Learning Through Market Drops

October 22, 2025

Subscribe now

This is not financial advice. I’m just a guy walking his dogs, learning to trade, and thinking out loud.

It’s walks and stock talks with no tiktoks—although I will dance for you if you really want me to.

But nobody wants that.

THIS IS MY FIRST PODCAST ON HERE. And this article is the transcribed and edited version.

Beautiful fall day out here with my dog. It’s been a crazy week or two for stocks. The major indexes have been down, gold had a rare drop of about 11%. And you know, it’s my first year trading still. So it’s been interesting.

The First Big Drop

It’s really the first time I experienced a big drop and held. I didn’t sell, which is good ‘cause you don’t wanna realize your losses. Make them real by selling. You wanna buy high or buy wherever, but sell higher if you’re gonna sell.

Learning when to take profits is one of, it’s on my list. And not being lazy and actually putting in stop losses. Also, switching from Vanguard to Fidelity or probably Interactive Brokers. I need trailing stop losses, percentage based, not—you know, Vanguard only does specific price levels, which is really stupid, but their platform’s not meant for active trading.

So it’s just one of those things where you think it’s gonna be a big deal to switch. It’s probably not.

I just verified through Perplexity and it’s NOT. Plus IBKR has more insurance per account than Vanguard does- and one of the reasons I want to get on there is to trade international stocks that you can’t on some brokers. Fidelity and Schwab have more than Vanguard, but not all.

Even if there are delays, it’s probably good for me to be outta the market for a day or two anyway. I’m learning not to overtrade I’m kind of an overthinker, always adding onto a system to make it better and then confusing myself, breaking it- altho having started from scratch several times has helped me verify my ideas, learn the calculations better, and working with Perplexity has helped me find even better ways to calculate things.

For example, I’d been checking linearity of price movements with CORREL(), but P suggested this instead: R-squared linear regression…

=RSQ(B2:B252, ROW(B2:B252))

Why?

RSQ (R-squared) is better than CORREL for trend/momentum scoring because it quantifies how much of the movement in price is explained by a linear trend over time—it gives the proportion of price variance explained by the trend, not just the direction or strength of association.

RSQ tells you how well a linear model fits the series (0→10→1), so a high RSQ means prices are moving in a straight, predictable line—ideal for identifying strong, steady trends.

CORREL gives you the correlation coefficient (−1→1−1→1); it measures strength and direction of linear relationship, but is not sensitive to whether the strength is due to steady trending (explained variance) or volatile movement.

Downside of CORREL:

CORREL can show a high value for noisy series that just happen to move in the same direction, even if the underlying move is choppy or mean-reverting—it doesn’t penalize “wild” but coincident movement, and the sign can be positive or negative, which isn’t directly useful for ranking straightness or “pure trend”.

RSQ removes directionality and reports only the “straightness” of fit, so you can rank trends by their predictive clarity, not just association.

In short:

Use RSQ when you want to know how much of price movement is truly trend-like (regression fit).

Correlation just tells you if price and time move roughly together, not how clean or strong that movement is.

The good news is gold’s going back up at least a little bit. I actually bought a little bit more on the dip, first time.

I come in as kind of an iconoclast. Like I always wrongly think I’m smarter than everybody else.

And then eventually make my mistakes and then learn the hard way and then see the value of some of the advice I thought was stupid in the beginning, like buy and hold, don’t sell low, blah, blah, blah.

Building My System (With Google Sheets and 50 Perplexity Searches a Day)

Playing around though with, I use a lot of Google Sheets, you know, until I get kind of my formula and my system that works for every situation. At that point I’m gonna go algorithmic, well, or automated trading or something like that. We’ll see. But I’m doing it manually for now.

And I’ve been adding in some things like, you know, if you play momentum, you wanna get into a stock that’s going up, but when it goes too high, it’s gonna come back down. And one of the things I’m grateful that I picked up from various people on Substack elsewhere, Perplexity searches—I search Perplexity about 25 times a day or 50—is the whole difference between institutional and retail traders.

Institutional vs. Retail: The Unstoppable Momentum

Institutional, that means all those large entities that invest massive sums of money in the market, such as pension funds, mutual funds, hedge funds, insurance companies, endowment funds, commercial banks, and sovereign wealth funds. These groups are the “smart money” and control the majority of trading activity—over 80% of S&P 500 capitalization is held by institutions.

But anyway, they buy so much and a lot of it’s often… It’s in retirements and pensions and things like that, so it’s stable money.

They can’t change it often. They can’t change it fast. They don’t wanna move a market or freak a market out, so they tend to accumulate and take profits slowly. Which does create sort of unstoppable momentum, up or down- unstoppable by retail traders (and retail’s you and me unless you work for big fund or something like that).

What we gotta watch out for is when retail traders are looking for stock that’s going up. It might be going up due to institutional accumulation. And then retail gets excited. The news gets excited, people buy more. It kind of overextends gets up into rarefied air where there is no support.

The support below it where institutions bought a lot, it gets further and further away as they drift towards space. It’s almost like a rocket high in the atmosphere. It’s gonna run outta gas and plummet back down.

So there are some formulas for figuring out how far up it should go. I ran across one the other day and didn’t really fully digest it. But I also look at things like what is the recent growth or decline in return and in risk adjusted return, which is of course return divided by volatility such as standard deviations. Those are all good.

And here are some tips from Perplexity on buy/sell/hold criteria (these were a bit slanted toward gold/IAU):

The Decision Framework: Hold vs. Sell

Based on the technical research, here’s your systematic approach:

HOLD Signals: Price testing but holding above 200-day SMA - long-term trend intact VMA declining after price decline - selling pressure exhausting Bullish RSI divergence forming in oversold territory with support level holds Volume declining during pullback - not institutional distribution Consolidation timeframe aligns with historical patterns (3-5 months for gold) Successful retest of broken support-turned-resistance levels Higher lows forming despite short-term weakness - constructive price action SELL Signals: Break below 200-day SMA on heavy volume without quick recovery 8% decline from your entry point - hard stop loss rule VMA rising during price decline - increasing panic/institutional selling Bearish RSI divergence in overbought conditions confirmed by breakdown Failed retest - price breaks back through supposed support Consecutive lower highs and lower lows with no constructive bounce attempts Consolidation extending beyond historical norms (beyond 5-6 months for gold) without resolution REDUCE POSITION/WAIT Signals: Trading in middle of range between key support and resistance - no clear setup Mixed signals from multiple indicators - wait for confirmation Low volume consolidation during summer months (seasonal pattern for gold) Approaching but not yet breaking critical support levels - wait for clarity

Technical Indicators: The Lagging Problem

I did a lot of stuff back in March through May with TradingView’s technical indicators, and people have made a lot of their own, and I programmed some with the help of AI. I was deep into that and kind of came out of it going: a lot of these are just lagging indicators that don’t really project future momentum in any way.

That’s why I went to the Google Sheets and created some stuff that’s about long-term trends. You know, it’s tough, for example, if something’s going up a lot right now, but for the last year it’s been going down. It’s hard to trust that kind of change unless there was an actual fundamental change in the company. It could just be news hype. It could be people kind of rotating into what seems good right now. They’ll rotate out of it.

So looking for the long term stability signals, things like earnings per share that are positive. Of course, that doesn’t work for ETFs. They don’t have EPS. So I screen for that, and I gate them by saying they have to be >-0.01, so the ETFs still qualify.

I need to get a better data source so that I can add fundamentals in. ‘Cause I’m using Google Finance and it doesn’t have very many fundamentals. Looking at volume is another thing that I do. Volume can confirm whether something’s the price movement is reliable.

About the Dogs (A Brief Interlude)

“Come on sis. Let’s go back.” There’s some other random dog walking around without a leash. Yeah, our dog was named Sister when we got her. It’s very weird. But we have rescue dogs. We got rescue dogs and we keep their names.

So we’ve got Sister who we call Sis. We got Coco, that’s not a bad name, but she’s honey colored, not chocolate. She’s a chocolate lab times a Great Pyrenees. Somehow she came out looking like a yellow lab, but her name’s Coco. And then we got Toots who looks like a Tootsie Roll maybe, I guess. I don’t know.

But it’s funny that she’s called Toots ‘cause that seems like a kind of a jovial name. And she is very, very anxiety based. So I think it’s funny. We’re working on her. She’s getting a lot better.

De-correlation: The Johnson & Johnson Example

Anyhow, de-correlation is what I wanna talk about with Johnson and Johnson because one of the things I’ve studied is like, you can correlate the price movements of say, the NASDAQ to gold, right? Find out, oh, maybe gold’s about 65% correlated, right?

If it was 0%, the prices wouldn’t be the same at all. If it was negative 100%, it would mean the prices were exactly the opposite of the NASDAQ. That’s not really a good hedge because your prices cancel each other out, so I tend to look for 30 to 60% correlation.

The Solver: Puzzle Piecing Your Portfolio

And then there’s a way if you find really good stocks long term and have their correlations that you can run what’s called a solver. Come here, come on. Solver is this mathematically complex thing that you can do through—you could do linear ones with Google Sheets. Had to go to Excel to get the non-linear GRG, so that I can solve according to maxes or mins or averages.

‘Cause what I wanted to do is take all the prices of all the stocks I’m looking at, find the proper allocation that kind of puzzle pieces ‘em together in a way that maximizes say, average return per day. But I also wanna minimize volatility.

So I actually do a formula that’s like average daily return divided by the standard deviation squared or cubed ‘cause I wanna emphasize that, so that way we’ve got a risk adjusted return sort of metric and we’re trying to match all the pieces of the portfolio to really reduce any daily decline in the portfolio while maximizing the overall average return. So that’s a good one.

Market Polarities: Thinking in 3D

But in doing all those sheets, I looked at de-correlation and there are a few stocks like MO, which is Altria, and Johnson & Johnson—they’re pretty de-correlated from all of what I would call the poles. They’re sort of polarity. I kind of visualize like a 3D market where gold’s going in one direction in this three-dimensional cube. NASDAQ’s going another.

I mean, the S&P 500 and Dow Jones are so similar to the NASDAQ and the NASDAQ’s really the most extreme form of the three. So I usually don’t look at the other two in terms of polarities. But you can also look at crypto polarity with something like WGMI, you can look at oil or energy.

And then on the flip side, for risk off, because in a truly risk-off environment, like we’ve had the last few days where gold and tech and everything seems to be going down, you know—it’s interesting, gold is typically a good hedge because it’s a function of despair, or fear, but it’s also now become a momentum play for profits. And so its prices have become based on hope and similar to NASDAQ. Tech is all about hope, in my opinion.

Risk-Off Plays and the Healthcare Anomaly

So we wanna de-correlate from those things, but we wanna find some risk off ones. Like typically when people are afraid they buy energy, healthcare, consumer staples, you know. So I’ll have a polarity for XLP, XLU, XLV, which is health.

Health is weird. It’s been a little squirrely these days. It’s been squirrely this year, but it’s interesting. So if people have hope, you see lately, risk off, kind of hope is biotech ‘cause it’s a combination of the hope of tech and the safety of healthcare, right? So, yeah. XBI, it’s not been a reliable thing all year long.

A lot of those, those are the 12 SPDR sector stocks, right? Or there’s 11 of them I think. I don’t know if you had real estate. I think it’s 12. But any of those XLs, that’s what those are.

Why Johnson & Johnson Makes Sense Right Now

So anyway, you have all these different polarities and then you measure each stock, how correlated is its price movement to those poles. And then you can kind of say, well, it looks like the market is kind of 1.3 to one ratio NDX to IAU, which is gold.

Um, wow. Yeah. I might have lost some of you guys on the acronyms, but I think this will be coming out with a written piece. So, de-correlation is really important. Like if things are all over the place and people don’t know which way to go, and they’re rotating quickly from one to another, that’s when you want a stock that’s going up without being too correlated to all those poles.

And that’s where, you know, you can kind of run like a calculation on the correlations to those poles and see like if they have a low standard deviation, that means that it’s pretty similarly correlated to a lot of things, which in my opinion is kinda like de-correlated. That’s why I think about it that way. That means it makes sense that Johnson & Johnson’s going up, you know. MO not so good right now. Don’t fully understand why.

But the other way to look at that is that Johnson & Johnson has so much diversity of products that they’re not relying on tech. They’re not relying on people only buying risk off. People will buy shampoo, whether the economy’s good or bad, right? So that’s your consumer staples play. Et cetera. But they’ve got more than that, right? ‘Cause they’re still going up when it’s risk on, when tech stocks go up, J&J is still going up.

Those are just a few things I wanted to review. ‘Cause I’ve been trying to learn how to always know what stocks to be in, regardless of the environment. That’s my goal.

Learning From Hiccups: Don’t Play Favorites

So all these little hiccups like corrections, declines, they’re all a great way for me to learn. And I’ll tell you one thing that you learned from that is do not play favorites.

Like, I have some stocks that I loved for a while. It’s like, I don’t know. I mean, it’s the only good analogy and I’m married, but the only good analogy I think is like dating or stuff. Like you can have a, you can have, this woman was so great with her, you know? But right now, she’s in treatment. She’s crazy. I hope she comes back though. It’s really good. Got a really good time.

That’s a horrible analogy. I’m sorry.

But there are stocks like, man, I was so into Howmet Aviation earlier this year. They were great for a long time and then they weren’t, and then they were, and I missed it, you know? ‘Cause I don’t have good monitoring in place. You know, it’s, there’s always things to work out with my system.

Unfortunately, you know, the Google sheets kind of start crashing when you have more than 200 stocks in my screeners, which use, I don’t know, some 200 columns. Let me see where we at right now in this one. 225 columns. Yeah. And then all those Google Finance calls don’t work. Sometimes when everybody’s looking at the same stocks, either ones that are good or bad, then you’ll lose your data.

So I’m gonna get, I think it’s called FMP. It’s not that expensive. Get a different data partner that also has the fundamentals in it. Start using Excel, get used to Excel again.

My Advice (Not Financial Advice)

Anyway, my advice to you, and I’m not a financial advisor at all, and I’m not gonna give you any stock picks or anything like that, is to figure out how to have in your portfolio the things that really truly do hedge you over time.

It’s stupid because like if you’re attached to nothing going down, you can’t be like that. Some part of your portfolio has to be down all the time if you’re hedged, diversified in the right way. So, but you know, I think finding ways to make sure that your hedge still makes sense, you know, because traditionally gold’s a hedge against tech, but they’ve become so correlated. It’s not as good of a hedge. Right?

And let’s say crypto, like the ETFs, like WGMI are very NASDAQ correlated. So they’re up gigantically when tech is up. But when tech goes down a little, they drop gigantically too. So that’s not a hedge. It’s actually adding more risk on tech risk, but more gains. Just things to be aware of.

What is the true hedge? What is truly uncorrelated or de-correlated? I gotta figure out the right term for that. These are all important.

And so, you know, if I have any advice, it’s

Stay invested. Stay smart. And keep trading.

Alright, talk soon. Bye.

(Why am I hanging up like it’s a phone call, it’s a podcast. I don’t know. Hey.)

Crazy Good: Joy, Productivity, and Profit is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Thanks for reading Crazy Good: Joy, Productivity, and Profit! This podcast is public so feel free to share it.