AI 对冲基金 · 大师信号(存档)

This is the kind of second-derivative move I pay up for: quarterly EPS went 69.33, 69.55, 76.64, 87.10, 93.54, 110.44, 136.27 — that's year-over-year growth of 35%, then 59%, then 78%, accelerating, not fading. The margin inflection is doing the heavy lifting: operating margin ran 47% to 60% and net margin 43% to 56% in six quarters, with the bulk of it in the last two prints, while revenue growth re-accelerated off the 20% Q4 blip to 36% and ROE lifted 35% to 40%. And they're doing it while deleveraging — D/E 0.56 down to 0.45, current ratio 2.46 — so the worst-case tail is a growth slowdown, not a blowup. At 30.75x the market is paying for a good business growing 30%; the most recent quarters say earnings are compounding at nearly triple that, and I'll take that asymmetry in size until the margin trend rolls.

This is a fast grower plain and simple: EPS went 223 in FY2024 to 327 in FY2025, and the last four quarters add to about 427 — better than 50% growth — with the two most recent quarters up 59% and 78% against the year before. You're paying 30.75 times earnings for that, so the PEG is roughly 0.6, and I'll take a P/E half the growth rate every day of the week. The story checks out down the line: revenue up about 35%, gross margin up from 58% to 68%, operating margin 47% to 60%, net margin 43% to 56%, ROE at 40%, book value compounding 24% a year, debt/equity actually falling from 0.56 to 0.45 with current assets 2.5x current liabilities — a balance sheet that can eat a bad year. The one thing that keeps me from calling it a 95 is that chipmaking is cyclical and these margins are the best I've ever seen here, so some of this growth is borrowed from a boom; earnings drive stock prices, and as long as these earnings keep marching, you're not overpaying for them.

Invert it: what kills this? Not the balance sheet — debt to equity fell from 0.56 to 0.45 while the current ratio sits near 2.5, so nobody is borrowing to manufacture the earnings. Not the returns — ROE climbed 0.29 to 0.40, gross margin 0.58 to 0.68, operating margin 0.47 to 0.60, all while revenue grew 36%; a company that raises price and volume simultaneously has a moat, not a marketing department. The one honest objection is capital intensity: FY2025 free cash flow per share of 191 against 327 of earnings means roughly forty cents of every dollar earned goes back into fabs just to stay in front, and semiconductors have humiliated optimists before. But book value compounding at 24% a year and a 30x multiple against that growth and those returns is a fair price for a great business, and I'd rather own this than argue about it.

This is a business with the fingerprints of a real toll bridge on it: returns on equity climbing from 29% to 40%, net margins from 43% to 56%, gross margin up a full ten points while revenue compounds better than 30% — you don't raise price and volume simultaneously unless the customer has nowhere else to go. Book value per share has compounded 24% a year on modest leverage of 0.45 and a current ratio near 2.5, so the growth was bought with earnings rather than borrowed money, and free cash flow of roughly 217 per share over the trailing year puts the stock near 19 times owner earnings against a 30.75 P/E — a fair price for that quality, not a steal. My hesitation is honest: this is a spectacularly capital-hungry business in a cyclical industry, where today's fortress is rebuilt every few years at enormous cost, and the multiple already assumes the current boom keeps rolling. On the numbers in front of me I would own it for ten years, sized so a two-year famine in the cycle wouldn't spoil my sleep.

The enterprise itself is unimpeachable: a current ratio of 2.46, debt to equity of 0.45 and falling, book value per share compounding from 775 to 1248, and not a single deficit quarter in the record — by the tests of financial strength and earnings stability this passes handily. The difficulty is entirely one of price. At 30.75 times earnings and, by my arithmetic (136.27 x 30.75 = roughly 4,190 per share against book of 1,248.33), some 3.4 times book value, the product of the two multipliers is above 100 where I insist upon 22.5; the buyer at this level is paying for a continuation of 35-40% returns on equity that the cyclical history of this industry does not guarantee. I will add that the record furnished me spans barely two years, which is no substitute for the ten years I require before granting any premium at all. A sound company bought at a speculative price is a speculation, and I must mark it accordingly.

This is a fast grower hiding inside the world's biggest stalwart: earnings went from $13.71 to $18.01 a share in one year, better than 30%, on 18% revenue growth — and the last four quarters each came in ahead of the year-ago quarter, so the growth is visible, not promised. At 27.85 times earnings against roughly 30% earnings growth, the PEG is under 1, which means you're not paying up for the story; even haircutting to the 18% revenue growth, since some of the earnings jump came from net margin going 36% to 40%, the price is fair rather than foolish. The balance sheet is in good shape and getting better — debt/equity down from 0.80 to 0.71, 34% return on equity, book value compounding 21% a year — so a bad year wouldn't break anything. The one thing I'd keep an eye on: free cash flow per share slipped to $9.02 from $9.64 and the current ratio has drifted from 1.35 to 1.23, which tells me they're spending heavily; if earnings keep marching up, fine, but that's the number that would change my mind.

This is the kind of toll bridge I understand well enough to own: a 68% gross margin, 40% net margin, 34% return on equity, and revenue still compounding 18% at a size where most businesses have long since run out of road — book value per share has gone from $38.75 to $59.58 in two years, and earnings from $13.71 to $18.01, which is management putting retained dollars to work rather than burning them. Debt to equity has come down from 0.80 to 0.71 while all that was happening, so the returns aren't a leverage trick. At 27.85 times earnings you are paying a fair price, not a steal, for a genuinely wonderful business, and I'd rather do that than the reverse. My one reservation, and it keeps this out of the 90s: free cash flow per share slipped from $9.64 to $9.02 even as earnings jumped 31%, so roughly half of reported profit is being plowed back into plant — if those outlays don't earn their keep, today's price starts asking for perfection.

This is one of the few businesses that deserves the word great without argument: 34% return on equity in every single period shown, 68% gross margins, 47% operating margins, book value compounding at 21%, and revenue growth accelerating from 15% to 18% at three and a half trillion of size — that combination does not happen by accident, it happens because of a toll booth. Inverting, the one thing that would kill this is the cash: free cash flow per share fell from $9.64 to $9.02 while earnings per share jumped from $13.71 to $18.01, so barely half of reported profit is turning into cash and the rest is going into the ground as capital spending. I'll accept that so long as the capital keeps earning 34%, but if ROE starts sliding while the spending continues, the whole thesis is a fiction and I'd want out. At 27.85 times earnings growing north of 30% with leverage falling from 0.80 to 0.71, you are not being asked to believe anything stupid — that is a fair price for a great business, which is all I ever ask for.

The inflection already happened and the tape hasn't paid for it: revenue growth went from 12-13% in the middle of FY25 to a locked-in 17-18% for four straight quarters, and EPS went 13.71 to 18.01 — up 31% — with the most recent quarter at 4.82 versus 3.67 a year ago. You're paying 27.85x for a 31% earnings grower doing 34% ROE and 40% net margins while D/E actually falls from 0.80 to 0.71 and book value compounds at 21% — that multiple is priced for a mid-teens grower, not this one, and there's no leverage problem to blow me up. What keeps me from backing up the truck is the capex tell: gross margin slipped 69% to 67%, operating margin rolled from 0.49 in Q1 to 0.45 in Q4, and FCF/share went backwards year over year (9.02 vs 9.64) even as EPS surged — the growth is being bought, not just earned. Good-sized long, not a max position; if op margin keeps sliding and free cash keeps diverging from EPS, the story changes fast.

The enterprise is unimpeachable — a 34% return on equity, 40% net margins, and not a single deficit quarter in the record shown — but I am asked to buy the shares, not admire the company. At $18.01 of trailing earnings and 27.85 times those earnings, the price is near $502 against book value of $59.58, a multiple of roughly 8.4 times assets; the product of the two exceeds 230, where my limit for a defensive buyer is 22.5. The balance sheet is merely adequate rather than strong: a current ratio of 1.23 has never satisfied my requirement of 1.5, and debt sits at 0.71 of equity. One pays here for a projection of what the future must deliver, and I have never found the future obliging enough to warrant surrendering the entire margin of safety in advance.

This is the inflection I get paid for: revenue growth had decayed to the low-30s through mid-2025 and has now snapped back to 62-63% in the two most recent filings, and it's not empty growth — gross margin went 27% to 36-38%, operating margin compressed from -50% to -25%, and net margin from -49% to -21%. ROE has gone -0.45 to -0.08 and TTM losses per share are a third of what they were at the FY25 mark, so the earnings trajectory is building, not fading. Just as important for staying alive: D/E collapsed from 2.10 to 0.20 and current ratio is 5.5x — they funded through equity, book value per share went $0.64 to $5.84, so the deteriorating-fundamentals-plus-leverage trap that blows up accounts isn't here. The offset is real and keeps me from swinging with both hands: at $37.5B on a company still burning free cash (-$0.18/sh, worse than last quarter) the market already believes the story, so the multiple is doing the work — I'd own it sized for the trend, not for the valuation, and the day growth decelerates back toward 40% I'm gone.

Here's my problem: there are no earnings to buy. Revenue is growing beautifully — 62% and accelerating — gross margin has climbed from 27% to 36%, and they've cleaned up the balance sheet nicely (debt/equity down from 2.10 to 0.20, current ratio 5.48), so this isn't a company in trouble. But the P/E is negative, free cash flow per share is still negative, and the market is paying $37.5 billion for a business that has lost money in every period shown. You can't run a PEG test on losses, and when I can't tell you what I'm paying per dollar of earnings, I'm paying for a story about rockets instead. Earnings drive stock prices in the long run — come back when the bottom line actually turns, because at this price the good news is already spoken for.

Rockets are a thrilling business to read about and a hard one to own: I see ten straight periods of losses, an operating margin still at negative 25% on the latest trailing twelve months, and free cash flow per share that has never once been positive. The book value per share has certainly climbed, but it climbed by selling stock to newcomers rather than by retaining earnings — that is a financing achievement, not a compounding machine, and the same story explains why debt-to-equity fell from 2.10 to 0.20. Give credit where due — gross margin improved from 27% to 36%, revenue grew 62%, and a 5.5x current ratio means no one is knocking on the door — but the market is asking $37.5 billion for a company that has yet to earn a dime, which is a price that requires everything to go right for a decade. I would rather watch this one from the sidelines and buy a business whose earnings I can already count.

Ten periods of data and not one dollar of profit or free cash flow in any of them — ROE negative in every single line, operating margin still negative 25% after all this alleged progress, and cash going out the door every quarter. The book value 'compounding' at 156% is not compounding at all; it is share issuance dressed up as capital creation, which is how you finance a business that consumes capital rather than earns it, and the same trick is what shrank debt-to-equity from 2.10 to 0.20. Yes, gross margin improved from 27% to 36% and revenue grew 62%, so the rockets evidently fly — but at $37.5 billion you are paying something north of twelve times book and dozens of times sales for a company that has never earned a cent, which requires believing that a capital-intensive business in a field littered with the bones of optimists will soon earn returns nobody in the industry has ever sustained. I would rather be approximately right about this one: a fine engineering story at an idiotic price is still a bad investment.

Here is an enterprise that has not earned a dollar for its owners in any of the ten periods shown, yet Mr. Market prices it at $37.5 billion — better than three times a book value of $5.84 a share that was built by selling stock to the public rather than by retaining profits. The balance sheet is genuinely strong (current ratio 5.48, debt to equity reduced from 2.10 to 0.20), and margins are mending, but strength of finances only tells me the company will survive its losses, not that I should pay for them; a negative P/E cannot be reconciled with any standard I would apply. Improving gross margins and 62% revenue growth are the raw material of hope, and hope is precisely what this price capitalizes — the defensive investor requires demonstrated earning power, and there is none. I would not own it at this price, and I would want to see several years of actual profits before I revisited the question.

Revenue growth bottomed at 56% in the July 2025 quarter and has accelerated four straight quarters to 106% — that's not a decelerating megacap, that's a business re-accelerating off a bigger base, and operating margin went with it, from 49% during the ramp-cost quarter back to 66% and holding. Quarterly EPS has compounded 0.78 to 1.10 to 1.32 to 1.78 to 2.42 to 2.48; run-rate the latest quarter and you're at roughly $9.90 against a 26.8x trailing multiple, so the market is paying about 21x forward for triple-digit top-line growth — the price says maturity while the numbers say the opposite, and that gap is the whole trade. Balance sheet backs the swing: 0.40 debt/equity, 4.6x current ratio, ROE 117% — none of the leverage that turns a wrong call into a blowup. The one thing I'd watch is FCF/share falling to 0.89 from 2.02 while EPS climbs, which smells like inventory and receivables building into the ramp; I'd size big here but I'd be out fast if growth rolls and cash conversion doesn't follow earnings.

This is a fast grower by any definition — TTM revenue up 106% and EPS marching from $0.80 to $2.48 a quarter over eight straight quarters, with growth reaccelerating rather than cooling. At a P/E of 26.77 against earnings growth north of 60%, the PEG is under 0.5, which means you're paying a stalwart's multiple for a fast grower's numbers; that's the setup I spent a career looking for. The story checks out down the line: gross margin steady at 75%, operating margin improving to 66%, net margin 62%, a current ratio of 4.6 and debt/equity of just 0.40 — a balance sheet that survives a bad year. Two things keep me from calling it a classic 90: free cash flow per share ($0.89) is running well behind EPS ($2.48) this quarter, which usually means receivables and inventory are swelling, and a $5 trillion company simply has less room to double than a small one — but earnings drive stock prices, and these earnings are real and cheap.

The economics here are the finest I've ever laid eyes on — 75% gross margins, 66% operating margins, 117% return on equity, book value per share compounding at 74% a year, and a fortress balance sheet with 0.40 debt-to-equity and a 4.59 current ratio. At 26.77 times trailing earnings with revenue doubling, the price isn't the absurdity people assume; the trouble is that I don't know how to handicap a semiconductor business ten years out, and returns like these are a loud invitation to every engineer and every customer with a checkbook to build the thing themselves. What does give me pause in the numbers is cash: trailing EPS of $2.48 against only $0.89 of free cash flow per share, down from $4.01 against $4.98 a year prior — earnings are running well ahead of the cash arriving at the door, which is usually inventory and receivables doing the talking. A wonderful business at a fair price, but one whose ten-year earning power sits outside my circle, so I'll admire it from the sidelines rather than pretend to know.

The operating numbers are as good as anything I have ever seen: 75% gross margins, 66% operating margins, return on equity above 100% in every period shown, book value per share compounding at 74%, and almost no debt. But every one of those periods is the same boom — there is not a single down cycle in this record, and you cannot call a business durable when the whole history is the good part of one capital-spending cycle. Two things bother me in the inversion: cash conversion collapsed in the latest quarter, with free cash flow per share of $0.89 against $2.48 of earnings after running 80%+ for two years, while debt-to-equity went from 0.33 to 0.40 — earnings up, cash down, leverage up in one quarter is exactly the pattern you find when a company is financing its customers' enthusiasm. At 27 times earnings the price is not obviously stupid against this growth, but $5.2 trillion requires the buildout never to pause, and I don't know that, so it goes on the too-hard pile.

The balance sheet is beyond reproach — a current ratio of 4.59 and debt to equity of 0.40 leave no question of solvency, and every period shown is profitable. But the defensive investor buys earnings and assets, not enthusiasm: at a P/E of 26.77 on $2.48 of trailing earnings the shares fetch roughly $66 against $9.50 of book value, near seven times tangible net worth, so the product of the two multiples is some 187 where I would tolerate 22.5. A return on equity above 100 percent and revenue doubling year over year are not durable facts but invitations to competition, and the record offered spans barely two years — far short of the decade of stable earnings I require. A sound enterprise, plainly, but the price capitalizes a future no one has demonstrated; there is no margin of safety here, and I must record the overvaluation as the disqualifying fact it is.

Revenue growth has gone 30% to 39% to 63% to 93% across eight trailing periods — that's not a good business, that's a business accelerating into a second derivative inflection, and I'll pay up for that all day. Operating margin ripped from 1% to 47% and net margin to 55% while TTM EPS more than doubled off 0.20 a year ago to 0.44, with ROE tripling to 38% and debt/equity actually falling to 0.18 — growth this steep with no leverage and a 7x current ratio means the blow-up risk is in the multiple, not the balance sheet. At 147x trailing the market clearly believes something good, but it's capitalizing a snapshot that EPS momentum is going to make look cheap within a couple of prints; the trend is still running ahead of what the price assumes. The one thing that keeps me from full size is the margin line flattening at the top end — 46 to 47 is where acceleration goes to die — so I own it big but I watch that number like a hawk, because at this multiple the day the inflection stops is the day you're out, not the day you average down.

The business itself is admirable — 85% gross margins, operating margins that climbed from 1% to 47% in eight quarters, returns on equity marching from 11% to 38%, book value per share compounding at 41%, almost no debt at 0.18 to equity, and a current ratio north of seven. That is the arithmetic of a company with real pricing power and an owner's balance sheet; I have no quarrel with the enterprise. My quarrel is with the ticket price: $409 billion against $0.44 of trailing earnings and $0.50 of free cash flow per share is roughly 147 times earnings and 130 times cash — a price that assumes this 93% growth and margin expansion continues for a decade without a stumble, and leaves nothing for the buyer if it doesn't. I also confess this sits at the edge of what I understand about how such software earns its keep ten years out, so I'll admire it from the sidelines rather than call it cheap or call it broken — a wonderful company, but not at a fair price.

The business itself is close to wonderful: 85% gross margins, operating margin up from 1% to 47% in eight quarters, revenue accelerating to 93%, ROE compounding from 11% to 38%, book value per share growing 41% a year, almost no debt, and a current ratio above 7 — and the cash is real, with free cash flow per share running ahead of earnings. I do note the net margin now exceeds the operating margin, which means a chunk of the reported profit is interest and tax accounting rather than selling software, so the headline earnings flatter the enterprise. But at $409 billion you are paying 147 times earnings and roughly 45 times book for about $2.8 billion of net income — to earn a decent return from here you must believe the earnings multiply more than tenfold and that nothing intervenes, and I have never gotten rich believing that sort of thing. A fine business at a preposterous price is not an opportunity; it is a temptation, and I decline it.

This is a fast grower and no one can argue otherwise — revenue up 93% and accelerating from 29% two years ago, operating margin from 11% to 47%, EPS marching 0.06, 0.09, 0.14, 0.20, 0.25, 0.36, 0.44 quarter after quarter, and a balance sheet with almost no debt (D/E 0.18) and a 7x current ratio, which is exactly the kind of fortress that lets a growth story survive a bad year. The trouble is the price tag: 147 times earnings on roughly $2.8B of profit is a story stock's multiple, and when I see a 55% net margin sitting above a 47% operating margin, I know part of that bottom line is coming from somewhere other than selling software. Even if you credit the full triple-digit earnings growth, the moment it normalizes toward a still-terrific 40-50% the PEG goes to three, and you're paying for five years of perfection you may or may not get. Wonderful company, wonderful numbers, but I don't get paid to buy the story at 147 times — I'd know what I own here and I'd wait.

The enterprise itself is unobjectionable: a current ratio of 7.23, debt to equity of 0.18, and an unbroken record of positive and rising earnings would satisfy my tests for financial strength handsomely. But an investment operation is a price paid for a value received, and here the price is 147 times trailing earnings against a book value of $4.11 a share — roughly sixteen times net assets, so that the product of the two multipliers exceeds my limit of 22.5 by a factor of a hundred. At such a figure the buyer receives an earnings yield under seven-tenths of one percent and is paying not for demonstrated earning power but for a continuation of 93% revenue growth that no company sustains indefinitely. A superb balance sheet is no margin of safety when the entire purchase price is the capitalization of hope; the defensive investor must decline, and the speculative buyer should understand exactly what he is doing.

The operating inflection here is real and I won't pretend otherwise: TTM revenue growth flipped from -34% to +199% and +185% in the last two prints, gross margin climbed off the 21% trough to 43%, and the operating loss as a multiple of revenue halved from -12x to -5.5x. But you don't get paid for being right about a trend the tape already worships — $5.1B of market cap against revenue this small is several hundred times sales, which is the market saying it already believes every bit of this acceleration and then some. Underneath, book value per share is going backwards (1.75 to 1.61), free cash flow per share is getting worse, the current ratio has collapsed from 37x to 3.9x in two quarters, and the positive EPS prints are fair-value noise, not earnings. Inflection yes, asymmetry no — low debt keeps it off my short sheet, but the correct position on a right-trend/right-price standoff is none.

This isn't a fast grower, a stalwart, or a turnaround — it's a concept stock, and I can't put a PEG on a company with no E. Operating margins run around negative 500% of sales, meaning they spend roughly six dollars for every dollar of revenue they book, and the two quarters that show positive net margins came with deeply negative operating margins, so that's accounting noise, not profit. Book value per share nearly doubled, but it climbed by selling stock, not by earning money — that's dilution dressed up as progress, even if it does leave a clean balance sheet with 0.21 debt-to-equity and plenty of current assets. Revenue is growing fast off a base so small it barely matters, and the market is asking 5.1 billion dollars for the story; earnings drive stock prices in the long run, and there are none here to drive anything.

Invert: what kills you here? A company with roughly $28 million of revenue carrying a $5.1 billion market cap — call it 180 times sales — while its operating margin runs minus 546 percent, meaning it spends about six and a half dollars to book one dollar of revenue, and has done so every single period shown. Free cash flow per share is negative in all ten periods, ROE averages minus 90 percent, and the only two 'profitable' quarters show positive net margin sitting on top of operating margins near minus 600 percent, which tells me the earnings came from mark-to-market accounting, not from selling anything to anybody. Book value per share nearly doubling is not compounding; it is shareholders being asked to write fresh checks, and the fortress balance sheet — D/E of 0.21, current ratio of 3.94, down from 37 — is simply the proceeds being consumed. Revenue actually shrank through most of 2024 and 2025 before the recent bounce, so this is not even a great business at a silly price; it is a capital-destroying science project at a preposterous one, and no amount of enthusiasm about quantum physics changes arithmetic.

I don't understand this business well enough to own it, and what the numbers do tell me is discouraging: an operating margin of roughly negative 546% means the company spends something like six dollars for every dollar it takes in, and free cash flow per share has been negative in every period shown. The book value per share has compounded nicely, but that growth came from selling shares to the public rather than from earning money — the return on equity has been negative in all ten periods, averaging -90%, and the current ratio's fall from 37 to under 4 tells me the last raise is being consumed. At a $5.1 billion market capitalization against $1.61 of book value per share and no earnings at all, the price already assumes a decade of flawless execution in a field whose economics nobody can yet forecast. The balance sheet is clean and the cash gives them time, which is why I'd call this a pass rather than a short — but I'd rather stay comfortably out of it than pay for a lottery ticket.

There is no earning power here to capitalize: the trailing multiple is negative, operating margins run between minus five and minus fifteen times revenue, and the only two quarters showing positive net margin sit alongside deeply negative operating margins — accounting gains, not business results. The balance sheet is the one respectable feature — current ratio 3.94, debt to equity 0.21 — but I note the book value per share nearly doubled while the company lost money every year, which means shareholders bought that strength with their own dilution rather than earning it. A market capitalization of $5.1 billion against a book value of $1.61 per share, cumulative deficits, and negative free cash flow in every period shown is a price that capitalizes hope for quantum computing, not demonstrated results. The defensive investor requires positive earnings across the whole record; this enterprise has never produced them, so no margin of safety can exist at any price above liquidating value, and this is far above it.

The franchise itself is about as good as capitalism produces — 82% gross margins that haven't budged an inch in three years, average returns on equity of 35%, and book value compounding at 23% a year. But invert and look at the newest filing rather than the average: operating margin fell from 43% to 31%, ROE from 0.41 to 0.30, debt-to-equity climbed from 0.51 to 0.72, and free cash flow per share collapsed from $18 for the full year to $0.69 on a trailing basis. That is a company shoveling every dollar of owner earnings, plus borrowed money, into capital spending whose returns nobody has yet demonstrated, and the net margins bouncing from 5% to 48% between quarters tell me the reported earnings deserve a squint. Twenty-five times earnings is not a silly price for a great toll bridge, but I don't pay for earnings I can't see in cash, so this one sits on the too-hard pile until the spending either shows up as profit or stops.

This is plainly a wonderful business by the numbers — 81% gross margins that haven't budged in two years, returns on equity in the thirties, net margins near a third of revenue, book value per share compounding at 23%, and revenue still growing 28% at enormous scale. That's a toll bridge, and at 25 times earnings I'm not being asked to pay a foolish price for it. What stops me is what's happened in the last two quarters: operating margin has fallen from 43% to 31%, free cash flow per share collapsed from $3.54 to $0.69, debt-to-equity climbed from 0.51 to 0.72, and ROE slid from 0.41 to 0.30 — the owners' earnings are being poured into capital spending at a rate I cannot yet evaluate from these figures, and I don't buy a business hoping the reinvestment works out. I'd rather sit still and watch whether that spending shows up as earnings or as depreciation; a great franchise spending like this is a fine thing to own only if you know the return on the last dollar, and here I don't.

This was a classic fast grower, and the sales still look terrific — revenue up 28% and accelerating, gross margins parked at 81%. But I don't buy revenue, I buy earnings, and the earnings are going the other way: operating margin has fallen from 43% to 31% in a year, trailing EPS has slipped from about $28 to under $27, and free cash flow per share has collapsed from over $21 a year to roughly $16 and dropping fast as the capex bill piles up. Meanwhile debt/equity climbed from 0.51 to 0.72 and the current ratio thinned out — the balance sheet is still decent, but every line is moving the wrong direction. At 25 times earnings you're paying a growth price for a company whose profits are currently shrinking; that's a PEG that doesn't work, and until the spending starts showing up in the bottom line I'd rather watch it than own it.

The top line is the best it's been in three years — revenue growth going 16% to 22% to 26% to 33% to 28% — and that's exactly what the 25x multiple is paying for. But look at what happened underneath in the June quarter: operating margin snapped from 41% to 31%, net margin 48% to 26%, ROE 41% down to 30%, and free cash flow per share went from $5.19 to $0.69. EPS printed $6.22 against $7.20 a year ago — the earnings line is going backwards while the revenue line accelerates, which means the incremental dollar of revenue is now costing more than it earns, and D/E has walked from 0.51 to 0.72 to fund it. A 25 multiple on trailing earnings that include a $10.51 quarter is not cheap once you mark it to the new margin structure; the market is still capitalizing the old one, and that's the gap I'd be on the other side of until the spend shows up in profits rather than just in the balance sheet.

At roughly $674 a share against $102.54 of book value, the market asks some 6.6 times net asset value and 25 times trailing earnings — a product of price-to-earnings and price-to-book near 166, when my own limit is 22.5; whatever the merits of the enterprise, the buyer at this price has purchased no margin of safety whatsoever, only the hope that today's prosperity persists indefinitely. The balance sheet itself I do not quarrel with much: a current ratio of 2.23 is adequate and earnings have been positive throughout the record. But the record is less serene than the averages suggest — net margin has swung from 0.05 to 0.48 between adjacent quarters, operating margin has slipped from 0.41 to 0.31, return on equity has drifted down from 0.37 to 0.30, debt to equity has climbed from 0.51 to 0.72, and free cash flow per share has all but vanished, from $21.23 in fiscal 2024 to $0.69 on the latest twelve months as capital outlays consume the earnings. A defensive investor pays for demonstrated results, not for projections; here he would be paying handsomely for the projections while the demonstrated cash is being spent elsewhere.

The top line is rolling over off an extraordinary base — revenue growth went 100% in Q3'25 to 27%, then 15%, and only 32% in the latest quarter. That's a decelerating series no matter how you smooth it, and operating margin at 0.44 is still below the 0.50-0.51 peak printed a year ago while ROE has faded from 0.28 to 0.24. The market is paying 49x for that, which is a multiple that assumes the old 50-100% growth curve is still intact; the trend flatly disagrees. What keeps me from pressing harder is the Q2'26 sequential bounce — EPS 0.38 to 0.64, margins and ROE both re-expanding — so this is a fade, not a funeral, but decelerating growth at 49x with debt/equity at a series-high 4.93 is precisely the setup where you lose big, and there is no version of this where I'm long.

This is a fast grower that's cooling off while still being priced like it's accelerating — revenue growth ran 50-100% through 2025 and has stepped down to 27%, then 15%, then 32% in the most recent quarters, and the trailing EPS I can count (0.62+0.67+0.38+0.64 = about 2.31 against roughly 1.99 a year earlier) is up maybe 16%, with a big chunk of the year-ago base flattered by that Q4 2024 quarter where net margin hit 90 cents on the dollar. Pay 49 times earnings for 16-33% growth and your PEG is somewhere between 1.5 and 3 — you're paying for the story, not the numbers. Debt to equity of 4.93 and climbing from 3.17, book value per share that just went backwards, and free cash flow per share that bounces from +3.88 to -1.77 quarter to quarter tell me the earnings here ride on how busy the market happens to be — that's a cyclical wearing a growth costume, and you don't buy those at their highest multiples. Fine business, wrong price; I'd want this at half the P/E before I'd own it.

This is a toll booth on other people's speculation, and while the tolls have been rich lately — 82% gross margins, mid-20s returns on equity, book value compounding 13% — I have no way to know what the traffic looks like in a dull market, and the record already shows it: ROE fell to 8% in the 2024 September period and free cash flow per share has swung from +$3.88 to -$1.77 without warning. Earnings that depend on retail enthusiasm and crypto volumes are not the kind I'd care to underwrite for ten years, and the 4.9x debt-to-equity, whatever portion is customer-related, leaves no room for a bad season. At 49 times earnings and roughly ten times book, the price already assumes the boom is permanent — I'd rather own a business where I can guess the next decade than one where I can't guess the next quarter. Wonderful company at a fair price, yes; a cyclical one at a perfect price, no thank you.

The returns look handsome until you ask where they come from: net margin that swings from 24% to 90% in four quarters is not a margin, it is an accounting event, and the 45% average is flattered by it. Free cash flow per share goes 1.99, -1.57, 0.70, 3.88, -1.77, -1.06 — a business whose owner earnings change sign every other quarter is not compounding anything, it is riding customer balances and a speculative trading cycle, and revenue growth decelerating from 100% to 32% tells you which way that cycle is turning. At 49 times earnings and roughly ten times a $10.61 book value, with debt at five times equity, you are paying a 2% initial earnings yield for the privilege of owning a levered transaction house dependent on other people's appetite for gambling. I would rather hold cash than underwrite that; the arithmetic requires believing the good quarter is the permanent one.

At roughly $114 a share against trailing earnings of some $2.31 and book value of $10.61, the buyer pays near 49 times earnings and nearly 11 times book — a product of more than 500 where I counsel no more than 22.5. There is no margin of safety here whatsoever; the price capitalizes an imagined future, not demonstrated earning power. The balance sheet gives no comfort either: debt to equity of 4.93 and a current ratio of 1.22, well short of the 1.5 a defensive investor should insist upon. Earnings, though positive throughout the brief record shown, swing violently — ROE of 0.08 two years back against 0.24 today, and revenue growth ranging from 15% to 100% — which is the signature of a cyclical, sentiment-driven enterprise rather than a stable one. Let the speculator have it at this figure.