Technical Analysis
Technical Analysis For Dummies
Rockefeller spends more pages telling you why indicators fail than telling you how to use them, and that ratio is why this beats most beginner books. The whole thing is built around one claim: your stop-loss rule decides whether you survive, not your indicator.
The one idea
The book's real argument is that technical analysis is a risk management discipline wearing a forecasting costume. Rockefeller states flatly that some of the best indicators work less than half the time under normal conditions, and that the biggest error beginners make is trusting the tools too much. She does not treat that as a problem to be engineered away. She treats it as the design constraint you build around.
So the sequence she teaches runs backwards from what most people expect. You do not find a signal and then figure out where to get out. You decide what a loss costs you, size the position from that number, place the stop at the same moment you place the entry, and only then care which indicator got you in. Chapter 5 says it outright: of the five steps in a trading plan, step four (setting the stop and target) is the one that decides the outcome, and you can trade mediocre securities with mediocre indicators and still net a gain if you control losses.
The second half of the argument is about self-deception. She shows a 10-day moving average crossover returning 68.6 percent over 1,000 days, then subtracts ten dollars per trade and watches it collapse to 28.6 percent, below the 43.4 percent you would have made doing nothing. She shows a moving average chart returning 61 percent and then tells you she rigged it by shopping for an ideal chart. She says the perfect backtested parameter stops being perfect the moment you add new data. For a book with a cartoon on the cover, the skepticism is unusually load-bearing.
What it actually teaches
Chapter 5 gives an explicit five-step trading plan, and the next twelve chapters exist to fill in the tools for each step.
FIND THE TREND
She gives rule-based trendlines so you cannot draw the line you want to see. For an uptrend: start at the lowest low, connect to the next low that precedes a new high, redraw as new highs come in, stop when new highs stop, then extend at the same slope. Entry is the third touch of that support line by a bar's low. Exit is any bar low falling below it. She notes two touches is a real trendline too, just a riskier one. The 20-day moving average is her one non-judgmental trend marker, because everyone gets the same line. The linear regression is the mathematically true trendline, but she warns it has no trading rule attached and calls mean-reversion systems built on it nonsense.
WRITE THE ENTRY RULE, THEN CONFIRM IT
Pick one ruling concept, trend-following or swing trading, and stop mixing them. Trend-following entries she actually names: price crossing a moving average, the moving average level rule (today's average lower than yesterday's ends the uptrend, which usually fires earlier than a crossover), the two-average crossover, Donchian's 5/20 system, and Crabel's opening range breakout. Then add one non-duplicate confirmation. Her own experience: adding 12-day momentum as a second requirement cut trade count 30 to 50 percent without sacrificing much profit, and the trades it removed were mostly losers. Do not confirm momentum with momentum. She cites Tushar Chande finding momentum and one common relative strength variant over 90 percent correlated, which means two lines telling you the same thing twice.
SIZE THE POSITION OFF THE STOP
Three sizing methods, all worked with numbers. The Turtles' 2 percent rule: on 10,000 dollars of risk capital you lose no more than 200 dollars on one trade, and she immediately points out that is only 50 consecutive losses to zero, or ten each if you trade five securities. Van Tharp's R: an 8 dollar security with initial risk R of 2 dollars gets a stop that costs less than R (say 1.50) and a target of 2R (4 dollars), but you then check the security's actual 20-period high-low range, and if that range is only 3 dollars the 4 dollar target is fantasy and you cut it to 3, accepting 2:1. Ralph Vince's fixed fractional: 10,000 in capital, a 40 dollar stock, a 2 dollar stop, and a 2.5 percent cap on portfolio loss (250 dollars) permits 6.3 shares, which she admits is not economically tradeable. That admission is the point. The math tells you the trade is too big for the account.
SET THE EXIT BEFORE YOU ENTER
The stop goes in with the entry order, always. She calls mental stops hogwash and a delusion, on the grounds that nobody watches every minute and traders with mental stops sit and watch the loss grow. Her exit menu: the last-three-days rule (out if price breaks the lowest low of the prior three days), a break of support or resistance, the confirmation point of a double top or bottom, the 10-day average as a warning to trim with the 20-day as the actual stop, Wilder's parabolic stop and reverse, an average true range stop set 25 percent beyond a normal range (so a 3 dollar average range triggers past 3.75), and LeBeau's chandelier exit hung one to three ATRs below the highest high since entry. On profit targets she is honest that nobody has solved it. Her options are a fixed multiple of R, capturing 75 percent of the average range, or letting an indicator call it, which she describes as winging it.
DECIDE HOW YOU GET BACK IN
This is the step everyone skips and she flags it as unsolved. She asked around 30 system designers and traders whether reentry should use the same rules as entry. Roughly half said same, half said different. So there is no answer, only your written answer. Alongside it she prescribes a trading diary: print the chart, write down every override you wanted to make, print it again later and mark whether you were right. Her line in the sand is a quarter. Override more than a quarter of your signals and you are still a technical trader but no longer a system trader.
What it looks like Monday morning
You stop entering trades without an exit order. Every position gets a stop typed in at the same second as the entry, and the size of the position gets calculated from that stop rather than from how much you feel like buying. That single change is the book's whole payload, and it is a Monday-morning change, not a research project.
Second, you re-run whatever indicator you already trust with a realistic cost per trade subtracted, because her Chapter 4 tables show the ranking of parameters flipping once ten dollars of slippage goes in. If the edge dies at ten dollars a trade, you found out on a spreadsheet instead of over six months of live fills. Third, you start the override diary, which costs nothing and is the only way to find out whether your gut is an asset or the thing draining the account.
| Moving average | Trades | Gain, no costs | Gain after $10 per trade |
|---|---|---|---|
| 10-day | 178 | 68.6% | 28.6% |
| 31-day | 32 | 59.3% | 49.3% |
| 35-day | 47 | 61.7% | 31.7% |
Buy-and-hold over the same 1,000 days returned 43.4 percent, so the 10-day winner beats doing nothing by 25 points before costs and loses to it after.
The biggest mistake that beginning technical traders make is attributing too much reliability and accuracy to technical methods.Barbara Rockefeller, Technical Analysis For Dummies
Where it fails
- The profit tables are cherry-picked and she says so in the fine print. Chapter 12 shows a moving average level rule returning 61 percent, a two-average crossover returning 36 percent, and Chapter 3 shows an on-balance-volume trade returning 42 percent in a month, which she annualizes to about 500 percent. Only after the tables does she write that she rigged the case by hunting for an ideal chart, and that for every one like it thousands exist where the same rule produces heartache. She also states she is conveniently not subtracting commissions and fees from any figure in that chapter. A beginner who skims tables and skips prose walks away with numbers that are worse than useless.
- Every pattern statistic comes from one borrowed study of 500 stocks, 1991 to 1996. The double bottom delivering a profit 97 percent of the time after confirmation, head-and-shoulders at 93 percent, the dead-cat bounce at 90 percent, the ascending triangle failure rate dropping from 32 percent to 2 percent if you wait for a close above the line, the symmetrical triangle breaking down 57 percent of the time: all of it is Tom Bulkowski's Encyclopedia of Chart Patterns dataset from a five-year window ending in 1996. That is before decimalization, before Reg NMS, before algorithmic execution. Bulkowski himself re-ran and revised his pattern numbers in later editions. Rockefeller quotes the older set and does not date-stamp them for the reader. Sizing a position on a 97 percent number sourced that way is exactly the overconfidence the rest of the book warns against.
- The market-structure chapter is now the most out-of-date thing in the book. Chapter 20 is literally titled Ten Ways the Market Has Changed and it stops in 2011. Algorithmic trading is described as maybe 75 percent of US volume, the May 2010 flash crash is the cautionary tale, retail forex leverage was just cut to 50 to 1, and there are nearly a thousand ETFs. The software appendix sends you to Metastock, TC2000, eSignal, NeoTicker and a ninjatraderpro.com URL. Nothing on zero-commission retail brokerage, payment for order flow, crypto, 0DTE options, or the modern data stack. Any specific number, threshold or vendor in this write-up is from the 2nd edition (2011) and the buy button may be selling a later printing with different figures.
- It assumes a charting package with an optimizer and a lot of screen time. Chapter 4 is built on being able to order software to test every moving average from 10 to 50 days over 1,000 days and get results in minutes, then re-test with a slippage assumption, then re-test on 500 days of out-of-sample data. That is a real assumption about your tooling in 2011 and it filters harder than it looks. The setup trading section in Chapter 16 goes further: it wants intense concentration during market hours, active stop management, and says you really should be flat when you go on vacation. If you have a day job, half of Part V is describing a life you do not have.
- The position-sizing math gets named and then handed off. She introduces R, fixed fractional sizing and maximum adverse excursion, then points you at Van Tharp, Ralph Vince and John Sweeney to actually learn them. The one fully worked sizing example produces 6.3 shares and she then tells you not to trade it. Money management is the step she says decides everything, and it gets roughly ten pages, less than she gives to candlesticks. If you take her thesis seriously, this is the book's own priorities contradicting themselves.
- The author is a currency trader and the fingerprints are everywhere. Rockefeller publishes a daily FX newsletter and consults on FX charts. When she gives a personal number it is an FX number: the pound, the euro and the yen trend roughly 60 percent of the time by her measures. The clock-based rules are about the 11 a.m. New York European close. The warning that big players hunt your obvious stops is FX market lore. None of that is wrong, but a stock trader is getting risk advice calibrated to a 24-hour market with no earnings dates and no halts.
- It quietly does not work for anyone who has read one other technical book. Chapters 6 through 15 are a competent survey of bars, gaps, candlesticks, patterns, trendlines, channels, moving averages, momentum, volatility and point-and-figure, at roughly fifteen pages each. That is enough to recognize a thing, not enough to trade it. Elliott Wave gets a sidebar and a shrug. Gann gets three pages of gentle debunking. Market profile, volume profile and Ichimoku are absent. And Chapter 17 tells you flatly not to build or buy a trading system, so a reader who came looking for one leaves empty handed.
Who it is for
Buy it if
You are new, you have an account funded and nothing written down about when you get out. This is the book that will make you place a stop with every entry and size the trade off that stop instead of off enthusiasm. It is also the right first book if you have already decided technical analysis is nonsense, because Rockefeller concedes most of the critics' points in Chapters 1, 3 and 4 and then explains what is left.
Skip it if
You already trade with rules and a written plan. Seventy percent of this will be revision, and the parts that are not revision send you to other books anyway. Skip it too if you want depth on one method. For chart patterns go to Edwards and Magee or straight to Bulkowski, whose numbers she is quoting secondhand. And skip it if you expect a system, because she spends a chapter arguing you should not build or buy one.
As an Amazon Associate I earn from qualifying purchases. It costs you nothing extra, and it does not change what gets recommended or what the verdict says.
Reviewed here is the second edition (2011). The link goes to the current edition. Rockefeller's five-step plan carries over; the specific pattern statistics quoted above are from the 2011 printing.
Back to all titles