whole confession. Dakota Lithium did roughly twelve million dollars in net revenue in 2025. Dragonfly paid four million for it, three quarters of that in its own stock rather than money. A third of one year of sales, mostly in paper. Healthy businesses do not trade hands at prices like that. Estates do, when the family needs the house emptied by the end of the month and the good silver goes to the first serious offer that clears the door.
And the release said plain type, which I respected, because it spared me the inference. Dakota Lithium’s revenue was significantly below prior years, they wrote, because of working capital limitations and inventory constraints that reduced product availability. From the language of the audited into the language of the dock, that reads. Dakota Lithium ran out of money to buy the batteries it needed to sell. Every furious forum post about a three-month backorder, every warranty claim left hanging from June to November, every angry Instagram comment that vanished twenty minutes after it went up, all of it was the visible waterline of a company that could not fund its own shelves. I had written exactly that in an assessment before I knew about the sale. The sale proved it in a sentence.
The life raft is already taking water
You would assume the buyer is the strong one. The shark eats the minnow, the natural order holds, the story runs the way stories are supposed to. So walk with me and the story of the daddy shark (boop do do dee doo).
Dragonfly reported its first quarter of 2026 on the fourteenth of May. Net sales of nine point seven million dollars, down more than twenty-seven percent from a year before. Gross margin of seventeen-point six percent, nearly thirty the year prior, meaning more of every battery dollar was eaten by the cost of building it than before. A net loss to shareholders of seven point seven million in a single quarter. That adjusted figure they call EBITDA, the number a company shows you when the real one is too upsetting to print, at negative four point six million. Charlie Munger hates that number with a passion. And the figure that cannot be adjusted because it is a fact and not an opinion, cash out the door from operations of eight point eight million dollars in ninety days.
Eight point eight million in cash gone in the first quarter alone. The whole company, every share added up, is worth around fifteen million dollars. So, it torched more than half its own market value in cash in ninety days just to keep the presses turning and the lights on. A gator that ate half its body weight in a season would not see the next one. That was the first quarter. The second quarter, the one that just reported, tells a more complicated story.
So how is it still breathing? This is where the story becomes either a survival miracle or a slow accident, and I hold a doctorate and read the filings twice and still cannot tell you which.
At the close of the third quarter of 2025, Dragonfly had three point eight million dollars in cash and a term loan carried at forty-five million, against a principal that, with piled-up interest, had grown toward seventy million. Put those side by side. Three point eight million in the bank. Seventy million owed. The auditors did the thing auditors do when the arithmetic looks like that. They raised substantial doubt about the company’s ability to continue as a going concern. A going concern is the accountant’s term for a business expected to remain in existence a year from now. Substantial doubt about it is the buttoned, lawyered, deniable way to write the word tombstone.
How it clawed out of that hole is the spine of part one, and I will not run the whole autopsy again here. The short version: in the back half of 2025 it sold about ninety million dollars of new stock, threw forty-five million at the loan, converted twenty-five million more into preferred shares held by the same lenders, got five million forgiven, and reverse split its shares ten into one in December to keep Nasdaq from delisting a stock that had fallen to seventy-one cents. The people who lent the money kept ending up owning more of the company, quarter after quarter, and the founder watched his own quarter of it dilute down toward nothing without selling a share. If you want that story told properly, it is the one I already told. What matters here is where it left the company standing when the Dakota deal arrived.
One more number and then I stop reading the autopsy. That loan has been amended seven times since 2022, a treadmill I walked through in part one. What matters now is the seventh amendment, signed the same day as the Dakota deal, because it is the one that pays for the acquisition.
This seventh amendment raised the interest rate to fourteen percent from twelve. It made all of that interest payable in kind through the end of the year, meaning not paid in cash but added to the pile of what is owed. And it pushed the date when the lenders will test the hard leverage covenant from March of 2027 out to September of 2027, another six months of not having to prove the ratios work. It also, in the small print, made Dragonfly pay the legal bills of the lenders’ own lawyers as a condition of the deal, and made Dragonfly sign a release forever discharging those lenders from any claim it might have had against them. The drowning company paid the lifeguard’s attorney and signed a waiver promising not to sue him. That is not a negotiation between equals. That is a company doing whatever it is told because the alternative is the bottom. Count the law firms feeding on this one carcass, by the way. Jeffer Mangels for the liquidator, Lowenstein Sandler for the buyer, Proskauer Rose for the lenders, Holland and Knight for the agent. Four sets of billable hours circling a four-million-dollar deal like moccasins around a dying deer, and every one of them gets paid before a single Dakota Lithium customer sees a nickel.
Fourteen percent, all of it payable in kind, on a principal that already ballooned once from unpaid interest. That is a debt that grows while you sleep, and the lenders are the ones who wrote the terms that way. The company is no longer servicing this loan. It is feeding it. Every quarter the number gets bigger and no cash leaves the building to make it smaller, which buys cash today against a larger reckoning in late 2027, the date they keep shoving down the calendar because they cannot yet reach it.
· · ·
Read the plumbing, not the banner
This is the bull shark (do do dee doo) of the brackish swamp water. The bull shark is the one that swims up rivers into water too fresh and too shallow for anything that size to belong in, and it survives there by being meaner and hungrier than the place it wandered into. On the last day of July, this particular bull shark, having burned through two-thirds of its cash in six months, its stock reverse split to keep its listing, its loan pried out of a coffin, swallowed a minnow that had run dry of the cash to stock its own shelves. The way it paid tells the story, if you are willing to read it slow and cold instead of swallowing the headline whole.
Dragonfly paid mostly in stock because it has no cash to spare, and we know it has none, because in the same breath, the same day, its lenders again changed the loan. They reset the minimum liquidity the company must hold to four million dollars through January and five million after that. They raised the rate and let all the interest accrue in kind rather than be paid. And they consented to the purchase, which a secured lender must do before you go spending on acquisitions while you owe it tens of millions. The company said these moves would preserve about one million dollars of near-term liquidity. You do not measure an acquisition’s benefit in the cash it lets you keep from your own loan payments unless the cash is the point.
The relief they won is like a reef and is shallower than it looks. The amendment did not remove the hard tests. It moved them to late 2027 and left a trap wired into the covenant waiting to trip. If the company’s liquidity is below fifteen million dollars at the end of any quarter once testing begins. Then a second covenant switches on, a fixed charge coverage ratio the company must then clear.
So, Dragonfly has been operating with cash in the low single-digit millions. A problem I wish I had. So the fifteen million is a bar it is nowhere near. So the harder covenant is not really a maybe, it is armed by default, waiting at the end of 2027 for a company that pays its interest in kind and burns cash to somehow show that it covers its fixed charges. They did not escape the tests. They bought eighteen months before the tests arrive, and set the trigger where they cannot help but trip it.
A company does not need its lender to let it stop paying interest in cash, and does not need its cash floor reset, to close an acquisition, unless the acquisition would otherwise push it through the floor. The one million of cash in the purchase price and the one million of liquidity preserved by the loan change are the same size, and that is not a coincidence. Dragonfly could afford this deal only because its lenders let it not afford the loan for two more quarters. The purchase and the forbearance are a single transaction wearing two hats.
Why do it? The chief executive, Denis Phares, who holds a doctorate of his own and whom I do not doubt believes every word, said the acquisition is expected to be accretive to adjusted EBITDA in the fourth quarter of 2026 and to support the goal of positive adjusted EBITDA in that quarter. There is the engine, whole and running. Dragonfly has told the market for a year that it will finally stop losing money, on an adjusted basis, by the last quarter of this year. It missed on volume because the recreational vehicle market that buys its batteries went soft. So it bought twelve million dollars of somebody else’s revenue, cheap, in paper, to bolt onto its own top line and make the fourth quarter close. You cannot grow into the number fast enough, so you buy the number.
The middleman is the final tell. The deal was facilitated, the release said, by an affiliate of a specialty financial advisory firm. Specialty financial advisory, on a small acquisition paid in locked up stock with covenant relief attached, is not the language of a growth deal run by the usual banks. It is the language of the corner of the street that handles distress, the number you call when the ordinary machinery will not touch the thing. Every pipe in this deal is stamped salvage, not conquest.
Whose promise is it now
Back to my dying batteries, and to the shell left standing on the curb. This is where the asset purchase stops being a financing detail and starts being a customer’s problem.
Dakota Lithium’s famous eleven-year warranty was always thinner than the billboard. I read the policy itself. Eleven years is defect liability. The real replacement promise is six, after which you get a repair if they judge it repairable or a coupon for sixty percent off a new one, which is not a warranty, it is a discount with better marketing. It does not transfer to the next owner. It voids above forty eight volts in series, which a real house bank on a real boat can pass. It excludes saltwater at the terminals without grease, which is the working definition of a boat. And it keeps the right to deny your claim if, on inspection, they decide the battery merely reached the end of its natural life, a clause loose enough to swallow almost anything.
That was the deal when the Seattle company stood behind it. But that company went into liquidation, and Dragonfly bought the brand and the customer list from a liquidator. Here the contract gets specific in a way that matters to anyone holding one of these batteries. The purchase agreement says Dragonfly assumed only the liabilities listed on a schedule, and everything not on that schedule, in the contract’s own words, it did not assume, including any liability arising from the seller’s business before the closing. A warranty written by the old company to a customer before the sale is exactly that, a liability from business before the closing. Unless someone bothered to write consumer warranties onto that schedule, they fall by default into the pile Dragonfly walked away from. And in a liquidation, where the buyer is picking assets out of a corpse, nobody volunteers to shoulder the unsecured promises made to retail customers, because doing so just hands value to the dead company’s other creditors.
The schedule itself was not filed. It was left out of the public record, which is allowed, so I cannot show you the line that would settle it. But I can tell you what the contract does to the customer whose warranty did not make the cut. It sends them back to the dead company. The old entity that indisputably owed those warranties is now a liquidating estate, and that estate has a claims bar date, a hard deadline written into the agreement, the twenty seventh of January, 2027. After that date, no one can file a claim against the estate at all. So Dakota Lithium’s owner whose warranty was not assumed has until late January to line up behind every other creditor of an insolvent company and ask for pennies, and then the door closes. That is not a warranty. That is a place in a line that is about to be painted shut.
Let’s be honest I am not saying Dragonfly will refuse an old Dakota Lithium warranty card. It probably will honor them, out of plain commercial sense, because refusing would torch a brand it just paid to own. But there is a difference between a company that must honor your warranty because a contract binds it and a company that may honor your warranty because it prefers the goodwill. The first is a right. The second is a favor. If you held an eleven-year Dakota Lithium warranty on Thursday, you were moved from the first category to the second one on Friday, and no letter came to tell you so.
The convergence
Sydney did ask me over coffee in the cockpit. While something big rolled the tea-colored water off the port bow, whether any of this changed what I would put in the boat. Fair question. It is the only one that matters under all the filings.
I wrote an entire essay pulling Dragonfly and Battle Born apart, concluding the company was trapped by its own timeline and using litigation and financial engineering to buy time. There is no way I’m not pulling for an American company to win but I won’t buy from a company that will possibly screw me over. In that other essay I called it The Melting Point. And the alternative I was weighing against Battle Born, the plucky Seattle brand with the fishermen, has now been swallowed by the very company that essay dissected. My shortlist folded into one corporate parent while I was busy refilling my bourbon. I can buy a Battle Born, or I can buy a Dakota Lithium, and as of this afternoon both of those choices report to the same strained ledger in Reno.
The chemistry does not care about any of it, and that is worth saying flat, with no booze in the equation. Lithium iron phosphate is lithium iron phosphate. The cells do not know the brand was sold. A Dakota Lithium battery bought today is the same stable, hard to light, long cycling chemistry it was the day before the deal, and still a better bet on a boat than the lead acid it replaces or the flammable cousins in the power tool aisle. The corporate distress does not make the battery dangerous. It makes the promise behind the battery uncertain, and those are different failure modes. One is whether the thing catches fire. The other is whether anyone answers the phone in year four.
For a liveaboard the phone call is not small when the power goes out. When a house bank quits its backed up by metropolitan power and is usually a bolt on. In a place where the next chandlery is a long haul and a headache away, the warranty is not paper, it is the line between a repair and a crisis. The battery bank is aa central hub to all the power consumers and power producers.
I run a windlass and a bow thruster that together can pull more than three hundred amps when the wind is up and the boat is not where I want it. The exact load that finds the weak spot in a battery pack. I need the company behind that pack to be a going concern in the accountant’s plain sense, still breathing in a year, still answering, still stocking the replacement. I need the product to be over built not built on a shoe string. This afternoon I learned a little more about whether the company now standing behind two of my three options is that kind of company.
What Thursday showed
Ok so the the bar was set, and set in public, so grade it against something firmer than the mood of the headline. Dragonfly had guidance in or around about thirteen point two million dollars in second quarter sales and an adjusted loss near one point nine million and had told the market for a year that cost cuts and a trucking ramp would carry it to break even by the fourth quarter.
Thursday afternoon, in a preliminary results release, the numbers were reported. On the surface they are a hit. Underneath, they are the confirmation of everything the contracts already said.
Sales came in at thirteen point two million, dead on the guide, and the adjusted loss was one point six million, better than promised, an improvement of three million over the first quarter. Congratulations to the team there. Read only that and it is a good quarter, and management wrote the release so that you would read only that. There were swamp puppies in the swamp.
The sales were down nineteen percent from a year ago, because both verticals shrank at once, the OEM business that sells into recreational vehicles down sixteen percent and the direct to consumer business down almost twenty-five. And the entire improvement in the loss came from cutting costs, not from selling more. They said so themselves, that the adjusted loss narrowed despite lower sales. This is a company getting thinner and calling it getting fitter. You can starve your way to a smaller loss for exactly as long as there is something left to cut.
The thing I saw was the cash, and the cash confirmed my thesis in a single line of the balance sheet. Cash fell from eighteen point three million at the start of the year to six point three million at the end of June. Twelve million dollars gone in six months, more than ten million of it straight out the operating door. Six point three million is what is left, and the covenant they signed with their lenders two weeks ago set the floor they must not cross at four million.
They are sitting two million dollars above the line they swore not to cross. That is likely why they paid for Dakota in stock. Maybe that is why they needed their lenders, in the same breath, to let them stop paying interest in cash and to lower that floor.
My essay did not have to guess at any of this. The cash balance printed it. And one number below it, the company’s book equity has gone negative, from a positive eleven and a half million at the start of the year to a deficit. They have spent through the value of the company and out the other side.
The cash burn did slow in the second quarter, and the year has been an exercise in spending the expensive currency. From what I think I see the cost cuts are real and the second quarter bled less than the first. That said a slower leak in a boat with six million dollars of freeboard left is not the same as a dry bilge, and the next quarter the management team guided lower, not higher, an adjusted loss of two point four million against this quarter’s one point six. Likely worse partly because of the cost of restarting the very brand they just bought. The thing they acquired to help the fourth quarter makes the third quarter loss bigger first.
The part where they sue the weather
And then there is the sentence in the release that that is why I started paying attention as I was making buy decision. It is the strangest thing in the whole document and because it is what drove me here to write in the first place . The direct-to-consumer business, the retail side, fell almost twenty five percent. I can imagine a lot of dynamics in consumer contraction, tariffs (which I’ve talked about previously) all having impact. The company itself gave two reasons for it. One was macroeconomic pressure on consumer demand (makes sense). The other, in their own words, was negative third-party online commentary regarding certain of their products, which they believe has adversely affected customer sentiment (tell me Will Prowse is a burr under your saddle without saying it). And then this. They have initiated legal proceedings to address this commentary (Yep they’re after a YouTuber).
You’ve got to be honest with this. A public company, reporting that its consumer sales fell by a quarter, has named online critics as a cause and gone to court over what they wrote. A company that is losing retail revenue is suing the people who talked about its products. Am I in jeopardy? I write about this company’s products and its finances. I write financial analysis, grounded in the company’s own filings and its own contracts, which is the most protected and the most boring kind of commentary there is out there. Apparently, analysis of a company’s products is the kind that now comes with a legal notice stapled to the earnings release. I am not looking for trouble but have to say it plainly rather than tiptoeing around it, because tiptoeing would be its own kind of repression and censorship, and I figure the numbers do the arguing for me.
The reported numbers make the argument on their own. I might be wrong, but this is how I read it after a few sips of bourbon. The recreational vehicle market that feeds Dragonfly’s largest segment fell more than fourteen percent this year, by the company’s own citation. Its cash halved. Its OEM sales, which no online reviewer touches, fell sixteen percent right alongside the retail decline.
Ok so the critics did not shrink the recreational vehicle market. The critics did not burn the cash. The critics did not write the seven loan amendments or set the covenant floor two million dollars under the current balance. A company whose troubles are this broad and this structural, reaching for online commentary as the thing to sue, is a company pointing at the weather to explain why the boat is low in the water. For all I know they may even win the case.
I’m trying to think of how suing Will Prowse is going to help the company bounce back. Winning the case will not put a dollar back in the account, and the account is where this story really lives. Losing the case will burn cash. Prosecuting the case will do worse and burn public sentiment like a V8 airboat burns gas. Looking at the water the swamp puppies agree and there is a few lawyers agreeing like the Anacondas do.
Here think about the size of the thing they are suing over, because the size is the whole point, and the company handed us the numbers to do it. The direct-to-consumer line fell from about five point nine million dollars a year ago to about four point five million this quarter. That is a drop of roughly one and a half million in the quarter, and if the rate held for a full year it would be somewhere under six million dollars. That is the entire consumer decline.
Now be honest and we’re guestimating but it isn’t the part caused by critics. It is the whole thing, macroeconomy and commentary and everything else combined. The company was careful not to split it. They named two causes, the soft consumer economy and the online commentary, and then declined to say how much of the fall belonged to each, which is a strange thing to leave out if you are building a lawsuit around one of them and you have the sales data to prove it. The law suit will have to show harm and the court will want to see that in dollars. With the numbers reported this way we can’t see the breakdown chose not to show it.
So, the honest ceiling on what online commentary could have cost this company is the full consumer decline, under six million dollars a year, and the real figure is some unknown slice of that, smaller by whatever share the economy did instead. Set even the ceiling against the rest of the release. The company burned twelve million dollars of cash in six months.
The single loan DragonFly has amended seven times carries tens of millions of cost. It seems they burned consumer credibility to swat a gnat that they could have co-opted. The consumer business at its absolute worst reading is a fraction of the cash that left the building while nobody wrote a word. DragonFly can sue over the slice and it can win, and the consumer segment can still decide whether the company lives. The number that decides survival or success is on the balance sheet and no reviewer has a login to it. The attribution itself is the part to hold at arm’s length. Negative commentary hurt our sales is an attribution the company asserted and did not quantify. It was offered by the only party that holds the data to test it, in the same release that announced the lawsuit. A reader is free to weigh it accordingly.
The arithmetic of hope
I keep returning to the phrase the accountants use, because it is illustrative of the whole story, more honest than any release, and nobody meant it as poetry, “Going concern.” A concern that is still going. A business the auditors will bet is alive in twelve months. Dragonfly had substantial doubt raised against its own going concern in 2025, it climbed out of the coffin on ninety million dollars of other people’s optimism. Then turned and bought a company that had quietly run out of the cash to stock its own shelves. Then paid in stock and in a promise from its lenders to let it skip interest for two quarters, all in service of one number in one future quarter.
It might still work, and I will keep that door open, because contempt is cheap and I have spent the year trying to spend the dearer currency, which is attention. As I said in a few things I’ve written. I can hold two ideas in my mind. I can hold a company in contempt for its current actions and still hope they succeed to mend their ways. Dragonfly has real technology, a dry electrode process that is truly its own. It has a trucking bet that is no longer only a promise, because deliveries under the Stevens Transport order started shipping this quarter, and the company expects that trucking revenue to more than double next quarter. That is the one clean piece of good news in the release.
Doubling a small number leaves a small number, and the company still needs to nearly double its total sales, to a seventy million dollar annual pace, just to stop losing money on an adjusted basis. If the trucking ramp holds and the fleets behind Stevens convert and the cost structure keeps its new discipline, two sinking companies lashed together might, against the Altman score and the six million dollars of cash, float. Stranger things have happened. I am living on one of them.
I see though the physics of a balance sheet that are no gentler than the physics of a battery. You can adjust EBITDA (which is why Charline Munger hates it). They did, and they beat the number doing it. You cannot adjust cash. The cash, last Thursday, at the bottom of the page they would rather you skip, says six million dollars and change. That is two million above the floor they swore to their lenders they would hold. Evaluated against a machine that drank twelve million in six months. Everything else in the release is a story told over the top of that number.
The batteries in the boat are still dying at their own unhurried pace. I have not decided what replaces them. The bourbon is Oceans because the rum ran out, and that is the better outcome, and nobody could tell me when the next case of either was coming.
Nobody ever can. That is the whole business, and it is also the whole swamp.
From the deck of EOTI I can see reality just fine. The water is brown, the shark is upriver where it does not belong, the lawyers are circling something in the shallows, and two sinking companies are holding onto each other and calling it a rescue. I can see all of it clear as noon. It just is not much fun to look at.
Sources and Notes
Every factual claim in this essay is drawn from the primary and secondary sources below. The financial and deal terms come from Dragonfly Energy’s own filings with the U.S. Securities and Exchange Commission, including the three transaction contracts filed as exhibits to the acquisition report, each of which was read in full rather than taken from a summary. The second quarter figures come from the company’s preliminary results release of 6 August 2026, which the company states is subject to adjustment and which precedes the full Form 10-Q it said it would file later in the month. All filings were retrieved in early August 2026. The one figure that still cannot be verified from the public record, the schedule listing which liabilities Dragonfly assumed, is identified as such where it appears. Nothing here rests on memory.
The acquisition filing and its three contracts
[1] Dragonfly Energy Holdings Corp., Current Report on Form 8-K, filed with the SEC 31 July 2026, event date 31 July 2026. The cover report. It describes the transaction and attaches the three operative contracts as Exhibits 10.1, 10.2 and 10.3, each read in full for this essay and cited separately below. Source for the overall frame: the four million dollar price, the split of one million in cash and 1,500,000 shares at two dollars per share, the seller being Clean Liquidation, LLC as assignee for the benefit of creditors of Clean Republic SODO, LLC, and the Section 4(a)(2) exemption for the shares. Signed by Denis Phares as Chief Executive Officer, Interim Chief Financial Officer and President. On the filing time, see [1a] below.
https://www.sec.gov/Archives/edgar/data/1847986/000149315226035543/form8-k.htm
[1a] Filing time, on which the essay’s reading of the intended framing rests. The transaction was announced and filed the morning of Friday 31 July 2026, before the market opened. The press release crossed GlobeNewswire at 07:30 Eastern. The SEC accepted the Form 8-K at 07:35 Eastern and it posted to EDGAR at 07:36 Eastern, per the EDGAR acceptance and availability timestamps on the filing. A pre-open Friday-morning release, as opposed to a post-close Friday-afternoon filing, is consistent with a company presenting news it wants the trading day to absorb, not bury. The 07:30 press-release timestamp is from the GlobeNewswire dateline at [5]. The 07:35 and 07:36 timestamps are from the EDGAR filing header for the report at [1].
[2] Asset Purchase Agreement, dated 31 July 2026, by and among Clean Liquidation, LLC as assignee for the benefit of creditors of Clean Republic SODO, LLC, Dragonfly Energy Corp. as buyer, and Dragonfly Energy Holdings Corp. as parent. Filed as Exhibit 10.1. Governed by California law, venue in Los Angeles County, with a jury-trial waiver. Source for: the as is and where is sale with no seller warranties, Section 5.3(c) in which the buyer acknowledges the risk of undisclosed product or patent defects, the same-day general assignment that put the seller into liquidation on the closing date, the Section 3.1 assumption of only the liabilities listed on Schedule 3.1, the Section 3.2 default that all other pre-closing liabilities are excluded, and the 27 January 2027 claims bar date for the liquidating estate. Schedule 3.1, which would show whether consumer warranties were assumed, was not included in the public filing. Seller’s counsel was Jeffer Mangels Butler and Mitchell. Buyer’s counsel was Lowenstein Sandler.
https://www.sec.gov/Archives/edgar/data/1847986/000149315226035543/ex10-1.htm
[3] Securities Receipt Agreement, dated 31 July 2026, between Dragonfly Energy Holdings Corp. and the recipients. Filed as Exhibit 10.2. Governed by New York law, venue in Manhattan, with a jury-trial waiver. Source for the finding that the 1,500,000 shares went to a secured lender rather than to the founders: Section 4.2, titled Release of Secured Claim, issues the shares in exchange for the recipient’s release of all liens and security interests against the estate and the filing of UCC-3 termination statements. The recipient representations, including accredited-investor status and a carve-out for a multi-managed investment vehicle, indicate an institutional lender. Section 4.3 sets the twelve-month lock-up and, in subsection (c), carves out a single exception permitting the shares to participate in a change of control of Dragonfly during the lock-up.
https://www.sec.gov/Archives/edgar/data/1847986/000149315226035543/ex10-2.htm
[4] Seventh Amendment to Term Loan, Guarantee and Security Agreement, dated 31 July 2026, among Dragonfly Energy Corp. as borrower, Dragonfly Energy Holdings Corp. and Battle Born Battery Products, LLC as guarantors, the lenders, and Alter Domus (US) LLC as agent. Filed as Exhibit 10.3. Governed by New York law. Source for the loan terms cited in this essay. The interest rate was raised to 14.0 percent from 12.0 percent and made entirely payable in kind during the PIK period from 31 July through 31 December 2026 (Sections 3.1.1 and 3.1.2). The maximum senior leverage ratio test was deferred to the quarter ending 30 September 2027 at a 3.00-to-1.00 ceiling (Section 3.3). The minimum liquidity covenant was reset to four million dollars from 31 August 2026 through 31 January 2027 and five million dollars thereafter (Section 3.4). A fixed charge coverage ratio switches on if quarter-end liquidity falls below fifteen million dollars once testing begins (Section 3.5). The amendment’s effectiveness was conditioned on consummation of the asset purchase (Section 4.2). The borrower pays the lenders’ and agent’s legal fees (Sections 4.3 and 4.4), and the borrower releases its claims against the lenders (Section 13). The recital lists the loan’s full amendment history: original agreement 7 October 2022, then amendments dated 28 June 2024, 29 July 2024, 30 September 2024, 31 December 2024, 26 February 2025, 20 October 2025, and this seventh amendment 31 July 2026. Lenders’ counsel was Proskauer Rose. Agent’s counsel was Holland and Knight.
https://www.sec.gov/Archives/edgar/data/1847986/000149315226035543/ex10-3.htm
Dragonfly financial results and filings
[5] Dragonfly Energy, press release, Dragonfly Energy Announces Strategic Acquisition of Dakota Lithium Assets, Expanding Into New Markets, GlobeNewswire, 31 July 2026. Source for the multi-brand statement that Dakota Lithium will run alongside Battle Born, the roughly twelve million dollars in Dakota Lithium 2025 net revenue described as significantly below prior years, the marine and outdoor and powersports and golf cart market description, the Q4 2026 EBITDA accretion and positive adjusted EBITDA goal, the Phares quotation, the roughly one million dollars of near-term liquidity the company said the loan amendment would preserve, and the disclosure that the transaction was facilitated by an affiliate of Resolution Financial Advisors LLC.
[5a] Dragonfly Energy, press release, Dragonfly Energy Reports Second Quarter 2026 Preliminary Results, 6 August 2026. The primary source for the results section of this essay and for every second quarter figure in it. Net sales of 13.2 million dollars (13.159 million), down 19.0 percent year over year, with OEM of 8.4 million (down 16.1 percent) and DTC of 4.5 million (down 24.7 percent). Gross margin of 33.0 percent, including a 1.1 million dollar tariff refund benefit in cost of sales. Net loss attributable to common shareholders of 5.5 million, or 0.43 dollars per diluted share. Adjusted EBITDA of negative 1.6 million, a 3.0 million sequential improvement and a 0.6 million improvement year over year despite lower sales. Cash and cash equivalents of 6.280 million at 30 June 2026, down from 18.270 million at 31 December 2025. Net cash used in operating activities of 10.641 million for the six months. Stockholders’ equity of negative 184 thousand, down from positive 11.530 million at year end. Q3 2026 guidance of about 13.5 million in net sales and adjusted EBITDA of about negative 2.4 million, the latter reflecting vacated-facility costs and the cost of restarting Dakota Lithium operations. The statement that heavy-duty trucking revenue is expected to more than double sequentially in Q3 and that deliveries under the Stevens Transport order commenced during the quarter. The RV shipment decline of 14.2 percent through midyear. And the DTC attribution and litigation statement quoted in the essay, that the decline was due in part to negative third-party online commentary regarding certain products and that the company has initiated legal proceedings to address it. These figures are the company’s own preliminary results and are expressly subject to adjustment. Dragonfly stated that the full Quarterly Report on Form 10-Q for the quarter ended 30 June 2026 would be filed later in August 2026, and as of writing it had not yet been filed.
[6] Dragonfly Energy, First Quarter 2026 results, press release and related Form 8-K, 14 May 2026. Source for the Q1 2026 figures. Net sales of 9.7 million dollars, down 27.3 percent year over year. OEM sales of 5.8 million and direct-to-consumer sales of 3.7 million. Gross margin of 17.6 percent versus 29.4 percent in the prior-year quarter. Net loss of 6.6 million and net loss attributable to common shareholders of 7.7 million, or 0.64 dollars per share. Adjusted EBITDA of negative 4.6 million. Operating cash outflow of 8.8 million. Operating expenses of 7.4 million down from 9.8 million. And the Stevens Transport purchase order valued at more than three million dollars covering nearly five hundred trucks. This is also the source for Q2 2026 guidance of about 13.2 million in net sales and adjusted EBITDA of about negative 1.9 million.
[7] Dragonfly Energy, Quarterly Report on Form 10-Q for the period ended 30 September 2025, and the accompanying Q3 2025 disclosures. Source for the going concern history. Cash of 3.838 million dollars at quarter end. Term loan carried at 45.423 million against principal and accrued PIK of roughly 69.974 million. Q3 interest expense of 6.409 million. Q3 net loss of 11.070 million. Nine-month net loss of 24.901 million. The substantial doubt about going concern subsequently described by management as mitigated. And the subsequent-event financing and restructuring detail, namely net equity proceeds of 26.925 million on 6 October and 51.928 million on 17 October 2025, a 45 million dollar term loan prepayment, 5 million in principal cancellation, and conversion of 25 million in principal into Series B Preferred stock on 4 November 2025. The loan-amendment terms cited in this essay are drawn from the Seventh Amendment at [4] rather than from this quarterly, which predates it.
[8] Dragonfly Energy, press release, Dragonfly Energy to Report Second Quarter 2026 Financial and Operational Results on August 6, 2026, GlobeNewswire, 30 July 2026. Source for the earnings date and time: results after market close on Thursday 6 August 2026, with a conference call at 4:30 p.m. Eastern.
[9] Dragonfly Energy, press release, Dragonfly Energy Announces 1-For-10 Reverse Stock Split, GlobeNewswire, 16 December 2025, effective 18 December 2025. Source for the ten-to-one reverse split, the reduction in shares outstanding from approximately 120.8 million to approximately 12.1 million, and the stated purpose of maintaining Nasdaq listing compliance. The roughly 44 percent premarket decline on the announcement and the pre-split close near 71 cents are reported in contemporaneous market coverage (Stocktwits and RTTNews, 16 December 2025) and in the roughly 90 million dollars of 2025 equity raises summarized in the CEO year-end shareholder letter (Investing.com, 8 January 2026).
Market data and analyst context
[10] Share price and market capitalization at the time of writing: DFLI traded in the range of roughly 1.14 to 1.29 dollars with a market capitalization near 14.5 million dollars in the days around 31 July 2026, against a 52-week range whose high reflects the pre-reverse-split period (Kraken and MarketBeat quote pages, retrieved early August 2026). The 200-day moving average of 4.82 dollars referenced against the Q1 report window is from StockTitan’s May 2026 coverage at [6].
https://www.marketbeat.com/stocks/NASDAQ/DFLI
[11] Cash runway and dilution context: the characterization of less than one year of cash runway, free cash flow of roughly negative 32 million dollars, share dilution of more than eleven times over the prior year, and a market capitalization under 100 million is from Simply Wall St’s DFLI analysis (retrieved via search, dated approximately one month before writing). The Altman Z-Score of negative 1.03 signaling elevated insolvency risk is from AInvest’s December 2025 coverage of the reverse split. These are third-party analytical readings and are labeled as such in the text. The underlying financial inputs trace to the filings at [6] and [7].
https://simplywall.st/stocks/us/capital-goods/nasdaq-dfli/dragonfly-energy-holdings
Dakota Lithium background and the warranty
[12] The Dakota Lithium corporate structure (the Clean Republic and Clean Republic SODO entities, the Dakota Lithium Materials trade name, the Seattle headquarters and Grand Forks operations, the lithium iron phosphate chemistry, and the University of North Dakota cathode research through the Energy and Environmental Research Center) is documented in the author’s prior assessment of the company and its underlying sources, including a North Dakota Clean Sustainable Energy Authority grant proposal and the Better Business Bureau business profile. The cell sourcing characterization reflects that assessment and trade-forum discussion.
[13] The Dakota Lithium warranty terms (the eleven-year defect warranty against a six-year guaranteed lifespan, the repair-or-sixty-percent-discount structure in years seven through eleven, the non-transferability, the 48-volt series cap, the saltwater terminal and normal-end-of-life exclusions, and the customer-paid freight) are from the published Dakota Lithium warranty policy document as read for the author’s prior assessment. Whether those obligations transferred to Dragonfly is governed by the Asset Purchase Agreement at [2]: excluded by default under Section 3.2 unless listed on the withheld Schedule 3.1, with the seller now a liquidating estate subject to the 27 January 2027 claims bar date.
A note on method The deal terms in this essay were checked against the three transaction contracts filed with the acquisition report, not against news summaries. All three, the Asset Purchase Agreement, the Securities Receipt Agreement, and the Seventh Amendment to the loan, were read in full. The press release used the softer language of a strategic acquisition and did not disclose that the seller was a liquidating estate, that the shares went to a secured lender rather than the founders, or that the loan interest had been switched to payment in kind. Those facts come from reading the contracts. The one thing that remains truly unverifiable is the contents of Schedule 3.1 of the Asset Purchase Agreement, which lists the liabilities Dragonfly assumed and which was lawfully omitted from the public filing. Every conclusion in this essay that depends on that schedule. That is the whole business, and it is also the whole swamp.
From the deck of EOTI I can see reality just fine. The water is brown, the bull shark is upriver where it does not belong, the lawyers are circling something in the shallows, and two sinking companies are holding onto each other and calling it a rescue. I hope they can start bailing soon. I can see all of it clear as noon. It just is not much fun to look at.
About the Author
