General commentary for a general audience, not personalised advice. Share prices are US closes on Friday 6 Nov 2026 (Yahoo Finance): Palantir $160.38, Super Micro $28.76. The mandatory convertible depositary shares (SMCIP) closed at $50.16 the same day (StockAnalysis). Company figures are from Palantir's 2 Nov 2026 results release and Form 10-Q for the quarter ended 30 Sep 2026, and from Super Micro's 3 Nov 2026 results release and Form 10-Q for the quarter ended 30 Sep 2026. Anything marked as our calculation or our assumption is ours. The terminal reads cited below are the TraderMagz board as of 9 Nov 2026, 07:19 UTC.
Palantir grew 96% last quarter, kept 47 cents of operating profit on each revenue dollar, and produced more than a billion dollars of cash. The share price still only works if cash flow keeps compounding at a pace most large software companies never sustain for a decade. Super Micro grew faster still, to $19.9 billion of sales and a $1.6 billion accounting profit, and its share price only works if very little of that profit becomes cash a shareholder can keep.
Two quarters, two bets
Both companies reported the September quarter in the first week of November. Palantir's $2.31 billion of revenue and $0.44 of adjusted earnings were above LSEG estimates of $2.18 billion and $0.41 (CNBC, 2 Nov). Super Micro's sales and $1.81 of adjusted earnings were above LSEG estimates of $16.39 billion and $1.23 (Reuters, 3 Nov). Neither stock held a celebration. Palantir closed at $178.34 on 2 Nov, the day of its report, and at $160.38 on 6 Nov (Yahoo Finance). Super Micro closed at $31.39 on 3 Nov, the day of its report, and at $28.76 on 6 Nov. The tape is a poor referee in a week when our own market-health read was unhealthy. The multiples are the referee.
Palantir sells software that large institutions use to run data and decisions, in government and, increasingly, in business. Revenue was $2.313 billion, up 96% year over year and 20% from the June quarter. The United States was $1.937 billion of that, up 119%. Inside the US, commercial revenue was $984 million, up 148% year over year and 29% sequentially, and it passed US government revenue ($953 million, up 96%) for the first time (company release, 2 Nov). That crossover is real and it is easy to over-read. On the 10-Q's own segments, government as a whole was still $1.146 billion, 49.5% of revenue, and commercial was $1.168 billion. For the first nine months of the year the company says 51% of revenue was still government. The mix shifted. It did not flip.
The profit and the cash are the part that has to be faced. GAAP income from operations was $1.085 billion, a 47% margin. Adjusted operating income, which adds back stock-based pay and the payroll tax on it, was $1.427 billion, a 62% margin. GAAP net income attributable to common stockholders was $1.128 billion. GAAP and adjusted earnings were both $0.44 a share. They match because the add-back of stock-based pay is roughly offset by the tax the company would owe at the 23% long-term rate it uses in the adjusted figure. The actual tax provision was only $13.6 million on $1.148 billion of pretax income. Cash from operations was $1.205 billion. Capital expenditure was $8 million. Adjusted free cash flow was $1.216 billion, a 53% margin. (Free cash flow is the cash left after the spending needed to maintain and grow the business. Palantir's version starts from operating cash flow, adds back cash payroll tax on stock awards, and subtracts a very small amount of equipment spending.)
The guide went up with the quarter. For the December quarter the company expects revenue of $2.649 to $2.653 billion and adjusted operating income of $1.640 to $1.644 billion. For the full year 2026 it raised revenue to $8.530 to $8.534 billion, which it describes as 91% growth, US commercial revenue to more than $3.559 billion (at least 143% growth), adjusted operating income to $5.245 to $5.249 billion, and adjusted free cash flow to $4.8 to $5.0 billion.
Two quieter lines belong next to that. Total contract value closed in the quarter was $2.918 billion, up only 6% year over year, and US commercial contract value was $1.463 billion, up 11%, against commercial revenue growth of 148%. New bookings did not grow like revenue. US commercial remaining deal value, the stock of deals still ahead, was $6.517 billion, up 79% year over year and 4% from the June quarter, so the pile did not shrink. Noncancelable remaining performance obligations were $4.9 billion (10-Q). That figure leaves out contracts of twelve months or less, and it leaves out anything a customer can cancel. Palantir says many customers can terminate for convenience. Net dollar retention was 165% in the quarter, and the company counted 686 US commercial customers (Q3 2026 business update). Across government and commercial it had 1,090 customers in the trailing twelve months, and no customer was more than 10% of revenue (10-Q). One customer was 27% of receivables.
Super Micro builds AI servers and the racks around them. Net sales were $19.946 billion, against $5.018 billion a year ago and $11.1 billion in the June quarter (the year-ago figure is the 10-Q, the June figure is the company's comparison in the 3 Nov release). That is up 298% year over year, our calculation. GAAP gross margin was 12.7%, down from 17.5% in the June quarter and up from 9.3% a year ago. Gross profit was $2.541 billion. Operating expenses were $477 million, which is 2.4% of sales, so income from operations was $2.064 billion, a 10.4% margin (our calculation). Net income was $1.594 billion, an 8.0% margin (our calculation). Diluted earnings were $1.75 a share. Of that profit, $1.308 billion was attributable to common stockholders on a diluted basis after the preferred's share. Non-GAAP gross margin was 12.8% and non-GAAP diluted earnings were $1.81.
The cash flow statement tells a different quarter. Operations used $434 million. Purchases of equipment and a small investment were $47 million. Receivables rose $3.56 billion, to $9.69 billion. Inventory rose $2.48 billion, to $15.05 billion. Deferred revenue, which is cash customers have already paid for work not yet counted as sales, rose $3.05 billion, to $5.67 billion including the long-term piece. Receivables and inventory together are $24.7 billion. The common equity, at the 6 Nov close, is worth $19.0 billion.
One customer was 44% of net sales in the quarter, about $8.8 billion (our calculation). A year ago the top two were 17% and 12%. Two customers were 49% and 17% of receivables at 30 Sep. The company guided the December quarter to sales of $19.5 to $21.5 billion and fiscal 2027 sales, the year ending 30 June 2027, to $72 to $82 billion. Reuters reported that the prior full-year forecast was $65 to $72 billion (3 Nov). The 3 Nov release labeled the figures preliminary because the close was not finished. The 10-Q reports the same sales and the same $1.594 billion of net income.
What the price has to believe
A reverse DCF (discounted cash flow) runs a valuation backwards. The ordinary version forecasts cash flows and arrives at a price. This version starts from the price and asks what the business has to deliver for that price to be fair. Ours is deliberately plain. We use a 9% required return and, after year ten, 3% growth forever. Both are our assumptions, not the companies'. Change them and the levels move. The shape does not.
Palantir. We take the shares actually outstanding on 26 Oct 2026, the date on the 10-Q cover: 2,306.7 million Class A, 101.8 million Class B and 1.0 million Class F, together 2,409.5 million. At $160.38 that is $386 billion of equity (our calculation). Cash and marketable securities were $10.56 billion at 30 Sep ($3.62 billion of cash and $6.94 billion of securities). There is no meaningful debt. Enterprise value, the value of the business apart from the cash already in the bank, is about $376 billion.
The starting cash flow is the September quarter's adjusted free cash flow of $1.216 billion, times four, or $4.86 billion. That is a run-rate, not a trailing twelve-month total the company has published. The full-year guide, $4.8 to $5.0 billion, lands in the same place. We then ask how fast that $4.86 billion must grow for ten years so that, discounted at 9% and growing at 3% after that, it is worth $376 billion.
The answer is about 23% a year. At that pace, year-ten free cash flow is about $38 billion. If the cash margin stayed near 50%, that would be on the order of $76 billion of revenue, about nine times the $8.53 billion the company has guided for 2026. At an 8% required return the growth rate falls to about 20%. At 10% it rises to about 26%. Palantir also earned $89 million of interest in the quarter on the cash pile. If we strip that out, so the cash is not counted both as a pile of money and as the interest it throws off, the required growth is about 24%.
A 23% decade is not a repeat of this quarter's 96%. It is a fade that then refuses to keep fading. The paths around it are the whole argument:
The last row matters. A few very strong years, of the sort the company is having now, and then a long glide to 15%, is roughly what the price already says. The quarter is ahead of that glide path. It does not put the stock on sale. A future in which growth becomes merely excellent, 15% cash-flow growth for a decade from a business that already converts half of sales into cash, is a large gap the other way.
Super Micro. Common shares outstanding were 659.2 million on 31 Oct 2026 (10-Q cover). At $28.76 that is $19.0 billion. The 7% mandatory convertible preferred, which trades as depositary shares, is a separate claim: 86.25 million depositary shares at $50.16 is $4.3 billion. Bank debt and convertible notes were $8.6 billion at 30 Sep (the 10-Q lines sum to $8.65 billion) against $6.9 billion of cash, so net debt is about $1.7 billion. Enterprise value, common plus preferred plus net debt, is about $25 billion. On the midpoint of the fiscal 2027 sales guide, $77 billion, that is 0.33 times sales.
You cannot grow a negative free cash flow and get a sensible answer, and the latest quarter's free cash flow was negative. So we ask a different question. Take the $77 billion guide as year-one sales, hold sales flat after that, earn a constant free-cash-flow margin, discount it at 9%, and let cash flow grow at 3% only after year ten. The margin that makes this worth $25 billion is about 2.4%.
That is the price. Not a failed sales guide. A thin cash margin on a large one, and then not much growth. At a 12% required return, which is a fairer penalty for a levered hardware business if you want one, the margin the price needs is still only about 3.5%. If sales instead grew 10% a year, the 9% model would need only about a 1.2% cash margin. Set against the quarter, the accounting net margin was 8% and the cash margin was below zero.
These are not price targets. They assume the margin shows up as cash across the whole guided year, and they charge today's debt and today's preferred against that value. The gap between the 2.4% row and the 4% row is the mispricing, if you believe even half of the accounting profit can be collected. The gap between the 2.4% row and the 0% row is why the bear is not silly. Leverage makes both tails long.
The bear case, at full strength
Palantir's bear is the price, the customer mix, and the dilution. The product can keep working and the stock can still be the wrong price.
The valuation bear is the table above. You can believe every word of the quarter and still lose money if 96% growth decays to 15% and stays there, because 15% for a decade is not what $376 billion of enterprise value pays for. The business would still be excellent. The stock would still have been the wrong price.
The government bear has been softened and is not gone. US commercial passing US government is a milestone the company is right to mark. Government remains half the company, it is concentrated in the United States (84% of revenue, 10-Q), and a large share of the contracts can be ended for convenience. Remaining performance obligations of $4.9 billion are the noncancelable slice, not the whole future. A change in federal appetite, or a single large program slipping, moves a business that the market is treating as a decade-long compounder. No customer was over 10% of revenue this quarter, which is better than Super Micro's version of the same risk, and one customer was still 27% of receivables.
The dilution bear is real and, this year, smaller than the headline stock-based pay suggests. Stock-based compensation was $319 million in the quarter, 14% of revenue and 29% of GAAP operating income, and $786 million over nine months. That is why the adjusted operating margin is 62% and the GAAP margin is 47%. If you refuse to treat that as free and instead subtract it from the cash flow, our required growth rate goes from about 23% to about 27%. The share count itself has not been exploding. Basic shares outstanding rose from 2,391 million at 31 Dec 2025 to 2,409 million at 30 Sep 2026, less than 1% in nine months. The diluted share count used for earnings, 2,574 million, still sits 7% above the basic count. The overhang is there. The recent issuance has been modest, mostly because the share price is high enough that the same dollar of awards creates fewer shares.
Super Micro's bear is the margin, the cash, the customer, the competition, and a governance file that is still open.
The margin bear is the sequential comparison the company itself printed. Sales went from $11.1 billion to $19.9 billion and gross margin went from 17.5% to 12.7%. More volume did not buy a better gross margin. It bought a worse one. The 12.7% is better than the 9.3% of a year ago, so this is not a collapse to the floor. It is a reminder that AI servers are a pass-through of expensive parts. Operating margin only reached 10% because operating expenses are 2.4% of sales. That is a thin cushion if the gross margin gives back another few points.
The cash bear follows. A business can report $1.6 billion of profit and still consume cash when it has to buy the parts before the customer pays. Some of that is healthy, and the $5.7 billion of deferred revenue is customer money already in the door. What is left over is a $15.1 billion inventory pile and $9.7 billion of receivables against a $19 billion common equity. If growth slows before those balances turn into cash, the profit was a timing difference.
The customer bear is a single number. 44% of sales, and two customers approaching two thirds of receivables. We do not have the name. We do not need it to see the risk. Losing or repricing that relationship is not a rounding error. It is almost $9 billion of quarterly sales.
The competitive bear does not need a market-share table we do not have. Dell and Hewlett Packard Enterprise sell branded AI systems to the same buyers. The contract manufacturers that already build those racks can sell around a branded assembler. Reuters described the quarter as demand for Nvidia-powered systems (3 Nov). A 12.7% gross margin, falling as volume rose, is what that competition looks like when the expensive part of the box belongs to someone else.
The governance bear is documented, and it has two sides. Ernst & Young resigned as auditor on 24 Oct 2024, writing that it could no longer rely on representations from management and the audit committee (company 8-K). A special committee, with outside counsel and a forensic accountant, later said the evidence did not raise substantial concerns about the integrity of senior management or the audit committee, and that it did not believe the resignation was supported by the facts it reviewed (company disclosure in the fiscal 2024 Form 10-K). Both sentences belong in the file. What is current, not historical: the fiscal 2026 Form 10-K says three earlier material weaknesses were remediated in the year to 30 June 2026, and that one weakness in information-technology controls remained. BDO's report on those controls, dated 31 Aug 2026, was an adverse opinion. The September 10-Q says that weakness was still unremediated at 30 Sep 2026. Management says the statements are fairly presented in spite of it. The Securities and Exchange Commission sent document subpoenas on 19 Nov 2024 and 28 Apr 2026. On 19 Mar 2026 the US Attorney's Office for the Southern District of New York unsealed an indictment of three people employed by or associated with the company, on alleged export-control violations, and sent the company a further subpoena. The company says it is cooperating. That is not a restatement. It is an open file. On 15 Oct 2026 the company appointed Mark Long, formerly of Western Digital, as chief financial officer, succeeding David Weigand.
One structural fact sits underneath the governance point. Ablecom, whose chief executive is the brother of Super Micro's chief executive, manufactured about 97% of the chassis the company bought in the quarter (10-Q). Charles Liang and his spouse owned about 10.5% of Ablecom. Related-party purchases were $198 million, small next to $17.4 billion of cost of sales, so this is not where the gross margin went. It is a reminder of who the company is.
The dilution that matters at $29 is the preferred, not the convertible notes. The notes, $700 million due 2028, $1.725 billion due 2029 and $2.3 billion due 2030, convert at $61.06, $83.44 and $55.20. None were eligible for early conversion at 30 Sep. They are debt at this share price. The preferred is different. It converts on or about 1 June 2029. The initial price is $27.50 and the threshold appreciation price is $33.00. At $28.76, between those two, each depositary share converts into $50 of common stock, about 1.74 shares. Across 86.25 million depositary shares that is about 150 million new common shares, roughly 23% of the basic count (our calculation), on top of a 7% cash dividend that cost common stockholders $76 million of earnings this quarter. The company's December-quarter earnings guide assumes 753 million diluted shares. That count uses accounting rules that can include the notes even while they are out of the money. The economic dilution you can see today is the preferred.
What would change our mind
For Palantir, the view that the price demands a long glide path changes when that path gets easier in the numbers. One more strong quarter, on its own, does not do it. US commercial growth staying very high for several more quarters, with the GAAP operating margin near the mid-40s and the basic share count still barely rising, would mean the early years do more of the work and the downside in the table shrinks. That would change how asymmetric the price is. It would not make 23% for a decade an ordinary assumption.
We would treat the quality of the business as broken, which is a different claim from the price being full, if US commercial and US government decelerated together, if net dollar retention fell toward 100%, or if the diluted share count started to rise several percent a year while growth slowed. None of that is in this quarter. Bookings growing much more slowly than revenue, for a second and third quarter, would be the early version of it.
For Super Micro, the asymmetry disappears if the cash flow keeps refusing to show up. A gross margin sliding back toward last year's 9%, another quarter of negative operating cash flow after growth has cooled (so it is no longer a working-capital story), the 44% customer shrinking in a way that takes the guide with it, or the control weakness turning into a restatement, would mean the price was right to assume failure. A positive operating-cash-flow quarter with the gross margin stable at or above 12.7% would do the opposite. It would say the 8% accounting margin is starting to be collectable, which is the entire gap between the 2.4% row and the rows above it.
Where the terminal stands
The live board is a timing and quality machine. This note is a price-versus-expectations argument. They are allowed to disagree, and right now they mostly agree on timing.
As of 9 Nov 2026, 07:19 UTC, with the data marked fresh, both names are Wait. Palantir: conviction 70, timing 36, under a regime cap of 48. Fundamentals, on a snapshot dated 1 Oct 2026 that predates this quarter, are quality 8.8 and value 2.7. God Mode, the multi-timeframe technical read, is bearish, medium. The technology sector (XLK) is lagging. The defense theme is leading, and Palantir is lagging that group. Market health is unhealthy, at 46. The board's indicative first-add zone is 153 to 159, and the 6 Nov close of $160.38 sits just above it. The stock is also 16% below the 1 Oct close of $190.04 (Yahoo Finance), so the stale value score has not been refreshed for either the drawdown or the quarter.
Super Micro: conviction 73, timing 29, same regime cap of 48. Fundamentals on a 29 Sep 2026 snapshot, also before this quarter, are quality 2.8 and value 6.6. God Mode is bearish, medium. The infrastructure theme is lagging, and Super Micro is ahead of that group. The 6 Nov close is 30% below the 29 Sep close of $41.02 (Yahoo Finance). A value score taken at $41, before a quarter with $1.59 billion of net income, is not the number we would use today. We are not recomputing the board. We are saying the snapshot is old on purpose.
Both verdicts are Wait because the regime is poor, not because the two prices mean the same thing.
The ranking
We rank by how asymmetric the payoff is versus what the current price already assumes, over a 2 to 5 year horizon. Quality is a different list. On that list Palantir is in a different league, and the terminal's 8.8 versus 2.8 says the same thing even on stale snapshots. Timing is a third list. The board says Wait on both, and this ranking does not overrule it.
- Super Micro. The price is already set for the profit to disappoint. About a 2.4% cash margin on the guided sales, and then little growth, gets you to today's enterprise value. A 4% cash margin, still only half of this quarter's accounting net margin, and with no growth at all, is about 90% higher on the common in our model. The cost of being wrong is severe, because debt and a $4.3 billion preferred sit in front of the common, one customer is 44% of sales, and the cash flow this quarter was negative. That is a wide, levered distribution. It is the more asymmetric mispricing of the two.
- Palantir. The business just did the thing the bull describes, and the price already pays for a long version of it. About 23% free-cash-flow growth for a decade matches the price. A few explosive years and then 15% also matches it. A clean 15% decade, which would still be a fine outcome for a company of this margin, is about a 40% hole. Beating the embedded path by a couple of points is worth closer to 15%. At this price the skew runs toward disappointment if the fade is ordinary, and the quarter being excellent does not close that gap. It is what the gap is made of.
What we are watching
In the order of how much each print would tell us:
- Super Micro's next gross margin and its operating cash flow. Stable or better gross margin and positive cash flow is the evidence that the 8% net margin is collectable. Another drop in the gross margin, or another cash outflow once sales stop accelerating, is the evidence that 2.4% was generous.
- Whether one customer is still 44% of sales, and whether receivables stay concentrated in two accounts.
- Palantir's December quarter against the $2.65 billion guide, and whether US commercial can grow again at anything like 29% sequentially. A second quarter of contract value growing at a low double-digit rate, while revenue grows much faster, would mean the company is recognizing old deals faster than it is signing new ones.
- Palantir's basic and diluted share counts. The dilution bear is quiet as long as basic shares creep up by about 1% a year. It gets loud if that changes.
- The open Super Micro file: any update on the IT control weakness, the SEC subpoenas, or the export-control matter, and whether the new CFO changes the speed of the close. The 3 Nov release was still labeled preliminary. The 10-Q then matched it. A clean next close would be information.
Bottom line
Palantir is priced for a decade that stays excellent. The September quarter was better than that path, and the price already assumes several more years in that neighborhood followed by a glide path that does not break. In our model, 15% free-cash-flow growth for ten years leaves the equity about 43% lower, and 25% leaves it about 17% higher. Super Micro is priced for the profit to stay thin and the cash to stay scarce. The same model says a 2.4% cash margin on a flat $77 billion of sales matches the price, and a 4% cash margin, with no growth, is about 90% higher, with a real chance of a much worse outcome if the cash never arrives.
On asymmetry versus what is already in the price, we rank Super Micro, then Palantir. On the quality of the business the order is the other way round, and on timing the terminal says Wait for both in an unhealthy market. The risk attached to the ranking is that Super Micro's right tail is only collectable cash, and this quarter did not collect it.