Samwise Quick Reference Handbook
To streamline our daily blogs and conserve space, we’ve organized key resources into a convenient, collapsible dropdown menu below. A sort of Quick Reference Handbook if you will -- as our friends in aviation might call it. By clicking the menu below, you’ll have qu...
Please login to view this page.
Hey, what do you mean by retest going poorly? That it’s likely to break down?
Also if it doesn’t make new highs, would you count this as day 5 of the correction already? Or is this some kind of consolidation top stuff?
Yes this would be day 5 and this is exactly how every other correction has been counted
When we count the number of days of a correction, we count from the very top of the market to the very bottom
We look at the very top print and count that print as day one
So if a correction takes 25 days but eight of those days the market sits at the highest then the correction is really 17 days but we count is as 25 day correction.
In terms of the cycle, we count top to bottom. We are looking at what is the top most print to bottom most print .
We then look at for rallies. What is the bottom most print to top most print.
So April 7 to whatever day represents the highest print is how long the rally lasted.
Sometimes there can be long consolidations, and we count those consolidations as part of the rally up until the day, it reaches its highest point
Then the second half of the consolidation beyond the highest would be counted as part of the correction
At least that’s how we have it laid out in the tables that you see
Also, when I say the retest is going poorly, I mean the market made an aggressive move to 600 but then failed to take it out so it’s not going well from a bulls perspective.
When the market does inevitably peak, has it been driven by a catalyst in the past, like negative news or macro events? Or just tanks on a random day.
Could it be that the longer a rally is, the more likely a negative external catalyst occurs, which causes the breakdown? In other words, rally duration is determined by likelihood of a negative catalyst occurring? Similar to a geometric distribution where the prob of first success -negative catalyst- cumulative becomes higher as time goes on.
I know you mentioned profit taking and exhaustion in a previous post, but intuitively, this sounds like a slow grind downward. I’m wondering what usually triggers sell offs.
There can be catalysts that shift the market such as the Fed, but historically the market has proven its ability to peak one day and roll over the next without a specific driver. Sam discussed this phenomenon in previous briefings.
The deeper in the money an option goes, the closer to full value it goes. Ideally we want a spread that costs $1.00 to $1.20 because at $1.00, you have a potential 9x return.
The $570-$560 spread with the QQQ trading at $540 a share would be worth $8.00+ easily. But if we can’t buy in that low $1.00 range, then it’s not worth it.
NVDA was lagging behind QQQ for a bit but it’s now at new ATH
NIVIDA to 200! QQQ to 700! LET’S GO!!
It doesn’t look like there’s anything from preventing that from happening. Keeps going up and up. Another all time high
this AI supercyle by the administration is gonna push the bull market even further. Sam has better insight, and this is just my opinion but it feels kinda obvious at this point that everyone knows this is the start of the bubble, and the administration wouldn’t just give a correction or pop the bubble at any moment knowing the fact that they will have to win midterm election. We are far due for a correction at this point. we are still due for one, but not for a while because of midterm election next year. I personally think we are in fact super lucky to have no correction at least for the rest of this year. Lowered rates will print more and more calls
NVDA ATH
what an interesting day for Nvidia. it traded above much overbought in the hourly pretty much the whole day. Is there any implication from this?
Anyone else missing today’s briefing? I’m only seeing yesterday’s. it didn’t update
There isn’t one yet
Everything okay? Not that a post is needed today.
There’s usually a post every day the market is open. I like being updated daily. So you’re the one that decides if “Sam” puts out a daily briefing?
My hunch is Chris was actually addressing Sam as he gets notified of comments. As in, “Everything okay Sam?”, asking him if he is ok since he didn’t post today.
Correct DT. Thanks for clarifying my comment. I’ll try to be more careful. ????
Thanks for clarifying! I interpreted your comment incorrectly
Thanks Derek. I interpreted Chris’s comment incorrectly
Sam deserves a day off.
It looks like Nvidia traded above 70 hourly RSI for almost the whole day since opening. Does this mean Nvidia is severely overbought currently?
According to my charts, it was only over 70 for two out of today’s seven candles. Sam considers severely overbought to be up around 80.
Hello Sam,
A lot of doubts lately, including a major discovery I made.
CoreWeave 8-K Released — New $6.3 Billion Contract with NVIDIA
On September 9, 2025, CoreWeave filed a Form 8-K with the SEC, announcing the signing of a new Order Form with NVIDIA, under the existing Master Services Agreement (MSA) between the two parties.
Key Points:
– Total Amount: $6.3 billion
– Commitment Term: Through April 13, 2032
– Purpose: Provision of Reserved Cloud Services to NVIDIA
Points to Watch:
this contract extends or extends previous cloud delivery agreements between the two companies.
The term “reserved” potentially implies a firm consumption commitment by NVIDIA, but the 8-K document remains vague on the terms (no details on a “take-or-pay” clause, penalties, or guaranteed minimum threshold).
In the absence of published contractual details, it is unclear whether NVIDIA is obligated to use or pay for unused cloud capacity.
Possible accounting consequences:
– NVIDIA will likely record this commitment as an off-balance sheet “non-inventory purchase obligation”;
– This type of obligation can be recognized as an expense only if the service is consumed, or failing that, as a provision if deemed uncollectible;
– This contract could artificially inflate revenue if NVIDIA sells GPUs to CoreWeave at the same time, while reserving cloud capacity with this same customer-turned-supplier.
Conclusion:
This 8-K reinforces the idea of a circular relationship between the two companies. NVIDIA sells hardware to CoreWeave, then commits to using the cloud hosting that hardware.
But without details on the performance clauses, we cannot definitively conclude that there is a firm obligation to consume the cloud, a point that remains central to assessing the quality of NVIDIA’s revenue.
One to watch in NVIDIA’s 10-Q (Q3 FY2026)…
The article https://www.boursier.com/actions/actual … 67445.html discusses the announcement of Nvidia’s $100 billion investment in Open AI and raises the question of circular investments in AI.
NVIDIA indicates that this investment would not go towards direct purchases of its own products.
They’re talking about leasing. Strange, isn’t it hours of computing that Nvidia is credited with?
Here’s a tweet too of Dylan PATEL and MUSK…
I fear a spread with all this financial engineering involving NVIDIA, Oracle, Corewave, and many others—surely also quantum and modular nuclear companies like Oklo and Nuscale—as well.
A small spark could trigger the end of this exceptionally long rally, something unseen in over 20 years—except for the COVID rally, which saw mini-corrections along the way.
With the gap around $450 for the QQQ, this could really sting…
A story to follow and be vigilant in these tumultuous times.
Many sectors are soaring: AI, semiconductors, nuclear, energy, quantum, weapons, drone defenses…
It’s starting to add up, and Berkshire’s signal about buying and selling more down-to-earth sectors raises questions…
Here’s my understanding of your recent posts and the situation:
1. Probable timing of the correction
=> Cycle duration
Day 121 reached (September 29).
Longest rallies without a serious correction:
Covid 2020: 114 days.
All-time high 2006–2007: 151 days (unique case).
The 120–128-day zone is historically the end of a rally corridor → this places the highest probability between September 30 and October 8, 2025.
=> Short-term technical
Weak rebound from $588.50 → double top structure around $602–603.
Current candlesticks = doji + declining volumes → typical of buyer exhaustion.
Daily RSI at 66 → not yet “imminent crash,” but in the late stage (“VERY LATE STAGE” on your chart).
=> Scenario probabilities
Imminent correction (test as soon as this happens): 55–60%.
Lateral movement to $600 → then rejection in early October: 30%.
Exceptional extension towards $620–630 before correction: 10–15%.
2. Expected correction window
=> Probable start: between September 30 and October 3, 2025 (retest failure of $602–603).
=> Active phase: October 7-21, 2025, typical duration of a QQQ correction leg (10-15 sessions).
=> Target zone:
$570–560 (–5% to –7%) = standard correction.
$550 (–8.3%) = major consolidation scenario.
If $550 breaks → extension towards the Major Gap zone $495–505 by Q1 2026.
3. Clear Summary
The correction is not expected in November, but rather between late September and mid-October.
November is already the time when we can expect to see the bottom and prepare for the rebound.
Hello Sam,
Can you give me some feedback?
The opportunity cost seems enormous to me to have finally exited the markets since May/June, when we were expecting this correction.
No one wants a correction with a mid-term presidential election or a capitulation, and everyone is talking about it…
Have you revised your strategy accordingly, and therefore the analysis I summarized above?
Please don’t make personal attacks. It serves no purpose other than making you look like a fool.
The irony!
Is NVDA going to get pulled to the $200 century mark this rally?
$ 230 by the end of the year
Hopefully DieselCock is fine! Probably had to take a day off yesterday and will have an updated briefing today