Amazon’s Friday close is not the September print those older bull and bear pairs used, and the company has not posted the earnings date the tape is trading. From the Nasdaq official close of $248.00 on 23 October 2026, the 47 sessions through 31 December 2026 put the bull at $320.39, the base at $249.05, and the bear at $210.55. The base is $1.05, or 0.42 percent, above the close because the average daily log-return over the last 90 sessions is almost zero. The range is a volatility band, not a drift call.
The missing piece is the issuer’s calendar, and the sentence Andy Jassy published. FinanceFeeds wrote on 24 October that Amazon reports on 29 October, and Nasdaq’s quote page lists that date. The investor-relations event feed, fetched 25 October, does not list a third-quarter call. The latest event is the 30 July call. His Friday status does not say capacity is sold out. It says AWS Continuum is generally available.
Key facts
- Friday 23 October 2026 official close was $248.00, down $2.38, or 0.95 percent, from Thursday’s $250.38. Volume was 25,569,040. Nasdaq daily history, saved 25 October 2026. Nasdaq.
- The 20-day average is $254.43 and the 50-day average is $255.65. Friday is $6.43 under the 20-day line and $7.65 under the 50-day line. Same file.
- The prior five sessions, 16 October through 22 October, ranged from a $265.34 high to a $248.00 low. Friday’s low was $246.45. Friday’s high was $250.01.
- Year-end maths on the last 90 log-returns: bull $320.39, base $249.05, bear $210.55. N is 47 sessions through 31 December 2026. The 31 July jump, up 15.32 percent, is 38 percent of the squared deviations and stays in the sample.
- Second-quarter AWS net sales were $42.232 billion, against $30.873 billion, up 36.8 percent. AWS operating income was $16.621 billion. Form 10-Q for the quarter ended 30 June 2026. SEC filing.
- Trailing twelve-month free cash flow was an outflow of $7.604 billion at 30 June 2026. Operating cash flow was $161.403 billion. Net property purchases were $169.007 billion. Same 10-Q.
- Third-quarter guidance as of 30 July 2026 is net sales of $197.0 billion to $202.0 billion and operating income of $22.5 billion to $26.5 billion. Earnings release.
- Nasdaq lists an earnings date of 29 October 2026. Amazon’s event feed does not. The SEC submissions file shows no 8-K after 14 September 2026.
What just happened, and why the obvious reading is wrong
The obvious reading is a cloud stock into a known print, with capacity already sold and a $350 target sitting above a $248 close. The week does not hold that shape. Monday 19 October closed at $262.45, up 4.98 percent from Friday 16 October at $250.00. By Friday 23 October the close was $248.00, which is 5.51 percent under Monday and 0.80 percent under the prior Friday. The upgrade session was given back.
Benzinga and RedCandleRx wrote on 20 October that Morgan Stanley had named Amazon a top pick and raised a price target to $350 from $335. RedCandleRx named analyst Brian Nowak and a 2027 AWS growth figure of 47 percent. The Monday close in the Nasdaq file matches the $262.45 those posts cite. The note itself was not opened, and the $350 is not the bull in the headline.
The capacity line has the same flaw. On 24 October an account wrote that Jassy had said 2026 capacity was sold out, that 2027 was largely sold out, and that part of 2028 was gone. His 23 October status at 14:12 UTC does not say that. A paraphrase is not his sentence.
The date is a calendar entry, not an issuer fact. FinanceFeeds wrote on 24 October that Amazon reports on 29 October, after Microsoft and Alphabet, and Nasdaq’s saved quote file carries “Earnings Date : Oct 29, 2026.” The event feed’s latest earnings event is still the 30 July call, and EDGAR’s last 8-K is 14 September 2026. The quote page timestamp still reads 22 October, while the daily file’s completed 23 October bar at $248.00 is the close, and a weekend after-hours $249 is not.
Who is on the other side of the book
The commitments that move the equity are in the 10-Q. In the first quarter of 2026, AWS and OpenAI expanded an existing $38.0 billion commitment by $100.0 billion over eight years, including chips. In the second quarter, AWS and Anthropic expanded their collaboration by more than $100.0 billion over ten years, again including chips. Those lines are contracts. They are not a statement that the racks are already on.
The cash Amazon put in is separate from those commitments. It invested $8.0 billion in Anthropic notes from 2023 through 2025, then $10.0 billion of preferred stock in the second quarter of 2026, which cut a $20.0 billion facility to $15.0 billion. OpenAI Series C preferred was $28.7 billion over six months, and after 30 June Amazon funded a remaining $21.3 billion.
The release also names Uber and Pinterest on Trainium. FinanceFeeds reported on 8 September that Vistra has 4 gigawatts of nuclear under a 20-year contract to Amazon and Meta. That is the power contract on the other side of the buildout. Jassy’s Friday post names IBM, Robinhood, Capital One, Infor, and Luminate as Continuum preview users. A security product going generally available is not a capacity disclosure.
The 10-Q shows 10,783 million shares outstanding at 30 June 2026. At $248.00 that June count is worth $2.674 trillion. The August page on the $3 trillion club was a different print. This close, on that share count, is not in the club. The count is not a 23 October figure.
What the tape and the filings show
The series high in the Nasdaq file is $287.20 on 3 August 2026, when the close was $284.02. The series low is $196.00 on 17 February 2026. Friday’s $248.00 is 13.6 percent under the August high and 26.5 percent above the February low, and it sits under both moving averages.
International net sales were $42.197 billion, almost the same dollars as AWS at $42.232 billion. Operating income was not. AWS produced $16.621 billion, International produced $1.717 billion, and North America produced $9.123 billion on $116.177 billion of sales. The margins are 39.4 percent, 4.1 percent, and 7.9 percent.
| Line | Q2 2025 | Q2 2026 | Source |
|---|---|---|---|
| Consolidated net sales | $167.702 billion | $200.606 billion | 10-Q, filed 31 July 2026 |
| AWS net sales | $30.873 billion | $42.232 billion | 10-Q |
| AWS operating income | $10.160 billion | $16.621 billion | 10-Q |
| Operating income | $19.171 billion | $27.461 billion | 10-Q |
| Net income | $18.164 billion | $62.647 billion | 10-Q |
| Other income, net | $1.117 billion | $53.415 billion | 10-Q |
The release rounds AWS to 37 percent and $42.2 billion. Jassy’s sentence says 36.7 percent. The 10-Q ratio is 36.8 percent. The gap is rounding. The same release says the AI business and the chips business each exceeded a $25 billion run rate, which is not a 10-Q segment line.
Bull, base, and bear
The headline numbers are the distribution maths on the Friday close. They are not a Morgan Stanley target and not a copy of an older FinanceFeeds pair. The inputs are the last 90 daily log-returns, from 16 June 2026 through 23 October 2026. The mean, mu, is 0.000089969. The sample standard deviation, sigma, is 0.024494728. N is 47: the sessions from 26 October through 31 December, skipping only Thanksgiving on 26 November and Christmas on 25 December. The early closes on 27 November and 24 December count. Five October sessions, 20 November sessions, and 22 December sessions are 47.
Base equals $248.00 times e to the power of 47 times 0.000089969, which rounds half-up to $249.05. Bull equals $248.00 times e to the power of 47 times mu plus 1.5 times sigma times the square root of 47, which rounds to $320.39, or 29.19 percent above the close. Bear equals $248.00 times e to the power of 47 times mu minus sigma times the square root of 47, which rounds to $210.55, or 15.10 percent below the close. The bull is above the spot. The bear is below it. The drift is not so negative that the bull has to be rebuilt without it.
One session dominates sigma, and it stays in. On 31 July the close was $271.58, up 15.32 percent from $235.50 on 30 July, the session after the release. That day is 38.0 percent of the sum of squared deviations. Dropping it would cut sigma from 0.02449 to 0.01933. That is not done. The July jump is why the bull can reach $320.39 while the average daily return is a rounding error.
Checked against printed English pairs of $325 and $210, $340 and $190, $330 and $185, $370 and $207, and $395 and $205. A bull within $1 of any of those bulls would have been rejected even if the bear differed. $320.39 is $4.61 from $325. The bear at $210.55 is 55 cents from the old $210, and that pair collides only if the bull is also within $1. It is not, so the title stays the year-end pair. A ten-session path, which was not required, would have been $278.81, $248.22, and $229.72.
| Case | Price | Versus $248.00 | What has to be true |
|---|---|---|---|
| Bull | $320.39 | +29.19 percent | The next 47 sessions realize the drift plus 1.5 sample standard deviations, which needs days in the family of 31 July. |
| Base | $249.05 | +0.42 percent | The next 47 sessions match the last 90 days’ average log-return and nothing else. |
| Bear | $210.55 | -15.10 percent | The next 47 sessions realize the drift minus one standard deviation. A flat Anthropic mark plus a lower multiple fits this better than a retail miss alone. |
Benzinga wrote on 20 October that analyst Brian Nowak at Morgan Stanley framed a 3-to-1 setup with a $220 bear and a $410 bull, and that the firm’s target was $350, up from $335. $320.39 sits under $350 and under $410. $210.55 sits $9.45 under that $220 bear. The same post said Polymarket traders assigned roughly an 11 percent chance Amazon trades at $300 or higher in October, up from about 4 percent, and roughly a 53 percent chance it falls below $280. An October probability is not a 31 December level.
What Amazon has said, and what it has not
On the 30 July release, Andy Jassy, president and chief executive of Amazon, said AWS was “growing 36.7% year-over-year in Q2” and that the AI and chips businesses had “each eclipsed run rates of more than $25 billion.” He also cited advertising growth of 26 percent. That is the last set of words the company has put next to a quarterly number. It is three months old.
The only official forward band fetched here is the 30 July guidance. Third-quarter net sales were expected at $197.0 billion to $202.0 billion, growth of 9 to 12 percent, and operating income at $22.5 billion to $26.5 billion against $17.4 billion a year earlier. Excluding Prime Day, the release says growth would be nearly 400 basis points higher. A result inside the guide can be down from the second quarter’s $27.5 billion of operating income and still ahead of last year.
What the company has not said, in the pages opened here, is the date of the third-quarter call. The event feed stops at 30 July, and the 8-K index stops at 14 September. Jassy’s 23 October post does not mention earnings or a capex dollar. The sentence is: “AI has taken the gap between finding a vulnerability and someone using it from weeks down to minutes, leaving teams scrambling to keep up.” He then wrote: “That’s why we built AWS Continuum, now generally available.” A $220 billion capex figure from September pieces is not in the release or the 10-Q. The 31 July FinanceFeeds headline used it. A $190.4 billion Anthropic carrying value is not in this 10-Q, and it is dropped.
The cash and the mark
Free cash flow looks like a warning and is mostly a capex identity. For the twelve months ended 30 June 2026, operating cash flow was $161.403 billion, up from $121.137 billion. Property purchases, net, were $169.007 billion, against $102.953 billion, so free cash flow was an outflow of $7.604 billion against an inflow of $18.184 billion. The release ties the swing to a $66.1 billion increase in those purchases.
The earnings line has the opposite shape. Second-quarter net income was $62.647 billion, or $5.75 diluted, against operating income of $27.461 billion. Other income was $53.415 billion, of which the upward adjustment on private-company equity was $50.486 billion in the quarter, primarily the Anthropic preferred. The six-month tax provision included $15.9 billion of net discrete tax expense from those adjustments. A mark that large can reverse. It is not AWS revenue.
Two balance-sheet lines should not be added. At 30 June 2026 the Anthropic notes had an estimated fair value of $97.9 billion, against $45.8 billion at 31 December 2025, with an unrealized gain in other comprehensive income of $92.0 billion. Private-company equity not on the equity method, primarily Anthropic preferred and OpenAI preferred, had a carrying value of $122.3 billion, against $16.2 billion. The filing says an IPO or liquidity event would convert the notes and the preferred to nonvoting common, subject to an ownership cap. FinanceFeeds wrote on 17 October that some investors discuss $1.5 trillion for Anthropic rather than $2 trillion. That discussion is not a conversion date.
What happens next
Through 31 December 2026 the working path is $320.39, $249.05, or $210.55. Mu is about 0.00009 a day, so the base stays near Friday’s close unless the next 47 sessions average something else. The bull needs another day in the family of 31 July. The bear needs the volatility to print on the downside while the multiple stops paying for a mark that already hit income.
If a third-quarter release lands on the Nasdaq date, the test is the July guide. Sales of $197.0 billion to $202.0 billion and operating income of $22.5 billion to $26.5 billion are the company’s band. A result inside it can still disappoint a holder anchored on second-quarter net income of $62.6 billion, because that number was the mark. The same formula over four sessions through 29 October is a $267.01 bull, a $248.09 base, and a $236.23 bear. That checkpoint is not the title. The event feed does not list the call.
On Monday 26 October the live test is the event feed. If it is still empty of a third-quarter call, holders are using a Nasdaq date Amazon has not published. A print of $300 this month is a different contract from the $320.39 year-end bull.
Frequently asked questions
Where do the sessions through 31 December 2026 put a fresh Amazon bull and bear?
The bull is $320.39 and the bear is $210.55, with a base of $249.05, from Friday’s $248.00 Nasdaq close. The inputs are the last 90 daily log-returns, a sample standard deviation, and 47 sessions that skip only Thanksgiving and Christmas. The base is $1.05 above the close because the average daily return is nearly flat. These are distribution levels, not a bank’s target.
Has Amazon confirmed a 29 October earnings date?
Not in the documents fetched for this piece. Nasdaq’s quote page lists 29 October 2026, and a FinanceFeeds tape on 24 October used that date. Amazon’s investor-relations event feed, fetched 25 October, still shows the 30 July second-quarter call as the latest earnings event. The SEC submissions file shows no 8-K after 14 September. Until the company posts the call, 29 October is a calendar entry, not an issuer fact.
How is this pair different from the September FinanceFeeds calls?
The 29 September page used a $246.67 close and printed $325 and $210. The 14 September page used a $253.54 close and printed $340 and $190. This pair starts at $248.00 on 23 October and uses a fresh 90-day volatility, not those targets. The new bull, $320.39, is more than $1 away from $325, so it is not a collision with the old bull. The bear can sit near $210 without copying that page.
What did Andy Jassy post on 23 October?
He wrote that AI has cut the gap between finding a vulnerability and someone using it from weeks to minutes, and that AWS Continuum is now generally available. He named IBM, Robinhood, Capital One, Infor, and Luminate as preview customers. The status does not say capacity for 2026, 2027, or 2028 is sold out. Posts that attribute a sold-out line to him are paraphrases, and they are not quoted here as his words.
Why is free cash flow negative if operating cash flow rose?
Because property spending rose by more. Over the twelve months ended 30 June 2026, operating cash flow was $161.403 billion, up from $121.137 billion. Purchases of property and equipment, net, were $169.007 billion. Free cash flow, operating cash minus those purchases, was an outflow of $7.604 billion. The release ties the increase in purchases primarily to artificial-intelligence investment. The cash engine did not shrink. The buildout did.
Is the $350 target the same number as this bull?
No. Benzinga and RedCandleRx wrote that Morgan Stanley raised a target to $350 from $335. Benzinga also described a Nowak bull case of $410 and a bear case of $220. This bull, $320.39, is the 1.5-standard-deviation path through 31 December on Friday’s close. It is below those street bulls. The note those posts describe was not opened, so $350 stays a reported target, not an input.
This is not financial advice.
