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.
Thanks Sam. What’s your take on the second leg lower? Do you see it potentially staying down a long time? What’s your rough guess as to when we’d likely return to all time highs? April? Summer? Fall? Just ballparking my own mental planning. Thanks
You can extrapolate all of that from the correction table. Rebound should take no more than 5-days if it’s two legged. The second leg should be substantially shorter in duration than the first leg.
After that the rally back to the highs can take 25-50 sessions. The longest rebounds we had in the correction table were 60-70 sessions. But that was also for 20-30%+ rallies .
So I think 25-50 or 5-10 weeks to get back to the highs post correction.
Take a look at be corrections tables in the outlook. You can see how long it took for past post-correction rallies to reach their peak.
Note we haven’t even really started to rebound yet. If we don’t get a push back up to at least 492 493 then today’s rebound doesn’t count for anything.
Hi Sam – A push up to 492 would be a roughly +4.5% intraday. Why is such a large daily gain necessary for this rebound to have started?
I’m working on a way to make the tables sortable to make it a lot easier to look a various aspects of the correction.
Hi Sam, could you explain why gapping up doesn’t make much sense as a bottom?
Nearly all corrections end with a reversal of some kind (or at least a complex bottom like an inverse head & shoulders or double-bottom).
Reversal Days
A reversal day occurs when the market goes down hard on the session and then reverses course to erase the losses. That can happen in a number of different ways:
(1) Gap Down at the open, followed by a sell-off, followed by a full reversal of either ALL the losses or MOST of the losses. We either close barely down or green after a deep red day.
(2) Open flat, sell-off hard and then reverse hard to close out slightly red/flat or green.
(3) Gap Up, like today, sell-off throughout the day to give up all the gains and even go red, reverse course again, and close the day deep green at the highs. <– rare.
^All of these things give confirmation to traders that the market has tested a critical level and the investors have uniformly decided that at such prices, the market is objectively undervalued. The supply/demand curve shifts to the buy side.
But days like today where we just randomly gap up don’t really work. Especially after a long 13.64% correction.
Complex Bottoms
Another way the market bottoms is through some sort of complex bottom where we form an inverse head & shoulders, double-bottom or some other chart pattern. That has worked as well. That used to be the way to go actually. But lately — over the last decade maybe — the market has shifted to capitulation and reversals to bottom the market.
It used to be that you needed a complex bottom for the market to bottom. We never saw a low without a double-bottom or inverse head & shoulders at play.
If you go back to older corrections, you’ll see a lot of that going on. The more modern corrections tend to be v-shaped.
This valuable info. Perhaps you could find a place for this in your training material.
We got to about 472 just now, so maybe option (3) has occurred if we go up throughout the day?
Very possible.
Hi Sam – once the market closes today, would be helpful to understand what you think happened the last 3 days (market sideways) and get some commentary on reversal day options 1-3 above based on what you are seeing. Please and thanks!
Precious explanation.
Thanks man!
Could you expand a little on the gap on lows narrowing from the chart. I am a little confused as what I’m seeing. Just want to expand my ability to make complete sense of the charts you give us.
It wasn’t clear for me either when Sam wrote about this the first time a few days ago. By gaps he meant the vertical gap (price).
Another measurable consequence of the same reality is the RSI making successives higher lows.
I mean when you look at the distance between the lows on the chart, they were shrinking. For example, the distance from $509.43 to $496.93 is $12.50. That’s the distance from one low point on the chart to the next low point on the chart.
Now if you look at the distance between the next two low points — $496.93 and $487.73 — the difference there is only $9.20. Meaning, we only made a $9.20 NEW low after the rebound cycle.
Think logically here. Without even considering forward selling momentum at all, what happens logically if each time we make a new low, it shrinks and gets smaller and smaller.
Eventually, we get no new low at all right? Like if the distance between low points goes from $12.50 to $9.20 to $7.00 to $4.00 to $1.00, basically we’ll eventually shift to the other side.
Now when you see the distance between lows contract, it’s an indication that the overall selling momentum is slowly shifting to the buy side. It was contract all the way until the most recent leg down.
But even recently, a $13 new low isn’t that great to be honest. We went from a $480 low to a $467 low essentially (2.7% new low). That’s not a huge number at all. It’s still largely a momentum problem for the bears.
I guess the most recent low of $13 was throwing me off most as I thought it was bucking the trend. Interesting to know the previous contraction supersedes this outlier.
Hi Sam,
Would love to hear how you came to this price range for NVDA in your article:
Are you just projecting NVDA will retrace 50% of the correction drop (143 to 104), similar to the QQQ? Since NVDA has a higher beta against the market can we expect it to retrace more of the losses compared to QQQ? If possible, would love to hear your thoughts on how you project price targets for individual stocks in relation to the QQQ (e.g. if QQQ rebounds to 505 we expect NVDA to be at X price).
Thanks!
Yes of course. Nviida is going to track way way ahead of the market. Of course it should retrace 50% or more even. Nvidia has a much higher beta. That’s why it’s up far more today relative to the QQQ. My expectation is $125-$130. All along we’ve been in the $130 camp. That’s where Nvidia tends to run into resistance.
Hello Sam, From what you wrote about the $120-125 range for NVIDIA, I understand that NVIDIA could drop below $106, as this week saw a drop of more than $15. Previously, I understood that NVIDIA couldn’t retest the $99 support level reached in August. Did I misunderstand? Especially since your long position on Tarly and Arryn is at $100. In the case of a purchase at $106, the risk of selling at $120-$125 to recover below $106, isn’t $100 more important, especially given the catalysts in May with the AI and robotics conference with Jensen, where we know that if he speaks, the market listens, as was the case previously in Las Vegas with the drop in quantum computing and the half-yearly financial statements at the end of May. Thank you for your clarification.
So in august it reached $90. Unless we have a massive correction, Nvidia will never get down there.
Today Nvidia tracked way ahead of the NASDAQ-100. The NASDAQ-100 has barely moved off of its lows and NVDA is way way off of its lows.
So even if we get back to the lows int eh market, I don’t think Nvidia does. Nvidia may have already bottomed.
Thanks Sam but why do you want to sell at 120-125$ and buy more when it goes down ?
Hi Sam,
Can you elaborate on the momentum / psychology / trading(?) principles underpinning this statement:
If we’re expecting a rebound and then second leg lower then why can’t this type of behavior count towards that rebound?
Here is my understanding / take and maybe you can correct me:
The rebound is suppose to represent a potential easing up of negative sentiment and cause for the market to think the correction has ended. If the rebound is a swift upward movement with a quick peak and pull-back then:
Market participants will register this as “the market is still volatile”Market participants waiting for confirmation are not registering this move up as valid since the time spent trading in the range is smallNot enough market participants with poor positions (i.e “bag holders”) will have been flushed outThe hourly & daily RSI probably won’t look robust enough to constitute as a “rebound”Historical data doesn’t support this type of price action being a true “rebound” (?)
Although, the mechanics are still not super clear to me so I’d like to understand. Please make me understand 🙂
Not sure why my bullet points keep getting removed. I outlined my understanding through bullet points and they disappeared for some reason. Apologies if that part of the comment reads weird, just interpret abrupt joining of words as a bullet point.
That’s all correct. During the rebound, there’s always a sense that we could be wrong about the second leg down. I will personally question it. Once I get to that point where we’re selling but unsure if we’re going to pull-back and where we see no less than let’s say 5 comments of people asking me “are you sure we’re going to see a second leg down. What if we just hold through it? What if we just sell 1/3 instead.”
Subscribers will question it. That’s because we’ll have rebounded for 5-days and the correction selling will feel like ancient history. Time dilation will work in the opposite direction. And the BAM! Rug pull. That’s how it tends to go.
The rebound should be strong enough and long enough that people think that we’ve bottomed. A single day isn’t really enough for that.
Although a 2-day rebound where we get enough points certainly could count for it. We’ve seen it before. But we’d need enough forward momentum where the QQQ climbs above $500 a share.
Imagine a situation where we climb above $500 in just 2-days. T hat could easily get market participants pulled into the “correction is over” routine.
The smallest rebound I’ve seen in a 2-legged LARGE correction is like 5%. That would be equivalent to $491 on the QQQ. That’s the smallest rebound in a two-legged correction.
In a smaller correction of 6-8%, the rebound would be smaller, but the retracement would still be closer to 505. So there you might get a 3.5-4.5% rebound and a 50% retracement.
Follow up question: If our goal is to substantially unload our long positions at 505 because we think that’s where the market will peak (i.e. 50% retracement) then why wouldn’t we want to just do that if the market were to quickly jump up to 505 and then pull back, regardless of whether we treat that as a “proper” rebound? Is that trying too hard to “time” things?
We would. In fact, we’ll start lightening up way before that point. if we get up to the high $480’s or low $490’s, we’ll start selling positions.
If you’ve noticed, until recently, we’ve traded in and out of positions on each of the rebounds until the one last Friday. The rebound from $480 to $493 was the ONLY rebound we haven’t reduced positions into. We’ve reduced and traded every other smaller bounce. We’ll do that here as well.
That table is badass! LOVE that you can interact with it!
Thanks! Glad you like it. we’ve added it to its own dedicated page and are working on the visual aspects of it. So far we’ve made it compact. So far so good. I have a lot of ideas now on the types of data we can add to the site.
In response to the latest update: The correction table is looking good on my desktop!
We were able to make it compact. So it’s looking good. this might be the way to go with future tables for sure.
Important that we are distinguishing “correction bottoms” from “leg bottoms”.
The table above is all correction bottoms. Periods where ether market bottom and then rallied nearly 10% off of its lows. I began working on legs, and just through 2019, we surprised the 47 data points we have here going back to 2010. doing a correction legs analysis will take forever to complete and will have a lot of data points.
Which in teh end, there’s not a lot of value to for this correction because there aren’t many single legs that surpass this one. there really aren’t. the ones that do exist are covid, bear market and things like that.
what do you think that means about this one? trying to avoid “this time is different” thinking but also – is it?
Veeeeee !
Where was $NYMO at the low?
I think around -60. Not that low.
https://www.tradingview.com/symbols/SEED_DANWAGNERCO_INDICATORS-NYMO/
I think this data is delayed, 1 day old.
Oh that’s no good, does anyone know where to access the current NYMO chart for free?
Sam, QQQ filled the gap from the morning and rebounded. Would you now consider this as a “true” rebound?
Possibly. I still think we’re in for a second leg lower after a properly rally back to $500.
so you mean going back to 465 to retest the lows last time before exploding massively?
Not sure how low it will go after the rebound. that will really depends on how far the rebound goes when it actually happens.
Is this something you’ve tried doing by playing around with using stock data API services (e.g. AlphaVantage, Polygon, etc)? A little unfortunate they charge so much for 10+ year of historical data as that’s what we would need to do this.
I haven’t. But we’re often working with very specific parameters. We currently use Polygon.io for the portfolio tables.
That said, you don’t even need Polygon.io for this exercise—it can be done quickly using StockCharts.com. You just identify the key dates, like Zephyr did below, and compare them directly to past corrections.
As I mentioned, some of these rally durations don’t make sense because I’ve often posted the absolute high point instead of selecting the best high point. In many cases, 85% of the rally is completed within the first 25% of the total “rally duration.”
One thing I’ve done before in data tables like this is track how much the QQQ rallied in the first 10, 20, 30, and 40 days after a correction. I may do that again.
For example, we’d want to identify the highest price on the QQQ within 10 days of the bottom, 20 days, etc.StockCharts.com allows you to download all the data to Excel, and I’ve used that data with ChatGPT to pull out trends.
But honestly, you can eyeball a lot of it just by using StockCharts.com directly. You add price labels and then I use start and end dates within a 1-year period and keep moving it back each year until we catch every correction. It takes hours of work, but then one it’s done, it’s done
Purely based on the table, I can see why there could be a second leg and why a June 2025 and possibly May 2025 expiration would make sense. I filtered your table (attached image) to answer some exercises…
1. Identify all rallies that lasted 50 days.
◦ Jan 19, 2010
◦ Aug 9, 2010
◦ Nov 9, 2010
◦ Oct 27, 2011
◦ Apr 3, 2012
◦ Mar 7, 2014
◦ Dec 29, 2015
◦ Mar 13, 2018
◦ Oct 1, 2018
◦ Oct 12, 2020
◦ Feb 2, 2023
◦ Mar 21, 2024
◦ Feb 19, 2025
2. Determine the SMALLEST rally observed among those 50-day rallies.
◦ Mar 13, 2018 @ 15.91%
3. Count how many rallies exceeded 50 days in duration.
◦ 12
4. Identify specific corrections for rallies that lasted longer than 50 days and analyze those specific corrections.
Gives an average of…
◦ Pct. Loss: -10.817272727273%
◦ Trade Days: 24
◦ Bottom day: Likely a Monday or Friday
◦ Bottom Date: 4 of 11 beginning of month, 4 of 11 mid of month, 3 of 11 end of month, and very slight preference to November and April
◦ Rally % Gain: 22.096363636364%
◦ Legs: 2.0909090909091
5. I noticed that only some requirements were hit with the current correction…
◦ Current Pct. Loss:-13.65%
◦ Current Trade Days: 15
◦ Current Legs: 1
Which means that we most likely have another leg to go on average and this adds to ‘trade days.
So, looking at the sample size for 50 trade days and Pct. Loss less than -10% is too small for me.
Looking at an average of -13.64% Pct. Loss over 10 data points:
Gives an average of…
◦ Pct. Loss: -13.64%
◦ Trade Days: 24.1
◦ Bottom day: Likely a Monday
◦ Bottom Date: 3 of 10 beginning of month, 4 of 10 mid of month, 3 of 10 end of month, and slight preference to September
◦ Rally % Gain: 14.748%
◦ Rally Duration: 39
◦ Legs: 2
Which means that we most likely have another leg to go on average and this adds to ‘trade days.
Conclusion:
Combining these two analysis (4. and 5. above), 2 legs is the norm with Rally % Gain: 14.748% to 22.096363636364% after the correction is done, and an average rally of about 39 days or higher which would put us on a day before May 2025 or June 2025 expiration.
A 14.748% rally gain from the bottom would put us close but below QQQ highs (~536) so 22.096363636364% would be an extra nice gain to go past All Time Highs.
nice work @zephyr!
Nice job!
With these numbers and timelines, we might not quite profit from the QQQ 540/550 before the May. Hope the losses can be offset by the QQQ 525/535, and hoping we can hold that all way to expiry ????
Could today count as the beginning of a rally or does the fact that we closed lower than open (red candle) invalidate that?
Today is not a particularly positive day. It’s just OK. I think it could be the start of a multi session. Rebound towards 505. But I don’t think this is how the rally would start.
Yeah, today was a bit frustrating. So if I’m understanding correctly, today can still be the beginning of a bounce towards 500-505, but it’s highly unlikely to be the beginning of a correction bottom, and we should have a second leg lower after the bounce. Correct?
Say the market bounces to 503 and we execute your plan, where would be the values of Baratheon and Targaryen at that moment, approximately?
This may hurt NVDA.
https://www.zdnet.com/article/google-claims-gemma-3-reaches-98-of-deepseeks-accuracy-using-only-one-gpu/
Doubtful. Only because if it was damaging, the stock would already be down already.
The stock market doesn’t move slow slowly. Any news anywhere — it doesn’t matter if it’s on a tiny blog in a tucked away corner of the universe — it will be discovered, and the information verified. The impact would be felt pretty much immediately.
It could potentially be bad news if it ends up being something down the line, but I don’t think the impact is really there now otherwise the stock would be down
And right now we’re green in afterhours
If tomorrow’s PPI report also comes in under expectations it could solidify a rebound. I can see that happening tomorrow.