The Best Technical Analysis Books, Judged on Testability

Technical analysis has a testability problem, and the books divide cleanly along it.

Some describe rules you could program, run against history, and discover to be worthless. Others describe patterns that can always be recognised after the fact and never quite specified in advance, which means no result can ever contradict them. Both kinds sell well. Only one kind can make you better.

That distinction is the organising principle here. We have read a great deal of this material on this desk, most of it disappointing, and what follows is the short list that survived, with what is wrong with each one stated plainly.

Start with the one that sets the standard

Aronson separates subjective technical analysis, which is unfalsifiable because you can always say the pattern had not completed, from objective rules that can actually be tested. He then tests a large batch of rules and finds that most of the apparent performance is data mining.

Read this first and everything else you read afterwards is assessed differently. You stop asking "does this pattern work" and start asking "is this claim specified precisely enough to be wrong", which is the question that separates a method from a belief.

The weakness: it front-loads the difficulty. Several hundred pages of philosophy of science and cognitive bias arrive before any market content, and the statistical machinery, including Monte Carlo permutation tests and White's Reality Check, will go over most readers' heads the first time. Read the first part, skim the mathematics, come back to it.

The reference you will actually keep

This is the standard reference and it earns the position. Murphy is comprehensive, clearly written and honest about what each tool is for. If you need to know what a given indicator is claiming, it is in here and it is explained properly.

The weakness: it is a 1990s book about a pre-electronic market. The examples are end-of-day commodity charts, and there is nothing on intraday futures, modern liquidity, or the way an electronic order book behaves. Treat it as a dictionary rather than a method, and be aware that the sections on chart patterns are exactly the material Aronson is sceptical about.

The one that explains why levels exist

Most technical analysis describes shapes on a chart. Dalton describes the auction that produced them, which is a considerably more useful thing to understand. Initiative versus responsive participation, and the day-type taxonomy, let you tell a trend day from a balance day early enough for it to matter, and that is precisely the judgement that decides whether you fade a level or go with it.

If you have ever wondered why a level held on Tuesday and failed on Wednesday, this book is the answer and almost nothing else is.

The weakness: it is jargon-dense and repetitive, with TPOs, IB, value area and excess arriving quickly. Its original context is the CBOT pit, where time-based profiles made sense because volume data was scarce; most modern platforms give you volume at price directly, which is arguably the better tool.

The one about the mechanism underneath

Not a technical analysis book, and the most useful book on this page for anyone who trades intraday. Harris takes the order book apart instruction by instruction (market, limit, stop, market-if-touched, all-or-none, hidden and reserve) and then explains who is on the other side of each.

The reason it belongs here is that a great deal of what technical analysis describes as a pattern is a visible consequence of order book mechanics. Once you understand why stops cluster where they cluster, several chart patterns stop being mysterious and start being obvious.

The weakness: it is a textbook, priced and written like one, and there are no trade setups in it. It also predates the current market structure in places, though the mechanics it describes have not changed.

The broad one, with a caveat

Elder covers psychology, risk and technique together, which is unusual and valuable. The risk chapters are the strongest part: the 2% rule per trade and the 6% monthly limit, the second being the rarer idea and the more valuable one, a circuit breaker on the losing month rather than the losing trade.

The weakness: the rest is a tour of indicators, including MACD, Force Index and Elder-Ray, that a beginner today does not need and can be harmed by taking literally. Read the risk material, treat the indicator material as history.

What we deliberately left off

The encyclopedic pattern catalogues, the candlestick compendiums and most of the price action series are omitted on purpose. Not because chart patterns never work, but because those books are built the way Aronson warns about: patterns defined loosely enough that a reader can find them anywhere, with success rates quoted from samples nobody can reproduce.

If you want to know whether a pattern has an edge, the honest route is to define it precisely enough to test and then test it, which is a skill none of those books teach and the first book on this page does.

We also left off anything promising a system. A book that will sell you a complete method for the price of a book is describing something that either does not work or would stop working the moment enough people bought it.

The order to read them in

Aronson first, for the standard of evidence. Harris second, for the mechanism. Dalton third, for the auction. Murphy alongside all of them as a reference rather than front to back. Elder's risk chapters whenever you have an account open, which is when they become real rather than theoretical.

And then the part no book covers. None of them can tell you what this morning's auction is doing, and none of them can watch you talk yourself out of your own rule in real time. That gap is what our sessions are for, and you can sit in for three trading days before deciding anything.