Mark Minervini's SEPA and the 8-Point Trend Template

Most trading methods you can describe in a paragraph. Minervini's you can describe in a checklist, which is a rarer thing and the reason his name keeps coming up in momentum circles thirty years on. Eight numbered conditions, all of them pass or fail, sitting in front of a wider framework he calls SEPA. This is a plain write-up of what the rules actually say, what sits behind them, and one section near the end where I spell out which of the eight a chart screenshot can genuinely check and which one it flat out cannot. No affiliation, no affiliate links.

Quick Answer: The Method in One Paragraph

Who Is Mark Minervini?

A US equity trader who has been publishing the same core methodology since the 1990s and is described on his own Master Trader Program site as a two-time U.S. Investing Champion. He won the U.S. Investing Championship in 1997 and again in 2021, the second time in the $1 million and above stock division. Championship return figures circulate widely and not all of them agree with each other, so treat the specific percentages you see quoted with more caution than the fact of the wins themselves.

He came to this without a finance background, which is part of why the material reads the way it does. Everything is reduced to conditions a person can check rather than judgement a person has to develop. That shows up most clearly in the four books published under his name, the two best known being Trade Like a Stock Market Wizard and Think & Trade Like a Champion. He is still active: he runs a workshop, a mentorship tier called Minervini Private Access, and a charting and screening platform, Minervini Markets 360, alongside a steady stream of posts and videos.

Worth naming the thing that makes him different from most people in this category before we go further. The method is written down, numbered, and old enough to have been tested against several full market cycles by people who are not him. You can disagree with it on the merits, which is more than you can say for a framework that only exists as commentary over a live chart.

What Is the 8-Point Trend Template?

Eight conditions, checked in order, all of which have to pass. Seven of them are computed from price and two dates. The eighth is not, and that distinction turns out to matter enormously once you try to automate any of this, so the table calls it out directly rather than burying it.

Rules 1 through 5 are one idea expressed five ways: the three averages have to be stacked in order, fastest on top, with price above all of them and the slowest one pointed up. That is a longer-horizon version of the same stacking logic behind the golden cross and death cross and the same principle that makes a rising short-period average useful on an intraday chart, covered in the EMA day trading write-up. Minervini just insists on the whole stack rather than a single crossover, which is a much harder condition to satisfy and rejects a great deal more.

Rules 6 and 7 are the interesting pair, because together they define a narrow band. At least 25% off the low means the stock has already proven something. Within 25% of the high means you are not buying a broken name that has bounced. The commonly quoted figure for rule 6 runs anywhere from 25% to 30% depending on which edition or summary you read, and the exact number is less important than the shape of the constraint: he wants stocks that are strong but not yet extended past the point where a sensible stop can go. If you have ever wondered why a screener returns either garbage or nothing, it is usually because it is missing one half of this pair.

Rule 8 is a different animal. Relative Strength rating in this sense is a percentile ranking of a stock's price performance against every other stock, in the style Investor's Business Daily popularised, and it is a computed market-wide number rather than something you can plot. It is also, arguably, the criterion doing the most work, because it is the only one that compares the name to anything outside itself.

Stage 2, and Why a Filter Is Not a Signal

The eight rules are not arbitrary. They are a detector for one specific phase of Stan Weinstein's four-stage market cycle: Stage 1 is the base, Stage 2 is the advance, Stage 3 is the top, Stage 4 is the decline. Minervini buys in Stage 2 and nowhere else. Every criterion in the template is there to make Stage 2 falsifiable rather than a matter of opinion.

The distinction people get wrong is that the template is a rejection filter, not an entry trigger. Passing all eight tells you a stock is eligible. It does not tell you to buy today, at this price, with this stop. That is a separate decision, and in Minervini's framework the trigger is a pivot point coming out of a tight base, most often the volatility contraction pattern associated with his name. That page covers the entry mechanics on their own terms and there is no point repeating them here. The relationship is simple enough to state in one line: the Trend Template decides what goes on the list, the pivot decides when you press the button.

The practical consequence is a much shorter watchlist than most traders are used to. In a bad tape, very little passes, and the filter refusing to return anything is the filter working. Sitting out is a position. That is the same discipline argument behind the general momentum trading playbook, where relative strength and volume expansion do the selecting rather than a feeling about which sector is due.

Where SEPA Goes Past the Chart

SEPA, Specific Entry Point Analysis, is the full framework. The Trend Template is one leg of it. If you have read a summary that treats the two as the same thing, that summary skipped the parts that are not on a chart.

• Trend and stage
  The Trend Template and Stage 2 confirmation. The only leg that is fully visible in a price panel, and the one most people implement because it is the easiest to check.
• Fundamentals
  Accelerating earnings and sales growth, with an emphasis on the rate of change rather than the absolute number. A company going from 12% to 40% earnings growth interests him more than one sitting flat at 30%. None of this is derivable from a chart image.
• A catalyst
  Something specific that explains why the stock is moving now: a product, a contract, a new drug, a change in the numbers. Without one you are buying momentum with no story behind it, which tends not to survive the first hard week.
• A precise entry
  The pivot point out of a tight base, not a price you decided was cheap. The entry is a level the chart produced, which is what makes the stop meaningful.
• A defined exit
  Both the stop and the profit plan decided before entry. Predefining the exit is what converts the previous four legs into a trade rather than an opinion.

The gap between legs one and four and the rest is the whole reason this section exists. Plenty of people run the Trend Template, get a list, and buy off it. That is a technical momentum screen and it can work fine on its own terms, but it is not SEPA, and it is worth knowing which one you are actually doing.

The Stop Rule, the Sizing Rule, and Never Averaging Down

Three rules, and they are the least glamorous part of the method by a distance. The maximum loss on a position is capped in the region of 7% to 8% below the entry. Risk per trade is a small single-digit percentage of account equity, commonly cited at around 1% to 2%. And averaging down into a loser is not a technique he treats as advanced, it is a thing he treats as forbidden.

The 7% to 8% cap is worth thinking about carefully, because a percentage cap and a structural stop answer two different questions. A stop under the low of the last consolidation asks where the setup would be proven wrong. A percentage cap asks how much you are prepared to lose whatever the chart says. Both are legitimate, they frequently disagree, and the useful habit is to take whichever is tighter and skip the trade if the structural level is so far away that a sensible size becomes unworkable. That trade-off is the same one covered in the piece on where a stop actually belongs on a chart, and the sizing arithmetic that falls out of it is in sizing a position off risk per trade.

The pyramiding rule is the mirror image and it gets less attention than it deserves. He adds to positions that are already working, in decreasing increments, rather than starting at full size. That means the largest exposure ends up in the trades that have already proven something, and the smallest in the ones that die early. Combined with never averaging down, the effect is that your average position size correlates with how right you turned out to be, which is a structural edge that has nothing to do with picking better stocks. Sobering context regardless of the method: FINRA is direct about the risks of active trading, and no rule set changes the base rate.

How He Differs From Aziz, Cameron, TJR and Qullamaggie

These names get compared constantly, usually on the axis of who is more legitimate, which is the least useful comparison available. What actually separates them is what they teach, over what holding period, and how the knowledge is sold.

Against Andrew Aziz

Same instinct, opposite end of the clock. Aziz's method, written up in the profile of the ABCD pattern and Stocks in Play, selects a handful of pre-market gappers and resolves the trade inside the session, usually within the first two hours. Minervini selects off months of daily data and holds for weeks. Both put a hard filter in front of the pattern, and both would tell you the filter matters more than the shape. The difference is what the filter measures: today's unusual activity in one case, ten months of accumulated trend structure in the other.

Against Ross Cameron

The sharpest contrast in delivery. Cameron's program, covered in the Warrior Trading review, is built around live narration in a paid room, where the value is watching decisions happen in real time on low-float small caps. Minervini's value proposition is the opposite: a fixed, numbered rule set you can apply alone at the weekend with no room, no alerts, and nobody talking. If you learn by watching, the checklist will feel cold. If you learn by testing rules, the room will feel like noise.

Against TJR

Almost no shared vocabulary. TJR teaches ICT and Smart Money Concepts, laid out in the breakdown of what that framework actually says: order blocks, liquidity sweeps, fair value gaps, premium and discount. That is a language for reading intent inside a session. Minervini's eight rules make no claim about who is doing what or why. They describe an observable state of a chart over ten months and leave the motive out of it entirely. Two traders using both frameworks could look at the same daily chart and not exchange a single overlapping term.

Against Qullamaggie

The closest of the four, and worth being precise about the differences. Kristjan Kullamägi's setups, profiled in the write-up of his Breakout, Episodic Pivot and Parabolic Short, also select off daily charts, also demand a large prior move, and also surf rising moving averages. Three real differences. Kullamägi uses the 10 and 20 day averages where Minervini uses the 50, 150 and 200, so the horizon is shorter. Kullamägi's method is purely technical, while SEPA insists on an earnings and catalyst leg. And Kullamägi shorts parabolic tops, whereas Minervini's framework is long-only by construction, since Stage 2 is the only stage he takes. On the commercial side Kullamägi sells nothing at all, which removes the purchase decision entirely, while Minervini runs a full product ladder that is fair to evaluate on its own terms.

Can a Grading Engine Read the Trend Template?

Being precise, because this is where it would be easy to overclaim. SnapPChart reads one static chart screenshot that you upload. No live feed, no market-wide scanning, no fundamental data, no proprietary rankings. So the answer splits three ways and only one part of it is a yes.

The seven chart-derivable criteria can be read, on one condition: you have to plot the 50, 150 and 200 day moving averages before you take the screenshot. The engine reads the lines that are in the image. If they are not there, the moving average conditions are not checkable, in exactly the same way that a VWAP read needs VWAP on the chart. Rules 6 and 7 need a visible window that genuinely covers 52 weeks, which on most platforms means a daily chart zoomed out further than feels natural. Do both and the structural conditions are all present in the picture. The general mechanics of reading a stacked, sloping trend off an image are covered in how AI reads trend direction from a chart.

Rule 8 cannot be read, and there is no clever workaround. A Relative Strength rating is a percentile computed against the entire market, so it depends on data about several thousand other stocks that are not in your screenshot. SnapPChart does not compute it, does not approximate it, and does not have it. If you want that number you need a screener that maintains it. Saying otherwise would be inventing a capability, and this is the criterion where the honest answer costs something.

The fundamental and catalyst legs of SEPA are equally out of reach. There is no earnings history in a price panel and no news in a candle. So the scope of any product tie-in here is narrow and worth stating plainly: the technical Trend Template conditions plus the quality of the pivot entry, and nothing else. It is not an implementation of SEPA and it should not be described as one. What it is useful for is the thing your own eyes are worst at, which is applying identical criteria to the eleventh chart of the evening as to the first. That argument, in general form, is in grading a trade before you enter it and in the broader overview of where AI genuinely fits into a trading process. The product side lives at AI chart analysis if you want the neutral version.

One thing lines up better than you might expect. The engine grades momentum continuations only. A long is a pullback inside an established uptrend, a short is a counter-trend rally inside an established downtrend, and a chart in an uptrend never returns as a short however red the current candle is. Minervini's framework is long-only and Stage 2 only, so a chart that passes his template is by definition the exact category the long side of that grade was built to read. There is no reversal caveat to add here, which is unusual for these profiles. The caveats that remain are the missing Relative Strength rating and the missing fundamentals, and those are real. The swing timeframe also suits the workflow, since nothing is ticking while you upload a daily chart, which is the version of this discussed in AI for swing trading. What it will not do is find the names for you. It reads one chart, so the screening half of the method stays yours.