"Just Draw a Line" — The Best Advice Ever Heard at a Master Trader Program

This Q&A session began with a line that made the whole room laugh. Someone asked what to do about GCT (GigaCloud Technology)'s outside day. Minervini didn't answer directly. Instead, he told a story: years ago, at a Master Trader Program, students kept pressing him on how to read every possible price pattern and where exactly to measure a breakout from. Finally, David Ryan threw up his hands — "Look, man, just draw a line." The room burst out laughing. But Minervini's next sentence silenced the laughter: "But it was probably the best advice that he ever gave."

For the next 81 minutes, that line kept coming back. It explained why GCT should be stepped away from, why LLY only deserved a C rating, and why "buying a little higher" is often the right call. In the same session, there was also a live trade placed on the spot, a hard truth about stop losses, a rhetorical question about AI, and a man who built his own model explaining why he regrets ever introducing it.

📅 2026-09-17 (Thursday)｜MPA Member Q&A｜Guests: Mark Minervini, Brandon Hedgepath｜Runtime: ~81 minutes
⚠️ All prices, dates, and quotes below are taken from the video captions and cross-checked against daily chart data. Individual stocks are used for methodology demonstration only and do not constitute any buy recommendation.

---

I. Opening: The Market Is Still Under Pressure, But Some Names Are Already "Moving First"

Minervini, as usual, ran through his focus list first. Today's standout feature was gaps: MRNA (Moderna) was up over 10% intraday, and MTW (Manitowoc) and TH both gapped up as well. Among the names that didn't gap, MG (Mistras) was pulling back on shrinking volume — "just squatting, pulling back a little bit here," the pattern still looked good; HTFL was building a power play setup; ADM was bouncing nicely and might form a pullback buy point in the next day or two.

Then he laid the market's cards on the table: "This market is clearly under pressure." And — "we haven't even gotten a follow-through day yet."

About the STEM model: He started with an apology. A member (Ken) asked a pointed question: since STEM launched in August 2021, the green state has only accounted for 2.2% of the time. Orange and red periods can still be profitable but are clearly harder — has the past five years been different from his longer experience?

Minervini's answer: "I'm really sorry that I ever introduced any model." His reason wasn't that the model was wrong, but that people misuse it: the model was meant to be "one more layer of color" to help you see what the market is doing; but people treat it as a magic indicator to decide "whether I can trade today." He specifically distinguished two things: STEM is a rearview mirror with subjective judgment; the SPY model is purely mechanical, based on a set of trend indicators, and even it has limitations.

So his conclusion was hard: don't treat any model, anyone's opinion — including my own — as your leading indicator. What you should really use as feedback is your own trades and your own watchlist.

That idea was pushed to its extreme later in the session. Someone asked: with the 10-year Treasury still near 5% and oil still high, can the market really bottom and rally under this macro backdrop? His answer: "If they're making decisions on their stocks based on the 10-year, it's an absolutely no, never." If they're using the 10-year Treasury to decide individual stock buys and sells, the answer is: absolutely not, never. Then came the line that could be etched on your screen: "Stocks have the final say, always."

He also shared a detour from his early years: he started as a quant, with the sequence "first find a good market → then find the best sector → then pick stocks from the best sector," and the result was four to six consecutive losing years. Later he completely reversed the order — look at individual stocks first, let the stocks lead him to the sector; if a bunch of stocks are building setups, it's probably a good market. From that moment on, he "miraculously" started calling tops and bottoms accurately, earning a reputation as a prophet on CNBC and among Wall Street institutions — while he was barely even looking at the market.

---

II. Live Trade: MG — A Demonstration of "Buying the Squat"

When asked "If I were all cash right now, which ones should I buy?" Minervini's answer was: buy the ones that are moving first today. And since he was on the subject, he just demonstrated it live: "I'm gonna practice what we preach here. Let's, man, let's add MG right here." — I'm going to practice what we preach. Let's buy MG right here.

He wasn't buying a breakout — he was buying a pullback (what he calls a squat). The reasoning was clear: "I actually prefer at this point to buy this little pullback on a squat because then I wanna see how we close." At this position, I'd rather buy this little squat pullback, because then I can still see how it closes today.

Then he did the math on the spot, without skipping a step: this order was bought at $19.92, the prior low was $19.15, and the stop was placed at $19.10 — a difference of $0.82, about 4% risk. (Caption原文: "So only 1992, we're 82 cents away… about a 4% stop.")

Why half size instead of full size? Because the follow-through day hadn't come yet. He gave a framework for progressive exposure: if you're all cash today, start with 5%–10% to test the waters; if those positions start working, push toward 25% — either one large position of 20%–25%, or four or five small positions of 5%–10% each. "If that's working, I could care less what STEM models or what anybody says." — As long as it's working, I couldn't care less what color the STEM model is or what anyone says.

Where is this trade's "line in the sand"? The answer isn't at the stop price — it's in the structure: if it breaks below the prior low of $19.15, it forms an outside day that swallows the entire pivot, and you just abandon it. "I would just punt it." — Punt it on the spot, no debate.

MG: Live demonstration on 9/17, entry at $19.92, stop at $19.10 (~4%); a break below the prior low of $19.15 is treated as an outside day swallowing the pivot — abandon immediately.

---

III. "Just Draw a Line": GCT and LLY Teach the Same Lesson

Back to the question from the opening. A member (Ryan) said: he bought GCT yesterday (9/16) on the breakout, but it closed as an outside day; he remembered Minervini saying that with MG, an outside day in that situation might mean abandoning it — what about GCT?

Minervini's answer had two layers. The first layer was the math: if you really want to trade it short-term, you can certainly sell quickly. This is indeed not what you want to see, and the breakout wasn't successful — it just poked its head out and barely crossed those price levels.

The second layer was the line. "Just draw a line from the pivot to the left and you will very clearly see where the supply is." — Draw a line from the pivot to the left, and you'll see exactly where the supply is.

On GCT's chart: that line (the orange dashed line) is drawn at $53.70 — exactly the high of the September 16 outside day. Above that line, only a few scattered days touched it, and — "it never even closed above that level." Not a single day ever closed above that price. So that was its line of most resistance.

GCT: 9/16 outside day, poked above $53.70 then closed back below the 9/15 low; never a single close above that line — the "line of most resistance."

The next passage is the essence of the whole session. When does that line stop working? Minervini said: when you move the pivot up to the newly formed platform, the stock walks up the right side and then sideways for a while, you'll find — above the new pivot there are only very few days of trading, and the supply has been cleaned out. Then: "The line of most resistance turns into the line of least resistance."

He explained the process with an analogy: the established uptrend is the larger force; this short-term consolidation is the smaller force. The smaller force is the seller, the larger force is the buyer, and what we're waiting for is the larger force consuming the smaller force. Once the sellers are gone and there's no supply overhead, it only takes a little buying to push the stock up — and we know the volume of buying is large.

So he corrected the causality behind observations like "tip-off day": "It's not the cause, it's the effect." — That volume spike isn't the cause, it's the effect.

The same lesson, applied to LLY, gives the opposite conclusion. Another member (Fabio) asked: LLY (Eli Lilly) looks like it's about to break its downtrend and start forming a right side — should we look for a "little cheat" here? Amazon seems similar.

Minervini said: if you really want to bottom-fish LLY, it first needs to tighten for a few more days and form a proper pivot — and that's probably "the earliest point at which I would ever consider bottom-fishing a stock." But when he drew the line, the conclusion changed: looking left from the pivot, there's a dense wall of volume above — meaning a dense wall of supply overhead.

LLY: Drawing a line left from the pivot shows a dense band of volume above (= supply to be absorbed); the 50-day is still declining and the price is below it, so the rating is C.

Plus two hard criteria he wouldn't compromise on: LLY's 50-day is still declining, and the price is still below the 50-day — "That hasn't even gotten the intermediate-term trend to turn up yet." So its rating is C, and the strategy is: wait for improvement; if it can clear those highs, you can start taking a small bite.

And here lies the most counterintuitive line. He said the hardest thing for many people to understand is: buying where supply has been cleaned out, even at a higher price, is most of the time the better choice. "In hindsight you can go back and say, yeah, of course if I buy it up here, obviously I would have been better off if I bought it lower. But that's only in hindsight." Because only in hindsight do you know — for that "lower price," how much volatility you'd have to endure in between, whether you'd have been stopped out long ago. That lower price comes at a cost. Buy into supply, and you must be extra vigilant about its performance; buy after supply has been cleared, and you're buying certainty.

Of course, he also gave the entry threshold for "buying a little higher": if the stock has only formed a compressed right side right next to a decline, too close to the downtrend, that won't work either — the tightness of the right shoulder is always relative.

---

IV. "I Already Set My Stop — Why Am I Still Panicking?"

There's a second half to the GCT answer, and this part might be the most important thing to write down from the entire session.

So should you stay or go? He didn't give an answer — because you already gave that answer when you entered. "The answer is you set a stop loss. Did it hit the stop? If it did not, then you should allow it the room to fluctuate because you already made that decision." — You set a stop. Did it hit the stop? If not, you should give it room to fluctuate, because you already made that decision.

Then came the question that makes everyone's heart tighten: why, when you've already set a stop, do you still feel enormous pressure the moment the stock moves against you? His explanation: it means either your stop is too wide, or your position is too big — you think you can handle it emotionally, but you can't. "The pressure doesn't usually come from the stock price. It comes from position sizing." — The pressure usually doesn't come from the stock price. It comes from position sizing.

The solution follows naturally: either tighten the stop, or reduce the size of each position, then gradually "grow into" that size.

He also gave a very practical self-check tool — on the platform's Monalert, GCT clearly showed three lines that day: a reversal bar, an outside day, and a close below the 10-day line.

So what's the difference between MG's outside day and GCT's outside day? He put them side by side himself: GCT was "an outside day, that's it"; MG's was not just swallowing the previous day, but swallowing the entire pivot, the entire prior low — "that's my line in the sand. So I'm out." That's my pre-drawn line in the sand, so I'm out.

So the outside day itself is not the signal — what it swallows is the signal. Swallowing one day's range, you can tolerate; swallowing the entire pivot, that's your pre-written admission of error being triggered.

---

V. How to Actually Use the VCP Rating: Only Look at the Top 200–300

A member asked: in the VCP rating, which factors separate MRNA's 90 from others?

Minervini first poured cold water: it's a computer model, don't obsess over whether it's 77, 90, or 43. His usage is a bell curve: treat 70+ as the top 30% bucket; and more importantly — always only look at the top 200 to 300 ranked names, regardless of what percentage the score threshold cuts at. He did the math: even if the threshold drops to 5%+ and the list expands to 400 names, in a pool of 10,000 stocks, that's still the top 4% — already a very elite group.

The scores reported live this session are interesting as a firsthand reference: MRNA 90.6 (yesterday), Twilio 89, WTTR 77, HTFL 77, and MG — the live trade — only 59.5. Notice: he doesn't only buy high-scoring stocks. The rating is for quickly narrowing down "good patterns"; a 59.5 MG can still be bought because the position is right and the risk is small.

Following MRNA, there was a very practical exchange. A member (AK) said: he bottom-fished MRNA on 9/10 with a stop at $128; today was its first real breakout day, but it already had multiple R's of unrealized gain, and it might be one of the first names to break out in this pullback — should he trim to reduce risk, or hold for a big move? He admitted his old habit was "stops too tight, profits taken too early."

Minervini's assessment: "You nailed it as it's turning up off the low." You bought it very precisely as it was turning up from the low. As for whether to trim, he gave a very specific ruler: if you originally wanted a big move, your stop was only 2%–3%, and it's already up 10%, then I'd probably kick off 10%, 15%, or 20%, keep 80%–90%, and completely remove the risk from the entire trade — "completely de-risk it and free roll it," turning it into a "free carry" position.

MRNA: Member entered on 9/10 as it turned up from the low (stop $128); on 9/17 it gapped up, up as much as 11% intraday, VCP score 90.6 — the highest of the day — but they chose not to chase the gap.

---

VI. SPCX: Full Size, Sold Half, Added Back to Full — Then He Says "I'm Not Sure You Should Add Here"

A member (Dennis) asked Minervini to review his SPCX (SpaceX) trades: built full size on August 25, sold half on September 3, and added back to full size yesterday (9/16).

Minervini's first sentence was a cooldown: "I'm not sure I'd be pressing it here yet. I'd like to see that right side maybe tighten up a bit more." — I'm not sure I'd press it here yet. I'd like to see that right side tighten up a bit more. Then the hard metric: "The 50 days still in a downtrend." The 50-day is still declining.

But notice what he claimed first — because the idea of "buying a pullback near the lows" was his to begin with: "I'm the one who talked about buying it potentially right here off the lows on a pullback buy, and from a risk reward standpoint I thought it was worthwhile." From a risk-reward standpoint, I thought that spot was worthwhile — because just a small move up would be enough to cover the risk.

So for the member following through on that, his attitude was approval. What he didn't approve of was continuing to add at this position. And he described the nature of this trade more clearly than anyone — he imitated the voice in a retail trader's head: "Oh my God, this is SpaceX, this is, you know, Elon Musk, this could be a huge winner." His handling wasn't "sell everything" or "hold blindly," but: keep some, de-risk it, free roll it.

Most valuable was the analogy he gave — in the weekend video, he compared SPCX to two historical stocks: one like Meta, which rolled over after completing its right side; another like LinkedIn, which formed a steep V-shaped right side and then also rolled over. "Both of them rolled over." — Both eventually rolled over. "I think this one will probably be another one that rolls over. It's a matter of where it's from." I think this one will probably be another one that rolls over. It's just a matter of where it rolls from.

So the position discipline he gave is a declining curve: "The more it goes up the right side, the less I would be adding to it." — The more it goes up the right side, the less I would add — unless it's breaking out from an adjacent, newly formed base. And this kind of straight-up-the-right-side move will most likely run into that wall of supply on the left.

SPCX: 8/25 full size → 9/3 sold half → 9/16 added back to full; 50-day still declining, overhead supply not absorbed, Minervini explicitly does not recommend adding here.

---

VII. "Buying Half Size" — What Does That Actually Mean? It Has Nothing to Do with Position Sizing

This is a particularly easy detail to misunderstand. A new member (Mario) asked: on the platform, does "buying half size" mean actually putting on 12.5% or 25%?

Minervini's answer was crisp: "No, no. It has nothing to do with position sizing." It expresses the incremental nature of the action: "It's more of just communicating the incremental action." It's just communicating that this is a gradual action — usually when low-cheating, or when you're coming back from cash and testing the waters.

"Half size" is always half of your own full size: if your single-stock limit is 10%, half size is 5%; if it's 25%, half size is 12.5%. And he added: even if you've set 25%, in the current market you might only put on 1%–2%, or 5%, just to dip a toe.

The final line is the key to distinguishing "following calls" from "learning the method": "We're not providing a model portfolio. We're providing buys and sells as examples to show the process. That's your job." — We're not providing a model portfolio. We're providing buys and sells as examples to show the process. Position sizing is your job.

---

VIII. Someone Asked: Can You Teach AI to Analyze Stocks Like You?

This was the last and most interesting question of the session. Minervini's answer was four words: "I have not seen that." I haven't seen it.

But his reasoning is worth reading in full, because it's really about why the market will always have excess returns: the reason you can beat the index is pricing inefficiency. If you press a button and the computer figures everything out, that means everyone gets the same answer — the market becomes perfectly efficient. What happens then? Inefficiency moves somewhere else. And when the method stops working, people abandon it; once abandoned, it starts working again — "So now you abandon breakouts. Breakouts are too random now… Well, guess what happens? Breakouts start working." So you abandon breakouts because they're too random now; guess what happens? Breakouts start working again.

So his judgment: AI models will certainly have periods of good performance, and so will I; but there is no model or system that works forever — "That would defy the law of efficient and inefficient market pricing." And the existence of that law is precisely a good thing.

The most ironic detail: he doesn't believe in AI stock picking, but he uses AI to the extreme. He revealed the platform is running eight different GPT models, in his words "Pentagon-level" technology; they have neural networks for pattern recognition, a database of tens of thousands of trades containing his own personal trades and their own focus lists. "Our models are being taught by… a 43 year veteran trader telling it and showing it what to do." — What's teaching it is a 43-year veteran, trade by trade, telling it and showing it what to do. That's what most people don't have — not compute, but experience.

He also threw in a jab at his own spending: "I've never been so unprofitable until AI came out." Before AI came out, I was never this unprofitable. (Because all the money went to hiring people to write code.)

As for "AI is going to take over the world" fears, he closed with a family story: his wife is a photographer and uses AI for post-processing. One day she called him over to look — "AI cut her arm off and put it coming out of her stomach." AI cut off one of her arms and attached it to her stomach. So he concluded: "AI doesn't know where your arm should be, but somehow it's gonna take over the world and kill us all." Something that doesn't even know where your arm should be is somehow going to take over the world and kill us all.

---

IX. Other Q&A Quick Hits

① The red line for max drawdown is "double digits." Someone asked how much account drawdown one can tolerate, and whether to set a weekly limit. Minervini said he never exits trading just because of drawdown, but once he approaches double-digit drawdown, he must stop and reassess: "I really start trading a lot smaller and really, really trying to come back slowly." Trade much smaller, climb back slowly. His phrase was using progressive exposure to avoid "death by a thousand cuts" — "If you're three, four trades, five trades aren't working, there's no reason that you're trading the same size or trading larger." Brandon's addition was more colloquial: "Once you start going into a hole, you kind of want to stop digging. At least, get a smaller shovel." Once you're in a hole, at least get a smaller shovel.

② Don't chase after a gap. Both DAR and TH were asked about. TH was put on the list last week, gapped up this morning, but the gap wasn't large; Minervini said the plan for the day was "if it gaps up, buy nothing" — "We're very reluctant to chase that because very often you get that and then everything reverses and you just get hosed." We're very reluctant to chase, because often you chase in, then it reverses, and you get hosed. As for DAR: the base is big enough, the right side is developing, but it's a bit sloppy; after a 15% pullback it tightened to 6%. His approach: start with 1/4 size, add on a breakout.

③ Don't buy into supply. Someone (Abdurrahman) bought MANE that day. Minervini said from a bottom-fishing perspective, he understood what the person was doing, but "I generally wouldn't buy right into the supply like this, right off the lows." I generally wouldn't buy right into supply like this, right off the lows. He repeated the criterion: first push the price up to hit the supply, then it goes sideways there; if shares are being accumulated, the price will tighten and volume will shrink — that's what tells you the supply on the left is no longer flowing into the market.

④ Widening volatility is the "megaphone effect." Someone asked whether VGNT around 45 is a good buy point. He said the stock was tightening, but then became more and more loose: "So now this is what I refer to as the megaphone effect, where instead of tightening, your volatility is widening." This is what I call the megaphone effect: instead of tightening, your volatility is widening.

⑤ Channel lines are not a sell reason. Someone asked whether to trim when a stock hits the upper rail of an ascending channel. His answer: he wouldn't sell just because it touched the upper rail. But if it accelerates through the upper rail, then look at whether it's a climax run — 30%–60% gain in four to six weeks, with huge volume, possibly a gap, and all these conditions must be present simultaneously to count.

⑥ Record the trades you didn't take. Someone asked whether to log missed opportunities in the Greater tool. He said not only log them, but create a separate file: right-click and mark a "missed opportunity long." Because looking back at these records, you'll find you repeatedly miss the same spot, and the common thread is often "you're holding a strong personal opinion about the market, and that opinion is blocking you."

⑦ No single stock should become your obsession. Asked about ANET and NET, his wording was direct: ANET "isn't close enough" and "isn't on my list" because it's too loose, not tight enough; as for NET, he already has a small position and trimmed part of it at one point; the stock is indeed strong and heading to new highs, but chasing now doesn't pay — "It doesn't have that right side development." It doesn't have that right side structure, and if it naturally pulls back to the 50-day, that would be an 11.5%–12% drop, "that's too much for me, I'm not willing to sit in a 12% loss."

---

X. Quotes of the Session

• "Look, man, just draw a line."

• "The line of most resistance turns into the line of least resistance."

• "The pressure doesn't usually come from the stock price. It comes from position sizing."

• "Did it hit the stop? If it did not, then you should allow it the room to fluctuate."

• "Stocks have the final say, always."

• "I'm really sorry that I ever introduced any model."

• "The more it goes up the right side, the less I would be adding to it."

• "We're not providing a model portfolio… That's your job."

• "There is no model that works in perpetuity."

---

A Note to Myself

The most valuable thing about this session is that behind every conclusion is a number you can calculate. Buying MG, first calculate the $0.82, 4% risk; drawing a line, first count "only two days above it"; discussing trimming, first say clearly "kick off 10%–20%, keep 80%–90%"; discussing position size, first admit "a 12% potential drawdown I can't sit through, so I'm passing on this one."

What "just draw a line" really draws isn't a pattern — it's where you're willing to admit you're wrong, and how much you're willing to pay for it. Draw the line right, and panic has a boundary — because before you entered, you already wrote down "I'm wrong."

---

⚠️ Disclaimer: This report is for reference only and does not constitute investment advice. Markets carry risk; invest with caution.