AI 对冲基金 · 大师信号(存档)

This is the cleanest rate-of-change setup I've seen in a while — operating margin has gone 0.47, 0.49, 0.50, 0.51, 0.54, 0.58, 0.60 and net margin ripped from 0.43 to 0.56 in four quarters, with revenue growth troughing at 0.20 in 2025/Q4 and re-accelerating to 0.35 and 0.36. EPS is where the real violence is: 69.55, 76.64, 87.10, 93.54, 110.44, 136.27 — the last two prints are +59% and +78% year-over-year, and the sequential is still building, not fading. At 32x trailing the market is pricing a good compounder, but trailing EPS of ~427 is already stale against a 136 quarterly run-rate — that's roughly 25x on where the business actually is, and the multiple is being paid down faster than it's being re-rated. Leverage is going the right way too, D/E from 0.56 to 0.45 with a 2.46 current ratio, so there's no balance-sheet landmine under the inflection — margin expansion plus accelerating top line plus a multiple that hasn't caught the run-rate is when you size up, not nibble.

This is a textbook fast grower: revenue up 36% and quarterly EPS marching from 62.68 to 69.55 to 76.64 to 87.10 to 93.54 to 110.44 to 136.27 — nearly a double in two years, with no stumble in the sequence. Earnings are growing faster than sales because gross margin went from 0.56 to 0.68 and net margin from 0.40 to 0.56, so the growth is real operating leverage, not accounting decoration. At a P/E of 32 against earnings compounding 40-50%, the PEG is roughly 0.7 — you're paying less than the growth rate, which is exactly the situation I hunt for. The balance sheet backs it up: debt/equity down from 0.56 to 0.45, current ratio 2.46, ROE 40%, book value per share compounding 24%, and free cash flow that actually shows up. The one thing I'd keep an eye on is that this is a cyclical business and margins this wide invite competition and a bad year eventually — but a strong balance sheet lets a story like this survive one, and today I know what I own and I'm not overpaying for it.

Invert it: I look for the rot and I can't find much — gross margin went from 56% to 68%, operating margin to 60%, ROE climbed 29% to 40%, debt/equity fell 0.56 to 0.45, and current ratio sits near 2.5. That is a business getting better while getting bigger, book value compounding 24% a year with revenue up 36% and no accounting contortions to explain it. The one honest blemish is capital intensity — free cash flow per share runs barely half of earnings because the fabs eat the money, so this compounds on reinvestment rather than on cash handed back, and semiconductors have always been a cyclical business that punishes people who extrapolate the peak. At 32 times earnings against that growth and those returns, you are not being asked to believe anything stupid; I'd rather own this than most things trading cheaper.

This is the kind of business you can understand without a slide rule: it makes the chips everybody else designs, and the numbers say the customers have nowhere else to go — gross margin has walked from 58% to 68%, operating margin from 47% to 60%, and returns on equity from 29% to 40% while revenue grew a third. That combination of rising price with rising volume is what pricing power looks like on paper, and management has done it while cutting debt-to-equity from 0.56 to 0.45, holding a 2.5x current ratio, and compounding book value 24% a year — over half of reported earnings still landed as free cash flow despite the enormous appetite for new plants. At 32 times earnings you are not stealing it, and I'd note the foundry business eats capital and has always been cyclical, so today's record margins are not a birthright. But paying a fair price for a business earning 40% on equity and widening its lead beats a bargain price for a mediocre one, and I'd be content to own this for ten years.

The enterprise itself is beyond reproach: a current ratio of 2.46, debt to equity of 0.45 and falling from 0.56, positive earnings in every period shown, and book value per share compounding at 24 percent — this is a strong and honestly capitalized business, not a speculation in its operations. The speculation lies entirely in the price. At 32 times earnings and, by the arithmetic of a 0.40 return on equity against that multiple, roughly twelve to thirteen times book value, the buyer pays more than four hundred times the product Graham would tolerate at 22.5, and surrenders every particle of margin of safety to a forecast of continued 36 percent revenue growth and 60 percent operating margins. Expanding margins and rising ROE are precisely the conditions that flatter a cyclical semiconductor maker at the top of its cycle; I will not capitalize them into perpetuity. A fine company bought at this price is not a fine investment — the defensive investor should wait, and record the overvaluation as the adverse fact it is.

Thirty-four percent on equity, sixty-eight percent gross margins, and forty percent net margins held steady across every period shown — that is not luck, that is a toll bridge, and revenue growth accelerated from 12% to 18% while leverage fell from 0.82 to 0.71. Book value per share compounding at 21% with EPS up 31% to $18.01 means the capital is being retained and reinvested at rates almost nobody else can touch, so paying 27.6 times earnings for it is a fair price, not a foolish one. The one thing that makes me sit up: free cash flow per share went backwards, $9.64 to $9.02, while earnings ran away — the AI capex is swallowing the cash, and accounting profit that doesn't show up as cash is a promise, not a fact. I'd own it and I'd watch that line like a hawk; if FCF per share is still flat two years from now, the earnings are an illusion and I'd be wrong.

This is a fast grower hiding in a mega-cap suit: revenue growth actually accelerated from 15% to 18%, EPS went from $13.71 to $18.01, and every quarter is up year over year. At 27.6 times earnings against 31% reported earnings growth you get a PEG under 1 — though I'd haircut that, because the Q2 net margin of 47% sits right on top of the operating margin, which tells me something non-operating padded that quarter; on the cleaner operating trend you're paying roughly a fair price for 20%-plus growth, not a steal. The balance sheet is the kind I like — debt to equity coming down from 0.80 to 0.71, book value compounding 21% a year, 47% operating margins that haven't budged. The one thing that nags me is that free cash flow per share went backward ($9.64 to $9.02) while earnings jumped 31%; the cash is going into the ground as capex, and if that gap doesn't close within a year or two, the story changes.

The inflection already happened and it's holding: revenue growth troughed at 12-13% in the middle of FY25 and has printed 17-18% for four straight quarters, while EPS went from 13.71 to 18.01 — up 31% against an 18% top line, which is operating leverage doing real work. At 27.6x that's under 1x PEG on accelerating numbers; the multiple says the market is still underwriting a mid-teens grinder, and the trend disagrees. What keeps me from going max size is the rate-of-change underneath: gross margin has slipped 69 to 67-68, operating margin has walked down 49 to 45 across the FY26 quarters, and FCF/share actually fell to 9.02 from 9.64 — the capex cycle is eating the cash even as EPS rips, and part of the Q2 net margin pop at 0.47 looks non-operating. Balance sheet is a non-issue with D/E improving to 0.71 and book value compounding 21%, so there's no blow-up risk here — this is a real long, sized well but not the trade I bet the fund on, and I want to see whether the margin drift is investment or erosion.

This is unquestionably a wonderful business — a 34% return on equity held for ten straight quarters, 68% gross margins, 47% operating margins, and book value per share compounding at 21% a year while revenue still grows 18%. The balance sheet is sound and getting sounder, with debt to equity down from 0.80 to 0.71, though a 1.23 current ratio is tighter than I like. My hesitation is the cash: earnings per share climbed from $13.71 to $18.01 while free cash flow per share slipped from $9.64 to $9.02, so owners now get fifty cents of spendable cash per dollar of reported profit — at 27.6 times earnings that works out to roughly 55 times free cash flow, a price that assumes the enormous capital now being poured in earns the same returns the old software business did. I'd happily own this at a price that let me be wrong about that; at today's figure I'll admire it from the sidelines and wait for the capital spending to prove itself in the cash flow statement.

The enterprise is beyond reproach — earnings positive and rising in every period shown, EPS advancing from 13.71 to 18.01, return on equity near 34 percent, and book value per share compounding at 21 percent — but I am asked to buy demonstrated value, not admire it. At 27.61 times earnings and roughly 8.3 times a book value of 59.58 per share, the product of the two multipliers is near 230 against the 22.5 I permit; there is no margin of safety here, only a capitalization of expectations. The balance sheet, moreover, is merely adequate rather than strong: a current ratio of 1.23 sits below the 1.5 I require, and debt to equity of 0.71 leaves less cushion than the business's reputation suggests. A fine company bought at such a price is still a speculation, and I must call the price, not the company, the defect.

Revenue growth bottomed at 56% in FY26 Q2 and has gone 62, 73, 85, 106 straight up since — that's not a decelerating hypergrowth story, that's a second inflection, and operating margin rebuilt from 49% to 66% right alongside it while ROE climbed back to 1.17. EPS went 0.78 to 2.48 in five quarters and the market is paying 27.6x for it; a sub-30 multiple on a business reaccelerating through 100% top-line growth says the tape believes the old deceleration narrative, and the numbers flatly disagree — that's the asymmetry I want, and I'd be big. Balance sheet doesn't scare me either: D/E 0.40, current ratio 4.6, book value compounding 74%, so there's no leverage landmine under this. The one thing I'm marking: FCF/share fell to 0.89 from 2.02 and EPS was nearly flat sequentially into the latest quarter, so if working capital keeps eating the cash and that flat print becomes two prints, the trade is over and I'm out — but you size to the inflection you see today, not the one you fear.

This is a fast grower and it isn't a close call: revenue up 106% on the latest twelve months, earnings per share more than doubling year over year ($2.48 against $1.10), operating margin climbing from 49% to 66%, and return on equity above 100%. Yet the stock changes hands at 27.6 times earnings — a P/E less than half the growth rate, giving a PEG around 0.4, which is the kind of math that made money for me at Magellan. The balance sheet holds up the story too: debt/equity of 0.40, current ratio of 4.6, and book value per share compounding 74% a year, so a bad quarter wouldn't sink it. My one hesitation is that free cash flow per share ($0.89) is running well behind reported EPS this quarter — cash should follow earnings, and if it keeps lagging I'd want to know why before adding more.

A business earning over 100% on equity for ten consecutive periods, with 75% gross margins, 66% operating margins, and book value per share compounding at 74%, is not a puzzle — it is the best set of numbers I have looked at in a long while, and the balance sheet is unlevered at 0.40 debt-to-equity with a 4.6 current ratio. Inverting: the one honest crack is cash conversion — the latest quarter turned $2.48 of earnings into $0.89 of free cash flow against roughly 80% conversion in fiscal 2026, and debt-to-equity crept from 0.31 to 0.40; I read that as working capital built ahead of a revenue line that doubled, not as dishonesty, but I would want it to reverse. Nobody earns 117% on equity forever — capital floods toward returns like that — so the real risk here is paying 27.6 times what may prove to be cycle-peak earnings. But 27.6 times against 106% revenue growth and 62% net margins is not a price that requires believing anything stupid; it is a great business at a fair price, and I will not refuse that merely because the numbers embarrass me.

There is no arguing with the economics here: 75% gross margins, 66% operating margins, better than 100% return on equity, a current ratio of 4.6 and only modest debt at 0.40 — book value per share has compounded at 74% and revenue doubled again this period. At roughly 28 times trailing earnings of about $8 a share, the price is not what I'd call demanding for a business of that quality; it is the earnings themselves I cannot underwrite for ten years. Two things give me pause: free cash flow of roughly $5.28 a share against $8.00 of earnings means a third of the profit is sitting in working capital rather than the bank, and in one quarter net margin exceeded operating margin, which tells me gains below the operating line are doing some of the work. Semiconductors sit outside my circle — I can read that this is a wonderful business today, but I cannot tell you who is selling the picks and shovels in 2036, so I'll admire it and keep my thumb off the scale.

The balance sheet is beyond reproach — a current ratio of 4.59 and debt to equity of 0.40 leave nothing to criticize, and earnings have been positive and growing throughout the record shown. But I buy securities, not enterprises, and at a P/E of 27.60 on $2.48 of trailing earnings the shares fetch roughly $68 against book value of $9.50 — better than seven times book. The product of the two multipliers approaches 200 where I would insist on 22.5; a return on equity of 117 percent and revenue doubling in a year are precisely the conditions that invite competition and that the market mistakes for permanence. The record here spans a scant two and a half years of extraordinary results, and the price capitalizes their continuation rather than their demonstration — that is a speculation, however fine the company, and it affords no margin of safety whatever.

Revenue growth has accelerated eight quarters in a row — 30% to 93% on a trailing basis — and that almost never happens at this revenue scale; the second derivative is still positive, which is the only thing I actually care about. Operating margin went from 1% to 47% and TTM EPS compounded 0.20 to 0.44 in four quarters, so the incrementals are dropping through at 80%+ and the earnings trajectory is building, not fading. Yes, 143x is a rich number and the market clearly believes something, but it's discounting a business growing 50-60%, not one accelerating through 90% with 55% net margins, a 7.2x current ratio and 0.18 debt/equity — there's no leverage here to kill me. I'd be sized up on this, with the discipline that the day revenue growth stops accelerating the multiple de-rates hard and I'm out, not averaging down.

This is plainly a fine business — 85% gross margins, operating margins that have climbed from 1% to 47% in two years, revenue nearly doubling, returns on equity marching from 11% to 38%, almost no debt at 0.18 to equity, a current ratio above 7, and book value per share compounding at 41%. Management has funded that growth without leverage and turned it into real cash, and I tip my hat to them. But at $402 billion against 44 cents of trailing earnings and 50 cents of free cash flow per share, I am asked to pay roughly 143 times earnings and 15 times book — a price that already assumes a decade of flawless execution, and I notice net margin exceeding operating margin, which tells me some of those earnings came from below the operating line rather than from the business itself. I would happily own this enterprise; I would not pay this price for it, and I do not short a wonderful company merely because it is dear — so I sit on my hands and wait for the fat pitch.

The business itself is close to unimpeachable: 85% gross margins, operating margin up from 1% to 47% in eight quarters, ROE from 11% to 38%, revenue accelerating to 93%, book value per share compounding at 41%, essentially no debt and a 7x current ratio. That is a rare machine, and I will not insult it by calling it mediocre. But at 143 times earnings and roughly 125 times free cash flow — a 0.8% owner's yield against a 15x book multiple — the price already assumes this company keeps nearly doubling for a decade without a stumble, a competitor, or a defense budget that says no. I also note net margin of 55% exceeds the 47% operating margin, which means a good chunk of the reported profit is interest and tax artifacts rather than the business. Wonderful company, ludicrous price; I keep my hands in my pockets and go read something else.

This is a fast grower, no argument — revenue growth went from 30% to 93% over eight quarters, operating margin went from 16% to 47%, EPS keeps marching up, and the balance sheet is a fortress with debt/equity of 0.18 and a current ratio above 7. The trouble is the price tag: 142.93 times earnings against roughly 93% growth is a PEG near 1.5, and at $402 billion of market cap on maybe $2.8 billion of earnings you're paying about 78 times sales — that's a hot-story price even for a hot story. I also notice net margin (55%) running above operating margin (47%), which tells me a chunk of the profit is coming from below the operating line, not from selling software. Wonderful company, fully priced; the earnings would have to keep galloping for years just to justify today's quote, and elephants this size rarely gallop that long — I'd know what I own, admire it, and wait for a bad quarter.

The enterprise itself is unobjectionable to a defensive man: a current ratio of 7.23, debt to equity of 0.18, and earnings positive and rising through every period shown — this is no speculation on a business that has yet to prove it can earn. The price, however, is another matter entirely. At 142.93 times trailing earnings of $0.44, the shares fetch roughly $63 against book value of $4.11 — some 15 times book — so that the product of the two multiples exceeds 2,000 where I would insist on 22.5. That is not a margin of safety; it is a margin of hope, and the buyer at such a figure is paying today for a decade of growth that must arrive precisely on schedule and then keep arriving. I would let this one pass and note that the balance sheet's virtues, real as they are, cannot redeem a purchase price that leaves no room whatever for the future to disappoint.

Revenue growth has gone 9% → 13% → 14% → 17% → 20% across the last five quarters while operating margin ran 10% → 12% → 13% → 14% — that's top line accelerating and margins inflecting up at the same time, which is the single best setup in this business. EPS momentum is violent (Q2 5.81 vs 1.68 a year ago), and while the 31% net margin is obviously flattered by something non-operating, the 14% operating margin isn't — the core is genuinely widening. A 20x headline multiple (call it high-20s scrubbing the gain) on a name whose growth rate is doubling says the market is still underwriting the old 10-12% grower, and the trend flatly disagrees. The thing that keeps me from going max size is the cash: FCF/share has gone negative and the current ratio is 1.03 against D/E near 1.0, so they're funding this acceleration out of the balance sheet — if revenue growth rolls over before that capex earns, this turns fast, and I'd be out, not averaging down.

This is a fast grower that most people still file under 'expensive' without doing the arithmetic: revenue growth has accelerated from 9% to 20%, operating margin has climbed from 11% to 14%, gross margin from 47% to 52%, and full-year EPS went 5.49 to 7.20 — better than 30% — while the stock trades at 20.66 times earnings. A P/E of 21 against 30% growth is a PEG around 0.7, and that's the kind of math that made me money; you're paying less for the growth than the growth is worth. Two things keep me from pounding the table: the latest quarter shows a 31% net margin against a 14% operating margin, which tells me something below the operating line — a one-time gain — is flattering that EPS and the P/E with it, and free cash flow per share has gone negative while they spend heavily, so the reported earnings aren't all cash yet. Debt equal to equity and a current ratio barely over 1 is adequate rather than a fortress, so I own it on the operating story — more revenue turning into more operating profit — and I'd sell the day that margin expansion stalls.

This is plainly a fine business — returns on equity running 22-31%, gross margins widening from 47% to 52%, operating margins from 11% to 14%, revenue reaccelerating to 20%, and book value per share compounding near 40%. But I don't pay for reported earnings, I pay for owner earnings, and the latest twelve-month net margin of 31% against a 13% average tells me something one-time is dressing up that $5.81 of EPS; normalize it toward the historical margin and the 20.7x multiple is closer to 45-50x, while free cash flow per share has turned negative two periods running as the company pours money into capacity. A current ratio of 1.03 and debt roughly equal to equity leave little slack if that spending doesn't earn its keep. A wonderful company, no question — but at a price that requires the buildout to pay off exactly as hoped, I'd rather watch than write the check.

The operating business is plainly good — returns on equity of 22-31% every year, gross margin up from 0.47 to 0.52, operating margin from 0.11 to 0.14, revenue reaccelerating to 20%, and book value per share compounding from 24 to 51 in eight quarters. But the latest net margin of 0.31 against an operating margin of 0.14 means the bulk of that headline profit came from something other than selling goods and renting servers, and the 20.66 P/E that looks reasonable is resting on it; strip out the non-operating windfall and you are paying a good deal more than 20 times for the real thing. Worse, free cash flow per share has gone from +3.05 to +0.71 to negative in the last two periods — the company is now eating capital faster than it produces it, with a current ratio of 1.03 and debt equal to equity. I don't short a business with these returns, and I don't buy earnings I can't trace to cash; that puts it in the too-hard pile until the capex cycle shows what it actually earns.

At a P/E of 20.66 on trailing earnings of $5.81 the implied price is roughly $120 against book value of $51.14 — a price-to-book of 2.35, so that the product of the two multipliers is some 48, more than double the 22.5 I permit a defensive investor. The balance sheet compounds the objection: the current ratio has not once in this record reached 1.2, standing at 1.03 today against my minimum of 1.5, debt equals equity at 0.99, and free cash flow per share is negative in three of the last four periods, which means the reported profits are not arriving in cash. I note further that the latest net margin of 0.31 sits against an operating margin of only 0.14, more than double every prior period — earnings swollen by something other than the business of selling goods, and a multiple computed on such a figure flatters the buyer. The enterprise plainly earns well and its book value has compounded handsomely, but a defensive investor buys a margin of safety, not a record of expansion financed by borrowing and capitalized by hope.

This is a genuine fast grower and I'm not going to pretend otherwise — revenue growth accelerated from 14% to 50%, gross margin went 49% to 54%, operating margin more than doubled from 8% to 17%, and full-year EPS jumped from $1.01 to $2.66. The balance sheet is exactly what you want behind a growth story: debt/equity of 0.26 and a current ratio of 2.6, so a bad year won't kill them. My problem is the price tag — 132 times earnings on an $843 billion market cap. Even against triple-digit trailing growth I'd call that fully paid for, and the growth has to keep compounding at a rate nobody sustains for long just to justify what you're handing over today. Wonderful company, fully priced; I'd rather own it after a disappointment than pay up for the story.

Invert: what kills this is the price, and the price here is preposterous — 132 times earnings and roughly four-and-a-half times book for a business that has earned 3% to 10% on equity across the entire record, with a net margin average of 12% that owes as much to line items as to operations (one quarter shows a negative 17-cent operating margin alongside an 11% net margin, which tells you the earnings and the business are not the same animal). Note too that $1.41 of EPS on $41.18 of book is a 3.4% return on capital, so whatever the 10% ROE figure is measuring, it isn't what an owner gets; and free cash flow at $0.95 a share against $1.41 of reported earnings says the cash is thinner than the accounting. The balance sheet is sound — 0.26 debt to equity, 2.6x current — and revenue growth of 50% with gross margin up four points is genuinely improving, so this is not a fraud or a wreck; it is a mediocre-return, capital-hungry, fiercely competitive semiconductor business priced as if it were a royalty on the future. Buying it requires believing that a company compounding book value at 8% deserves 132 times earnings, and I would rather miss it ten times than pay that once.

This is a fine business getting better — gross margins have climbed from 49% to 54%, revenue is compounding at 50%, the balance sheet carries almost no debt at 0.26 debt-to-equity with a 2.6 current ratio, and book value per share has quietly compounded 8% a year. But the returns on equity averaging 6% tell me the capital being retained isn't yet earning its keep, and chip design is a business where the leader in one decade is the laggard in the next — that's outside the kind of durable, predictable moat I'm willing to underwrite for ten years. At $842 billion, 132 times earnings, and roughly 200 times the $0.95 of free cash flow per share, the market has already paid in advance for a decade of flawless execution. I'd rather own a wonderful company at a fair price than a promising one at a price that leaves no room for a single mistake.

The balance sheet is the one part of this I can praise: a current ratio of 2.61 and debt to equity of 0.26 would satisfy any defensive investor, and the book value per share has compounded steadily from 34.91 to 41.18. But at 132 times trailing earnings — with EPS of 1.41 implying a price near 187 against book value of 41.18, a price-to-book above 4.5 — the product of the two multiples exceeds 599 where I would demand no more than 22.5. Earnings, moreover, are thin and erratic for so grand a price: return on equity has wandered between 3% and 10%, operating margin turned negative as recently as the June 2025 twelve-month period, and full-year earnings per share fell to 1.01 in 2024 before recovering. One is not buying an enterprise here but a forecast, and I have never found a way to make a forecast serve as a margin of safety; the sound finances do not redeem a price that capitalizes hope.

The inflection here is real and it's violent — revenue went from shrinking 20-50% through all of FY25 to up ~190% in the last two TTM prints, gross margin ripped from 21% to 43%, and the operating loss ratio halved from -12x to -5.5x revenue. Balance sheet won't kill you either: D/E of 0.21 and a 3.9x current ratio means no leverage trap, so this is never a short. My problem is the price: a $5.1B cap against a revenue base so small that net loss is still ten times sales means the market has already paid for five years of this trajectory and then some, and the multiple isn't disagreeing with the trend — it's front-running it by miles. When the fundamentals are improving but the price already assumes perfection, the right size is zero; I'd rather wait for the growth to grow into the cap or for a break that hands me the asymmetry.

I can't categorize this one as a fast grower, a stalwart, or anything else, because there's nothing to categorize — the company loses money in almost every period shown, the trailing P/E is negative at -18.4, and free cash flow per share has been negative straight through. Revenue is up sharply (185% on the latest twelve months), but that growth lands on a -5.5 operating margin and a -$0.16 EPS, so revenue isn't translating into earnings; the two quarters that show positive net margin sit alongside deeply negative operating margins, which tells me the profit came from something other than the business. The balance sheet is actually the good news — debt/equity of 0.21 and a current ratio near 4 buy plenty of time — but $5.1 billion of market value against $1.61 of book per share and no earnings at all is a price paid entirely for a story. Earnings drive stock prices in the long run, and until there are earnings I don't know what I own, so I'd rather watch this one than pay up for it.

Quantum computing sits well outside my circle of competence, but I don't need to understand the physics to read the arithmetic: operating margins of minus 546% mean this outfit spends roughly six and a half dollars to bring in one, and it has done so in every period shown. There is no moat here because there are no durable earnings to defend — return on equity has averaged minus 90%, free cash flow per share is negative throughout, and what book value growth exists came from selling stock to strangers rather than from retained profits. The balance sheet is clean enough, with modest debt and a current ratio near four, but that only tells me how long the losses can be financed, not that they will ever stop. At $5.1 billion for a company with $1.61 of book value per share and no earnings to capitalize, the price already assumes a future I cannot verify from the numbers in front of me — I'd rather miss it than own it.

An operating margin of minus 546 percent means this company spends roughly seven dollars to collect one, and it has done so in every single period shown — the best operating margin in the entire history is minus 635 percent. The occasional positive net margin in 2026/Q1 and 2025/Q1 is not earnings; real operations never turned a dollar, so that is mark-to-market noise on warrants or securities, and I don't pay for accounting fiction. The book value per share 'compounding' at 93 percent is the tell — it grew from 0.37 to 1.61 while free cash flow per share was negative every quarter, which means the equity came from selling stock to the public, not from making money, and each round of financing buys the business another year of burning it. Five billion dollars of market value for a business with negative returns on equity averaging minus 90 percent requires believing something stupid about a technology that may not produce a commercial product in my lifetime; I'd rather own a decent rock quarry.

There is no earning power here to capitalize: the P/E is negative at -18.4, the enterprise has lost money in nine of the ten periods shown, and operating margins of -546% mean the company spends more than five dollars for every dollar of revenue it books — the two quarters showing positive net margin sit alongside deeply negative operating margins, so those profits came from something other than the business. Book value per share is $1.61 and has grown 93% annually not through retained earnings, which are negative, but evidently through issuing shares; a $5.1B market capitalization set against that book and against free cash flow that is negative in every period is a price paid entirely for a hope, not for a demonstrated result. The balance sheet is the one creditable item — a current ratio of 3.94 and debt to equity of 0.21 — but a strong balance sheet financed by stock sales only buys time; it is not value, and it certainly does not justify the price. I would not own this at any price near the present one, and the margin of safety is not merely thin but inverted.

The inflection here is real and it's violent — TTM revenue growth nearly doubled from ~36% to 62-63% in the last two prints, gross margin ratcheted from 27% to 36%, operating margin halved its loss from -48% to -25%, and the balance sheet went from 2.1x debt/equity to 0.20x with a 5.5x current ratio and book value per share up 9x. That's the operating-leverage curve I get paid to own, and there's no blow-up risk left in the capital structure. The problem is the tape has already done the math: at a -233 P/E on a -21% net margin, they're paying roughly 50x sales for a company still burning free cash, which prices in not just this acceleration but the next two years of it. Great trend, fully discounted price — I'd rather sit with a hair trigger and size up if the growth rate holds while the multiple compresses, than pay perfection for a business that still can't self-fund.

I can't run a PEG test here because there's no E — this company has lost money in every period shown, and a P/E of -233 isn't a bargain, it's an absence of earnings. The good news is real: revenue up 62%, gross margin climbing from 27% to 36%, and a balance sheet they've cleaned up beautifully (debt/equity from 2.10 down to 0.20, current ratio 5.5), so this thing can survive a bad year. But the market is asking $37.7 billion — roughly 13 times book — for a business still burning cash at the free-cash-flow line and running a -25% operating margin, which means you're paying a spectacular price for a story about earnings that haven't arrived. I like the company and I'd love to own the rockets; I don't want to own them at a price that already assumes the profits, because in the long run earnings drive the stock and here there aren't any yet.

I understand rockets about as well as I understand cryptocurrency — which is to say, not well enough to hand over $37.7 billion for a business that has never earned a dime. Give the folks credit where it's due: gross margins have climbed from 27% to 36%, revenue is compounding better than 60%, debt-to-equity has come down from 2.10 to 0.20, and the current ratio of 5.48 means nobody's knocking on the door. But the book value per share didn't triple because earnings piled up — it tripled because they sold stock to people willing to pay these prices, free cash flow has been negative every single period shown, and return on equity has never once been positive. A market price of $37.7 billion on a company still losing 21 cents on the dollar isn't a margin of safety; it's a bet that everything goes right for a decade, and I'd rather be roughly right about a certainty than precisely wrong about a hope.

Ten periods of data and not one dollar of profit — net margin averaging -35%, return on equity averaging -29%, and free cash flow per share negative in every single period without exception. The book value per share compounding at 156% is not compounding at all; it is the arithmetic of selling stock to strangers, which is also how debt-to-equity went from 2.10 to 0.20 — the shareholders paid down that debt, not the business. Yes, gross margin is climbing and revenue is growing 60-odd percent, and someone will tell me the rockets are marvelous, but $37.7 billion of market value against roughly three-quarters of a billion in revenue and a permanent cash burn means you are paying about fifty times sales for the privilege of funding other people's science project. I would rather sit on my hands: buying this requires believing the terminal economics of the launch business are something no one has yet demonstrated, and I do not pay fifty times sales for a hypothesis.

The record shown contains not a single period of positive earnings — ten consecutive trailing-twelve-month losses, with the latest EPS at -$0.08 and free cash flow negative throughout — so the P/E of -233 is not a valuation but an absurdity, and there is no demonstrated earning power whatever against which to measure a margin of safety. The balance sheet is genuinely respectable: a current ratio of 5.48 and debt-to-equity of 0.20 are the marks of a solvent enterprise, and I grant the company that much; but I observe that the book value per share rose from $0.64 to $5.84 chiefly by selling stock to an enthusiastic public rather than by retaining profits. At roughly $18.65 a share the market pays about 3.2 times a book value it largely donated, and capitalizes 62% revenue growth and improving gross margins that have yet to reach the operating line. That is a wager on the future, not an investment measured against the past, and the defensive investor has no business in it at any price the crowd is presently willing to pay.