Technical Analysis Of Stock Trends By Edwards And Magee

The Method That Actually Worked Before Bloomberg Terminals Existed

Edwards and Magee's framework is built on three foundational premises that most people skim past because they sound too obvious. Price moves in trends. Trends persist until they don't. History repeats itself because human psychology doesn't change between 1948 and 2026. The book itself, Technical Analysis Of Stock Trends By Edwards And Magee, was first published in 1948 and has gone through nine editions since, which tells you something about its durability. I spent about six weeks properly learning the Dow Theory foundations before I ever touched their chart patterns. That's the part nobody emphasizes enough. If you jump straight into head-and-shoulders formations without understanding what the authors meant by primary, secondary, and minor trends, you'll misclassify breakouts constantly. The difference between a secondary decline and a primary downtrend reversal is where most retail traders lose money. I learned this the hard way with a mid-cap tech stock in 2019 that broke below a clear support level, bounced for what looked like a bullish retest, and then dropped another forty percent over six weeks. I'd labeled it a pullback within an uptrend when it was actually the start of a new primary downtrend. The authors actually address this directly in the chapters on trend classification, but it's easy to gloss over while you're excited about the pattern-recognition stuff.

Core Mechanics Of Technical Analysis Of Stock Trends By Edwards And Magee

The system rests on three types of chart patterns: reversal patterns, continuation patterns, and transitional patterns. Reversal patterns include the head and shoulders (and its inverse), double tops and bottoms, and rounding bottoms. Continuation patterns are mostly triangles—symmetrical, ascending, and descending—and flags and pennants. Transitional patterns are less discussed but equally important: they mark the consolidation phase between a completed trend and an emerging one. Volume confirmation is non-negotiable in their framework. Every breakout needs volume expansion to be considered valid, and every breakdown should ideally see volume increase as well. The authors provide specific volume ratios they consider meaningful, though these vary by market capitalization and average daily turnover. For large-cap stocks trading over a million shares per day, a breakout with volume at 1.5 to 2 times the 20-day average is what they'd call convincing. For smaller names, the threshold is higher because false signals are more common. I use a rolling 50-day volume comparison now instead of the 20-day because it smooths out earnings-related spikes that distort shorter windows. Narrowing price ranges signal exhaustion. When you see a triangle forming with progressively tighter swings, the authors treat this as a coiling spring—the wider the initial range and the longer the consolidation, the more significant the eventual move. This isn't speculation on their part. They documented this empirically across thousands of chart examples spanning multiple market cycles from the 1920s through the 1940s.

Practical Implementation That Isn't In The Textbooks

The biggest gap between reading Edwards and Magee and actually using their system is trend duration management. The book tells you how to identify a trend. It doesn't give you a clean rule for when to exit a position based purely on trend structure. I developed a personal workaround that combines their trend-line methodology with a simple moving average filter. Once a primary trend shows two consecutive closes beyond a 50-day SMA on the opposite side of the trend line, I reduce position size by half. A third close triggers full exit. This isn't from the book. It's something I added after watching my stop-loss orders get repeatedly hunted during choppy markets between 2017 and 2020. Gap analysis gets short shrift in casual discussions of this material. Edwards and Magee dedicate significant attention to breakaway, measurement, and runaway gaps, and they're worth studying in sequence. A breakaway gap on high volume that holds for five to ten trading days is your earliest confirmed signal that a new primary trend is beginning. Most traders miss this because they're looking for a clean breakout above resistance rather than tracking gap behavior. I keep a simple spreadsheet logging every gap on my watchlist stocks, noting the date, direction, volume relative to the 50-day average, and whether the gap held. After about forty entries, you start seeing patterns in how gaps behave across different market conditions. The authors' concept of "priced-in" expectations is particularly relevant for earnings-adjacent trading. If a stock has been in a strong uptrend and then gaps up on earnings but immediately sells off, Edwards and Magee would classify this as the news being absorbed into the existing trend structure rather than a reversal signal. The market had already moved in anticipation. This distinction matters enormously when you're trading around quarterly results and your model keeps giving you false reversal warnings.

Where The System Breaks Down

Let me be direct about what Edwards and Magee's framework cannot handle. Algorithmic high-frequency trading has fundamentally changed market microstructure in ways the authors couldn't have anticipated. The volume patterns they described were generated by human traders and institutional order flow. Today, a significant portion of volume on large-cap names comes from market makers and algorithmic participants who don't respect the same support and resistance levels. This doesn't invalidate the system, but it does mean that volume confirmation signals are less reliable than they were in the pre-2010 era. I've found that adjusting my volume thresholds upward by roughly thirty percent brings my results back in line with what the original framework predicts. Another limitation is the sheer amount of subjective judgment required. Identifying a head and shoulders pattern seems straightforward on a textbook chart. On a real-time screen, you're often in the middle of the formation and can't tell if you're looking at a legitimate pattern or a distribution phase that will continue lower. I've wasted considerable time shorting stocks that turned out to be normal pullbacks within strong uptrends because I called a pattern too early. The workaround is patience. Wait for the neckline break and the retest before acting. The book advocates this, but it's psychologically difficult to do when you're sitting on unrealized losses from a premature entry. The framework also struggles with low-liquidity small-cap and micro-cap stocks. The patterns require enough trading volume to produce reliable technical signals, and many sub-$2 billion names simply don't have the participation needed. I limit my application of this system to stocks with at least fifty million shares in average daily volume and a market cap above two billion dollars. Outside that universe, price action is too easily distorted by single large orders.

Chart Setup And Daily Workflow

I run my charts on TradingView, which lets me overlay multiple timeframes and apply custom indicators without much friction. My default setup includes daily candles, a 50-day and 200-day simple moving average, a 50-day average volume bar, and trend lines drawn manually. I don't use automated pattern-recognition tools because they generate far too many false positives on the kind of data Edwards and Magee's system was designed to interpret. The human eye catching structural nuances in the price action is still more reliable than any algorithm for this particular methodology. Each morning, I scan for stocks that have recently completed a recognizable pattern—usually from my weekend review—and I mark their measured move targets using the authors' width-of-the-pattern method. A head and shoulders target is measured from the neckline down by the distance from the head to the neckline. A triangle target uses the width of the widest part of the triangle projected forward from the breakout point. These aren't guarantees. They're probabilities dressed up as geometry, which is exactly what the authors intended them to be. The measured move targets tend to hold more often on larger, more liquid names. I've seen target accuracy drop from roughly sixty-five percent on S&P 500 constituents to around forty percent on less liquid mid-caps, which aligns with my observation about the liquidity requirement I mentioned earlier.

What Beginners Miss

The most counter-intuitive point in the entire book is that a successful breakout requires the market to actually close beyond the critical level, not just trade there intraday. This seems simple but traders ignore it constantly, especially when a stock gaps through resistance and closes near its high the next morning. That close is what matters. Edwards and Magee are emphatic about this, yet I see people enter positions the moment price touches a breakout level on a live screen, then wonder why they get stopped out when the price reclaims the other side by the close. The second thing people skip is the chapter on market breadth. The authors use breadth indicators like the number of stocks making new highs versus new lows to confirm whether a broad market trend is genuine or narrow and fragile. This is arguably the most useful part of the system for timing entries and exits in index funds and ETFs. When the S&P 500 breaks out to new highs but only a third of its components are participating, the authors would consider the breakout suspect. That nuance alone has saved me from riding several false breakouts over the years. There's no single download or software package that implements this methodology fully. The book itself is available through multiple publishers and used copies are frequently under twenty dollars. Some technical analysis platforms offer pattern-recognition features loosely inspired by Edwards and Magee, but they're approximations at best. The system requires you to draw your own trend lines, identify your own patterns, and make your own judgment calls. That's not a bug. It's the entire point.