交易心理分析

Trading in the Zone

马克·道格拉斯 · Mark Douglas · 2000

心理心理·行为与叙事S1 · E03 · 第 3 集难度 ●●○○○240 页通读约 6 小时P6

赌场不知道下一把谁赢，一年 1 亿美元的注照样净赚 450 万。The casino has no idea who wins the next hand and still nets $4.5 million on every $100 million wagered.

本集角度 · 把每一笔交易当成概率分布里的一次抽样

📖 微信读书 · 打开书页本季全部 10 集

ABOUT这本书

道格拉斯 1982 年起在芝加哥做交易教练，这本书 2000 年出版，240 页 11 章。他的判断很直接：多数人亏钱的原因在心态，拿着「这一笔必须对」的想法去玩一个只有概率优势的游戏。第七章是核心，用 21 点讲赌场怎么在纯随机的结果里赚稳定的钱，4.5% 的优势加足够大的样本，然后给出五个基本事实，第一条是任何事都可能发生。第十一章是作业：选一个 edge，连做 20 笔，风险事先定死。后半本讲信念，有点绕。他 2015 年去世，这本书仍是交易心理里引用最多的。

Douglas coached traders in Chicago from 1982; this 2000 book runs 240 pages in 11 chapters. His verdict is blunt: most people lose because they bring a this-trade-must-win mindset to a game that offers only a probabilistic edge. Chapter seven is the core, using blackjack to show how a casino earns steady money from random outcomes, a 4.5% edge plus a large enough sample, then laying out five fundamental truths, the first being that anything can happen. Chapter eleven is homework: pick one edge, trade it twenty times, fix the risk in advance. The second half, on how beliefs form, meanders. Douglas died in 2015; the book remains the most cited in trading psychology.

POINTS要点

单笔结果随机，一组结果稳定，赌场靠 4.5% 的优势加足够多的手数，不靠猜下一把One hand is random, a series is steady; the casino lives on a 4.5% edge and enough hands, never on guessing the next one
你不需要知道接下来会发生什么也能赚钱，这是五个事实里最难信的一条You don't need to know what happens next to make money, the hardest of the five truths to believe
不预设风险的人，是因为相信自己知道结果，观点越强止损越松People skip predefining risk because they believe they know the outcome; the stronger the view, the looser the stop
作业是同一个 edge 连做 20 笔，风险先定死，仓位按 20 笔全亏还能接受来算The homework: one edge, twenty trades, risk fixed in advance, sized so that losing all twenty is still acceptable

EXCERPTS原文摘句 · 5 条

结果只是概率性的事件，也能产出稳定的结果，只要赔率在你这边，而且样本足够大。

Events that have probable outcomes can produce consistent results, if you can get the odds in your favor and there is a large enough sample size.

Ch.7 · 印刷页 p.102 · PDF p.121

全书的赌场原理，一句话

在 21 点这个游戏里，赌场对玩家大约有 4.5% 的优势 […] 假设一年里赌场所有 21 点桌上总共压了 1 亿美元，赌场净赚 450 万。

In blackjack, the casinos have approximately a 4.5-percent edge over the player […] Suppose a total of $100 million dollars is wagered collectively at all of a casino’s blackjack tables over the course of a year. The casino will net $4.5 million.

Ch.7 · 印刷页 p.102 · PDF p.121

优势不大，靠样本变成稳定的钱

一、任何事都可能发生。二、你不需要知道接下来会发生什么也能赚钱。三、对任何一组定义 edge 的变量，输赢的分布是随机的。四、edge 只是说明一件事发生的概率比另一件高。五、市场里每一刻都是独一无二的。

1. Anything can happen. 2. You don't need to know what is going to happen next in order to make money. 3. There is a random distribution between wins and losses for any given set of variables that define an edge. 4. An edge is nothing more than an indication of a higher probability of one thing happening over another. 5. Every moment in the market is unique.

Ch.7 · 印刷页 p.121 · PDF p.140

五个基本事实，抄在交易日志扉页那种

普通交易者不会在进场前预先定好风险，因为他不觉得有必要。他能觉得「没必要」的唯一原因，是他相信自己知道接下来会发生什么。

The typical trader won't predefine the risk of getting into a trade because he doesn't believe it’s necessary. The only way he could believe “it isn't necessary” is if he believes he knows what's going to happen next.

Ch.7 · 印刷页 p.111 · PDF p.130

不设止损的心理学，两句话拆完

潜在的风险是这 20 笔全亏。这显然是最坏的情形。它发生的概率和 20 笔全赢一样，也就是说不太可能。但它是一种可能。

the potential risk is that you will lose on all 20 trades. This is obviously the worst-case scenario. It is as likely an occurrence as that you will win on all 20 trades, which means it isn't very likely. Nevertheless, it is a possibility.

Ch.11 · 印刷页 p.197 · PDF p.218

样本 20 笔的作业，仓位要按全亏来定

GODZILLA哥斯拉怎么读

从东京的利率台子上看

「知道风险和接受风险是两回事」，台子上坐过几年的人读到都会停一下，亏大的单子全是知道、没接受。管着一本日本国债的人最懂，止损写在纸上和真的按下去之间隔着一层皮，这本书讲的是那层皮。五个事实里第二条最难，你不需要知道接下来会发生什么也能赚钱。2026 年秋 10Y 在 5% 附近、油价过百、AI 资本开支吵成一团，人人观点很强，观点一强就舍不得预设风险。过时的是第十一章的 S&P 大合约，2021 年退市了。我们台子的 edge 是客户流和资产负债表，不是指标，样本 20 笔、风险先定这两条照样管用。

Knowing the risk and accepting it are two different things; anyone who has sat on a desk stops at that line; the big losses are always knowing without accepting. Anyone running a JGB book knows that between a stop on paper and a stop actually hit there is a layer of skin; this book is about that layer. The second of the five truths is hardest: you don't need to know what happens next to make money. With the 10-year near 5%, oil above 100 and AI capex a shouting match, everyone holds a strong view, and strong views skip predefined risk. Chapter eleven's big S&P contract was delisted in 2021. Our edge is client flow and balance sheet, not an indicator; the twenty-trade sample still applies.

FOR适合谁

分析已经够用、执行总漏气的人，读第七章和第十一章。刚入门还没有自己 edge 的人先别读，书里默认你已经有一套东西可以连做 20 笔。

For those whose analysis is fine but whose execution keeps leaking: chapters 7 and 11. Beginners without an edge of their own should wait; the book assumes you already have something to trade twenty times.