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.
Hi Sam,
You’ve mentioned markets start corrections when its ready to do so. You’ve mentioned market corrections sometimes start by just a cascade of sell-offs that lead into more selling and suddenly we’re in a correction. I’ve always been confused as to what starts these cascading sell-offs, if not by way of external events. Jokingly, do a group of hedge fund managers one day decide its time to take profits and that triggers more selling? Although the eventuality of the correction happening is not random, sometimes the trigger that starts it seems to be more random? Hopefully that makes sense ????
Maybe phrased more succinctly: what causes the stock market to “feel ready” to start corrections?
Thanks!
I’m sure Sam will elaborate better than I can, but technical and statistical analysis tends to work well for a couple reasons, especially for calling bottoms or pivot points (relative to using nothing at all), because the market is not that random. For one, when lots of people are using similar signals, to some degree it becomes self-fulfilling.
The other reason, which may be more important, is that behaviors of large numbers of people can be inferred by looking at indicators like volume, RSI, directional trends and lots of other signals. Experts like Sam can get a pretty good feel if the market seems to be exhausted on the sell or buy side, which inevitably leads to a reversal.
The simplest explanation is that if everyone has already bought, the stock has to go down because only sellers are left. And when everyone has already sold because they think it can’t get any worse, stocks have to go up because there’s only buyers left.
Getting the timing close more often than not is the reason why we’re subscribed to the newsletter.
I think it’s a mix of what Mr. Obviously is stating together with the repetitiveness of human nature. It’s a multitude of different factors.
(1) Retracement of the Previous Correction
This rally was always going to be a minimum 30% rally due to the size of the preceding correction. For sure that is always going to be the minimum if we weren’t entering a bear market. No bear market means, 30% rally at some point because that’s what it takes to get back to the highs after a 25% sell-off.
In 2019, we had very similar circumstances. The QQQ sustained a 23-24% correction in late 2018 and then rallied 30% during the first half of 2019 to get back to the highs. The market ultimately peaked at around 34% — just past the highs — and then sustained an 11-12% correction May 2019. That correction & rally is probably the most identical to today.
If the preceding correction were only 10-12%, then we’d only need a 12-14% rally to get back to the highs. That’s not the say the rally will only last 12-14% — it can go a lot further and usually has to in order for us to constantly push to new highs — but it is to say that the rally minimum is a smaller than the rally minimum here after a 25% rally. IT takes less to get back to highs in that situation.
So that is to say that the size of the rally generally corresponds to some degree to the size of the preceding sell-off.
Correction Retracement size forms part of the equation.
—-
(2) Built-in Profits
Then we have built-in profits. During that big rally from the lows put to highs, a lot of people have a lot of built-in profits. At some point, there will be a sell-off that triggers profit taking. If the market starts to pull back very heavily off of its highs, major investors will be incentivized to preserve those profits by selling.
(3) Time Cycle
The time cycle that we observe says a lot about the market. The fact that almost all rallies end by day 100 informs market participants, algorithm and trading programs when to be quick on the trigger.
If we’ve rallied 70-80-days, then at least the informed market participants are going to get very happy.
Imagine if the QQQ goes up 45% on the rally and we’re sitting at day 114. Any small sell-off is going to trigger selling as everyone paying attention will think “this is the correction.” In that sense, the time cycle is both intrinsic to the market and self-fulfilling.
Had Israel attacked Iran at Day 100, I think we could have seen a lot more selling as more investors would be inclined to dump at that point.
(4) Returns
As you can see from the NASDAQ-100 (QQQ) charts, the market has a tendency to peak at the 30% market at the high end. 30-36% represents the majority of the top 5-10 rallies going back to 2010. All of those top 10 rallies peaked at 30-36%. Covid is the outlier not just as a rally but really as a data point. We only include it to point out that we’ve seen a rally go for 84% afer 6-7 pretty substantial pull-back that were short-lived. But that rally stands alone in multiple ways. First, it was also the largest sell-off we’ve seen on the QQQ at 31%. That means, just to get back to the highs, it’s a 50% return rally. That’s just to return to the highs. No other correction had that set-up going in. Second, no only was there immediate resolution of the issues that sparked the sell-off to begin with, the fed engaged in considerable QE which drove markets higher. Third, it’s not obvious that it should be included given the multiple 6-7% pull-back we’ve seen on the way up.
Outside of that rare set-up, the last 15-years has followed a clear-cut repeatable pattern of producing retracements on 10-36% returns.
—-
Now in terms of what sparks it, it could be a single fund deciding to unload or a group of funds unloading together once one begins to sell. Particularly when the market has already met the parameters of a full rally.
Sometimes is minor news leading to a knee jerk reaction — again due to the market already expecting a pull-back to begin soon.
It’s not always obvious what the catalyst itself is. Sometimes we get consolidation tops ahead of a correction other times we go from rally mode right into sell-off mode like we saw with Covid.
If someone was in vacuum and had no access to news, there were multiple times where turmoil existed under the surface and the market acted as if everything was fine until the very last moment when it just decides to collapse. So in some sense, it feels like there is joint agreement to sell at a certain point — either by playing chicken or through recognition.
We saw this going into both the financial crisis and the covid crash on the QQQ. The Covid crash, the QQQ was on a slow grind making new highs each and every day and then BAM! It just started collapsing hard. 10% down in matter of a few days.
We saw someting similar in fall 2008. While the Dow and S&P were sort of grinding lower very slowly going into the crisis, the NASDAQ-100 was near its multi-year highs in August 2008. I remember Apple being at $190 a share in August ($200 was its all-time highs) and just a month and a half later it was down 60%. The QQQ highs was at $47.64 in Nov 2007 and it traded at a high of $42.09. This after a 25% correction earlier in the year. SO while its was about 10% off of its highs, it had substantially recovered from the Bear Stearns collapse earlier in the year and was trending higher going into the crisis.
The point being the QQQ — from the perspective of being in a full vacuum (no news) — looked as if it were trading toward its highs. It was rallying one moment and in the next it was collapsing.
Hi Sam, in regards to the latest update, I know you’ve discussed how overbought can actually be bullish and signal more upside (if it occurs early in a rally for example). Is the same true of oversold early in a potential correction, or is it more of a given that there’s almost always a rebound (even if it’s a minor and near-term) ahead of more potential downside? A more simple way to put it – do oversold and overbought signal similar movement in a potential correction/rally or are there differences?
So it’s supposed to be perfectly symmetrical. Meaning, this principle holds true for both rallies and sell-offs. But I’ve found that it is entirely asymmetrical.
Momentum rules suggests that a speeding car must always slow down before it stops. In the sam way, a rally must see a slowdown in upside momentum before peaking. A sell-off must see a slowdown in selling momentum before bottoming.
A technical analyst following strict momentum rules will suggest that you need negative divergence in a rally (slowing momentum) before there’s a peak. A higher price in the the market/stock with a lower peak in the RSI. So for example, if the RSI peaks at 74 with the QQQ at $520 and then after a pull-back and rebound, the next RSI peak is at 69 with the QQQ at $535, you have negative divergence. A higher price on a lower RSi.
On a sell-off, it’s the same thing in reverse. A technical analyst might argue that you need positive divergence, a lower price on higher RSI before we get a bottom.
But I’ve found that it’s very rarely true on the downside. Notice how during the March correction, we bottomed right at a 20-RSI on the dialy chart. We didn’t need to see new lows on a higher RSI. In fact, it never happened. The QQQ bottomed at $402 at peak oversold indicators. While we did get a retest, it wasn’t strict positive divergence.
I think a huge reason for that asymmetry is due to the fact that there is a finite amount of downside to the market — QQQ can only fall $534 points — while there is an infinite amount of upside. The QQQ can theoretically rise to $10,000+. That asymmetry leads to an asymmetry in the technicals.
In the overwhelming majority of the cases, oversold leads to a rebound in the market.
Hi Sam,
Could you elaborate on why we’re thinking there is a small probability of rolling over and/or continuing lower? Why are we not as worried about the correction starting at that point? Is the general idea that corrections don’t start when the hourly RSI is oversold?
Thanks!
So overall, even in a correction, the QQQ generally rebounds off of oversold conditions. Even if a correction were to have begun.
For example, suppose rather than up today, the QQQ was down $6-7 and reached deeply oversold territory. It reached a 20-RSI. Even in MOST corrections, we’d get a strong rebound at that point ahead of more sell-off. Supposing the QQQ reached a 20-RSI at $520, it might rebound to $528-$529 before then beginning another major leg lower. That’s how it general goes.
Corrections unfold with sell-offs > oversold > rebound > subsequent legs lower.
But as we observed in the Feb-April correction, that’s not always the case. There are times where the markets will continue lower in spite of reaching oversold conditions. That’ the slim probability part.
This is all moot of course as the market rebounded today. So we never got the set-up to begin with.
Right. So we’re back on the 540 lookout to buy those 500 puts.