[{"data":1,"prerenderedAt":1242},["ShallowReactive",2],{"\u002Fzh\u002Fglossary\u002Fwhat-is-volatility":3,"\u002Fzh\u002Fglossary\u002Fwhat-is-volatility-related":265,"\u002Fzh\u002Fglossary\u002Fwhat-is-volatility-langs":1239},{"id":4,"title":5,"body":6,"category":247,"date":248,"description":249,"draft":250,"extension":251,"meta":252,"navigation":253,"path":254,"related":255,"seo":260,"stem":261,"term":262,"updated":263,"__hash__":264},"glossary\u002Fzh\u002F5.glossary\u002Fwhat-is-volatility.md","什么是波动率？交易中如何衡量",{"type":7,"value":8,"toc":239},"minimark",[9,13,17,20,31,34,42,45,54,62,65,68,79,98,113,117,144,172,195,198,232],[10,11,12],"p",{},"波动率是收益围绕自身平均值的离散程度：看的是单个结果离均值有多远，而不是朝哪个方向。它用标准差来衡量，而且是对称的：偶尔单日赚 8% 的策略，和偶尔单日亏 8% 的策略，波动率完全一样。所以波动率不等于风险。它是一个乘数，决定了某个仓位大小会变成多大的风险。",[14,15,16],"h2",{"id":16},"工作原理",[10,18,19],{},"取一组收益（按日、按月或按笔），求出均值，再看各个结果离均值的典型距离。",[21,22,27],"pre",{"className":23,"code":25,"language":26},[24],"language-text","σ = √( Σ (rᵢ − r̄)² \u002F (n − 1) )\n\nrᵢ = 第 i 期的收益\nr̄  = 全部 n 期的平均收益\nn  = 期数\n","text",[28,29,25],"code",{"__ignoreMap":30},"",[10,32,33],{},"标准差的单位和输入一致：输入百分比，得到百分比；输入每笔交易的盈亏，得到账户货币。要在不同时间周期之间比较，就得按时间的平方根换算，所以用日数据算年化数值时要乘以 √252，也就是一年的交易日数。",[10,35,36,37,41],{},"实践中有两个特性很重要。第一，波动率",[38,39,40],"strong",{},"不是恒定的","，它会扎堆：平静的几周后面往往还是平静的几周，剧烈的日子则接二连三。第二，它是所有风险调整指标的分母：夏普比率这一类指标，正是拿收益除以这个数字，所以两个增长相同的策略，评价可能天差地别。",[14,43,44],{"id":44},"为什么重要",[10,46,47,48,53],{},"离开波动率，仓位大小就没有意义。每天波动 40 点的货币对上的 1 手，和每天波动 140 点的货币对上的 1 手，标签相同，却是两笔完全不同的押注。按手数而不是按预期波动来定仓位，最容易让账户的风险状况在交易者什么都没决定的情况下悄悄走样，这也是",[49,50,52],"a",{"href":51},"\u002Fglossary\u002Fwhat-is-position-sizing","仓位管理","规则按承担风险的金额来写、而不是按手数来写的原因。",[10,55,56,57,61],{},"波动率最终会变成",[49,58,60],{"href":59},"\u002Fglossary\u002Fwhat-is-drawdown","回撤","。下跌的深度，大致等于波动率乘以风险敞口，再乘以判断错误的持续时间。把所交易品种的波动率减半，仓位不变，您要扛过的回撤也会减半，收益也跟着减半。",[10,63,64],{},"最后，波动率会在您不知不觉中改变。在平静的季度里仓位刚好合适的策略，波动率一翻倍就成了重仓，它不用做任何新决定，就已经变得危险。止损拉宽，点差拉宽，同样的手数，承担的风险已经是当初允许的两倍。",[14,66,67],{"id":67},"数据显示了什么",[10,69,70,71,74,75,78],{},"在 ShowMyTrades 上有交易历史的公开账户中（2026 年 8 月），",[38,72,73],{},"夏普比率","中位数为 ",[38,76,77],{},"0.05","。这样的收益和周围的噪音几乎分不出来：中间水平的账户，每期的平均结果和这些期间的离散程度比起来微不足道。",[10,80,81,82,85,86,89,90,93,94,97],{},"分布的其他数字也是一样的结论：时间加权收益率中位数 ",[38,83,84],{},"+3.2%","，最深回撤中位数 ",[38,87,88],{},"9.7%","，",[38,91,92],{},"171 笔已平仓交易","，持仓时间中位数 ",[38,95,96],{},"2.4 小时","。优势很小，承受的风险却是所得结果的好几倍。",[10,99,100,101,104,105,108,109,112],{},"离散程度还解释了这些账户为什么赢得多、最后却剩得少。胜率中位数是 ",[38,102,103],{},"68.8%","，盈利因子中位数是 ",[38,106,107],{},"1.28","：大多数交易以盈利平仓，而少数亏损单亏得够大，足以吃掉大部分收益。",[38,110,111],{},"63.0%"," 的账户长期为正，算是多数，但领先的幅度，比背后结果的离散程度薄得多。",[14,114,116],{"id":115},"在-showmytrades-上的位置","在 ShowMyTrades 上的位置",[10,118,119,122,123,126,127,133,134,136,137,143],{},[38,120,121],{},"高级统计","的",[38,124,125],{},"表现指标","部分，",[49,128,130],{"href":129},"\u002Fglossary\u002Fwhat-is-standard-deviation-in-trading",[38,131,132],{},"标准差","和",[38,135,73],{},"上下相邻。这里的标准差以账户货币显示，衡量的是单笔交易结果的离散程度：标准差很大，",[49,138,140],{"href":139},"\u002Fglossary\u002Fwhat-is-expectancy",[38,141,142],{},"期望值","却很小，说明这个策略的结果很大程度上取决于样本里恰好有哪些交易。",[10,145,146,149,150,133,153,156,157,160,161,164,165,171],{},[38,147,148],{},"账户统计","面板在",[38,151,152],{},"增长",[38,154,155],{},"绝对增长","下面两行显示",[38,158,159],{},"日均 %"," 和",[38,162,163],{},"月均 %","。这两个是平均值，所以要对照",[49,166,168],{"href":167},"\u002Fguides\u002Fmonthly-returns-table-explained",[38,169,170],{},"月度收益","表来看，逐月的波动在那里一目了然：月均百分比相同的两个账户，每一行的数字可能天差地别。",[10,173,174,177,178,181,182,184,185,187,188,133,191,194],{},[38,175,176],{},"图表模块","是一个可以切换视图的面板：",[38,179,180],{},"逐笔交易增长","能清楚看出波动扎堆，",[38,183,60],{},"则显示这些离散付出了多大代价。",[38,186,121],{},"还提供",[38,189,190],{},"星期",[38,192,193],{},"小时","视图，可以看出结果集中在什么时候，而那通常也是波动集中的时候。",[14,196,197],{"id":197},"常见误解",[199,200,201,209,215,221],"ul",{},[202,203,204,205,208],"li",{},"“",[38,206,207],{},"波动率就是风险。","”风险是遭受您承受不起的亏损的可能性。波动率是双向的离散程度，完全不涉及您的资金规模。",[202,210,204,211,214],{},[38,212,213],{},"波动率低就安全。","”卖出尾部风险的策略（网格、马丁格尔、不对冲的套息交易），在那次注定让它们输掉的事件到来之前，波动率读数一直很低。",[202,216,204,217,220],{},[38,218,219],{},"夏普比率高，交易者就更优秀。","”它只说明在所测样本里，每单位离散程度换来的收益更高。只有 171 笔交易，这个估计噪音很大。",[202,222,204,223,226,227,231],{},[38,224,225],{},"波动率是交易品种本身的属性。","”它既取决于品种，也取决于时间段。同一个货币对一年里会几次切换市场状态，含同一种货币的货币对还会一起变化。请参阅",[49,228,230],{"href":229},"\u002Fglossary\u002Fwhat-is-forex-pair-correlation","外汇货币对相关性","。",[10,233,234,235,231],{},"波动率只有变成真实亏损时才真正要紧，这部分该怎么看，请参阅",[49,236,238],{"href":237},"\u002Fguides\u002Fmaximum-drawdown-explained","最大回撤详解",{"title":30,"searchDepth":240,"depth":240,"links":241},2,[242,243,244,245,246],{"id":16,"depth":240,"text":16},{"id":44,"depth":240,"text":44},{"id":67,"depth":240,"text":67},{"id":115,"depth":240,"text":116},{"id":197,"depth":240,"text":197},"指标","2026-08-19T00:00:00.000Z","波动率是收益围绕平均值的离散程度，用标准差来衡量。它本身不是风险，但同样的手数意味着什么，由它决定。",false,"md",{},true,"\u002Fzh\u002Fglossary\u002Fwhat-is-volatility",[256,257,258,259],"what-is-drawdown","what-is-position-sizing","what-is-maximum-drawdown","what-is-forex-pair-correlation",{"title":5,"description":249},"zh\u002F5.glossary\u002Fwhat-is-volatility","波动率 (Volatility)","2026-09-24T00:00:00.000Z","XCZ9I_SBNYpYPqBkPwcZX0S3YAivr3b6366b9-tzqnU",[266,466,727,1037],{"id":267,"title":268,"body":269,"category":247,"date":248,"description":456,"draft":250,"extension":251,"meta":457,"navigation":253,"path":458,"related":459,"seo":462,"stem":463,"term":464,"updated":248,"__hash__":465},"glossary\u002Fzh\u002F5.glossary\u002Fwhat-is-drawdown.md","什么是回撤？定义、公式与真实数据",{"type":7,"value":270,"toc":449},[271,274,276,279,285,288,291,325,327,330,337,340,343,353,355,360,372,375,377,395,410,412,442],[10,272,273],{},"回撤是账户价值从此前的峰值跌到之后低点的幅度，用峰值的百分比表示。它衡量的是账户离自己最好的时候跌了多少，而不是某一笔交易亏了多少。回撤是累积的，所以 20% 的回撤，可能来自四十笔小亏损，也可能来自一次巨亏。",[14,275,16],{"id":16},[10,277,278],{},"每个账户都有一条随时更新的高水位线，也就是它曾经达到的最高价值。回撤就是任意时刻账户低于这条线的距离。",[21,280,283],{"className":281,"code":282,"language":26},[24],"回撤 % = (峰值 − 当前值) \u002F 峰值 × 100\n",[28,284,282],{"__ignoreMap":30},[10,286,287],{},"这条线只升不降。账户一创新高，高水位线就移到新高，回撤归零。在那之前，只要低于峰值，每一天都算在回撤中，不管账户是在下跌还是在横盘磨蹭。",[10,289,290],{},"这个数字到底是什么意思，要看三组区分：",[199,292,293,303,309],{},[202,294,295,298,299,302],{},[38,296,297],{},"相对回撤与绝对回撤","。相对回撤就是上面的百分比。MetaTrader 报告的绝对回撤，是账户低于",[38,300,301],{},"初始入金","的金额，这是完全不同的数字：一个当前比峰值低 40% 的账户，绝对回撤可能显示为 0%。",[202,304,305,308],{},[38,306,307],{},"净值回撤与余额回撤","。净值回撤包含持仓的浮动亏损。余额回撤只看已平仓交易，所以一笔亏损只要不兑现，就永远不会出现在里面。",[202,310,311,314,315,319,320,324],{},[38,312,313],{},"当前回撤与最大回撤","。当前回撤是账户今天的位置。历史上最差的那一次是",[49,316,318],{"href":317},"\u002Fglossary\u002Fwhat-is-maximum-drawdown","最大回撤","，大多数",[49,321,323],{"href":322},"\u002Fglossary\u002Fwhat-is-a-track-record","业绩记录","引用的就是它。",[14,326,44],{"id":44},[10,328,329],{},"一个策略值不值得投资，回撤是硬约束，因为亏损和盈利的复利效果并不对称。跌 20%，要涨 25% 才能回本；跌 50%，要涨 100%。",[10,331,332],{},[333,334],"img",{"alt":335,"src":336},"亏损 10% 需要 11.1% 的盈利才能弥补，25% 需要 33.3%，50% 需要 100%，70% 需要 233%","\u002Fimages\u002Farticles\u002Frecovery-asymmetry.svg",[10,338,339],{},"跌幅一旦超过 50%，回本所需的涨幅就比亏损本身还大。",[10,341,342],{},"让人离开市场的，也是这个数字。几乎没有人会在急速暴跌中放弃，那时心里还抱着希望。人们是在不温不火的回升熬到第九个月时放弃的。深度只占一半，另一半是持续时间，而它在收益图上根本看不出来。",[10,344,345,346,133,348,352],{},"另外，回撤主要取决于",[49,347,52],{"href":51},[49,349,351],{"href":350},"\u002Fglossary\u002Fwhat-is-a-stop-loss","止损","纪律，远远超过入场质量。两位交易者信号完全一样、手数不同，胜率会一模一样，生存能力却天差地别。",[14,354,67],{"id":67},[10,356,357,358,231],{},"以下数字只描述在 ShowMyTrades 上发布的账户，并不是对全体交易者的调查。截至 2026 年 8 月，根据 15,436,464 笔已同步交易，有交易历史的已发布账户，历史最深回撤中位数为 ",[38,359,88],{},[10,361,362,363,74,367,89,369,371],{},"再看同一批账户赚了多少。",[49,364,366],{"href":365},"\u002Fglossary\u002Fwhat-is-time-weighted-return","时间加权收益率",[38,368,84],{},[38,370,111],{}," 的账户长期为正。也就是说，处在中位数的账户，一路上回吐的幅度，大约是最终收益的三倍：承受的风险比得到的结果大，而且大出的是倍数，不是一星半点。",[10,373,374],{},"让回撤数字有意义的，是这个比值，而不是回撤本身有多深。宣称收益很高、回撤却只有百分之二三的账户并非不可能，但它处在这个分布的最边缘，这种极端情况要靠账户页面上的其余内容来证明。",[14,376,116],{"id":115},[10,378,379,380,149,382,384,385,387,388,394],{},"在每个已发布账户页面上，",[38,381,148],{},[38,383,163],{}," 下方连续两行显示两个数字：",[38,386,60],{},"，基于净值计算，包含持仓的浮动亏损；以及",[49,389,391],{"href":390},"\u002Fglossary\u002Fwhat-is-drawdown-on-balance",[38,392,393],{},"余额回撤","，只基于已平仓结果。两者都是历史最大值，不是今天的读数，而且都来自经纪商的数据，不是账户所有者自己填的。",[10,396,397,398,400,401,404,405,409],{},"图表模块有一个",[38,399,60],{},"视图，用柱状图画出账户整个生命周期的每日回撤。下跌",[38,402,403],{},"持续了多久","，在这里直接就能看出来，不用靠推测，这是任何一个醒目的百分比都表达不了的。",[49,406,408],{"href":407},"\u002Ftools\u002Fdrawdown-calculator","回撤计算器","可以用您自己的余额算出回本需要多少涨幅。",[14,411,197],{"id":197},[199,413,414,420,426,436],{},[202,415,204,416,419],{},[38,417,418],{},"回撤就是我亏得最多的那一笔。","”不是。它是一条从峰值到谷底的累积路径，中间可能有几百笔交易，盈利单也包括在内。",[202,421,204,422,425],{},[38,423,424],{},"回本之后，我的回撤又回到零了。","”当前回撤是这样。最大回撤按设计永远不会降：这样，一个好季度就抹不掉一个坏季度。",[202,427,204,428,431,432,231],{},[38,429,430],{},"回撤低，说明风险低。","”新账户回撤低，通常只说明还没经受考验。这里已发布账户的已平仓交易数中位数是 171 笔；不到几百笔时，",[49,433,435],{"href":434},"\u002Fguides\u002Fhow-much-history-a-track-record-needs","回撤小反映的是样本量，而不是风险特征",[202,437,204,438,441],{},[38,439,440],{},"余额回撤才是真正的回撤。","”它只是更好看的那个。网格和摊平加仓类系统正是靠死扛亏损单不平，才让余额回撤一直很小。",[10,443,444,445,231],{},"回撤如何和账户页面上的其他数字对照着看，请参阅",[49,446,448],{"href":447},"\u002Fguides\u002Freading-a-trading-account-dashboard","如何读懂交易账户控制台",{"title":30,"searchDepth":240,"depth":240,"links":450},[451,452,453,454,455],{"id":16,"depth":240,"text":16},{"id":44,"depth":240,"text":44},{"id":67,"depth":240,"text":67},{"id":115,"depth":240,"text":116},{"id":197,"depth":240,"text":197},"回撤是账户价值从峰值到谷底的跌幅，用百分比表示。本文介绍公式、它为什么是累积的，以及数千个真实交易账户的数据。",{},"\u002Fzh\u002Fglossary\u002Fwhat-is-drawdown",[258,257,460,461],"what-is-a-stop-loss","what-is-a-profit-factor",{"title":268,"description":456},"zh\u002F5.glossary\u002Fwhat-is-drawdown","回撤 (Drawdown)","tI3oDk8FJPfBVP8sboa1RV13uVvVqDQQ3BpyQp8BpuI",{"id":467,"title":468,"body":469,"category":715,"date":248,"description":716,"draft":250,"extension":251,"meta":717,"navigation":253,"path":718,"related":719,"seo":723,"stem":724,"term":725,"updated":263,"__hash__":726},"glossary\u002Fzh\u002F5.glossary\u002Fwhat-is-forex-pair-correlation.md","什么是外汇货币对相关性？隐藏的集中风险",{"type":7,"value":470,"toc":708},[471,479,481,484,490,497,508,510,513,519,522,533,535,546,564,579,581,628,638,666,668,701],[10,472,473,474,478],{},"外汇货币对相关性，指两个货币对一起波动的程度，用介于 −1 和 +1 之间的系数表示。每个货币对都是两种货币的比价，所以只要两个货币对含有同一种货币，它们在一定程度上就是在押同一件事。持有",[49,475,477],{"href":476},"\u002Fglossary\u002Fwhat-is-correlation-risk","三个相关仓位","的交易者，并没有把风险分散到三笔交易上：他持有的其实是一个更大的仓位，而这个仓位的大小是碰巧凑出来的，不是自己决定的。",[14,480,16],{"id":16},[10,482,483],{},"这个系数等于两组收益序列的协方差，除以二者标准差的乘积。",[21,485,488],{"className":486,"code":487,"language":26},[24],"ρ = cov(A, B) \u002F (σA × σB)\n\nρ = +1.0        走势完全相同\nρ =  0.0        互不相关\nρ = −1.0        走势完全相反\n|ρ| > 0.7       计算仓位时把两者视为一个仓位\n",[28,489,487],{"__ignoreMap":30},[10,491,492,493,231],{},"背后的机制是算术，不是市场情绪。做多 EUR\u002FUSD、做多 GBP\u002FUSD、做多 AUD\u002FUSD，是做空美元的三种方式，美元一走强，三笔同时亏。做多 EUR\u002FUSD、做多 EUR\u002FGBP、做多 EUR\u002FJPY，则是做多欧元的三种方式。还有些货币对从构成上就是负相关的：EUR\u002FUSD 和 USD\u002FCHF 里的美元在相反的两侧，两个都做多，几乎等于什么都没持有，却要付两份",[49,494,496],{"href":495},"\u002Fglossary\u002Fwhat-is-a-spread","点差",[10,498,499,500,503,504,507],{},"它的危险来自两个特性。第一，相关性",[38,501,502],{},"不稳定","：最近一个月测出的系数，说的不是这两个货币对的固有属性，只是最近一个月的观察结果。第二，它",[38,505,506],{},"在压力下会升高","：正当最需要分散化的时候，资金成批涌入或撤出美元，各货币对的相关性都向 1 靠拢。",[14,509,44],{"id":44},[10,511,512],{},"相关仓位的风险是相乘的，不是相加的。合并后的风险敞口，遵循的是和任何双资产投资组合相同的公式。",[21,514,517],{"className":515,"code":516,"language":26},[24],"σ_total = √( σA² + σB² + 2 × ρ × σA × σB )\n",[28,518,516],{"__ignoreMap":30},[10,520,521],{},"ρ = 0 时，两个 1% 的仓位合起来约为 1.4% 的风险敞口；ρ = 0.9 时，合起来约为 1.95%，几乎等于两者相加。三笔这样的交易，单笔都在合理的风险限额之内，合起来却突破了账户的限额，而每一条单独的规则都没有违反。",[10,523,524,525,527,528,532],{},"这就是为什么",[49,526,52],{"href":51},"很有纪律的账户，照样会出现大幅回撤。规则一条没破，只是数的是仓位个数，而不是风险敞口；整个持仓组合的",[49,529,531],{"href":530},"\u002Fglossary\u002Fwhat-is-volatility","波动率","，是交易者自以为的两到三倍。",[14,534,67],{"id":67},[10,536,537,538,541,542,545],{},"以 ShowMyTrades 上有交易历史的公开账户为样本（2026 年 8 月），总交易量为 ",[38,539,540],{},"1,724,575 手","，账户已平仓交易数的中位数为 ",[38,543,544],{},"171 笔","。这个数量足以让集中度在品种分类统计里显现出来，也足以让一组相关仓位主导整份记录。",[10,547,548,549,552,553,555,556,559,560,563],{},"没有明说的集中度，会在回撤分布里露出马脚。",[49,550,551],{"href":317},"历史最深回撤的中位数","为 ",[38,554,88],{},"，但 ",[38,557,558],{},"38.2%"," 的账户曾从峰值下跌超过 20%，",[38,561,562],{},"17.6%"," 曾亏掉峰值的一半以上。单笔手数都不大的账户亏到这种程度，通常意味着许多仓位在同进同退。",[10,565,566,567,570,571,574,575,578],{},"用反向持仓抵消风险敞口而不平仓，是要持续付费的，这在总额里看得出来。这些账户共支付了 ",[38,568,569],{},"$4,782,670 的佣金","和 ",[38,572,573],{},"$862,547 的库存费","；在隔夜持仓的账户中，",[38,576,577],{},"86.3%"," 的净库存费为负。两个货币对互相对冲着持有，费用照样一天天累计，而当初开仓想承担的风险，大半却被抵消掉了。",[14,580,116],{"id":115},[10,582,583,586,587,590,591,594,595,133,598,601,602,594,605,608,609,612,613,616,617,620,621,133,624,627],{},[38,584,585],{},"分类统计","模块就是检查集中度的工具。它的",[38,588,589],{},"按品种","标签页列出账户交易过的每个品种，分",[38,592,593],{},"做多","、",[38,596,597],{},"做空",[38,599,600],{},"合计","三栏，每栏都有",[38,603,604],{},"交易",[38,606,607],{},"点数","和盈利，另有",[38,610,611],{},"盈利单（%）","。先看",[38,614,615],{},"品种","列：好几行都含 USD、方向又全都一样，那就是一个美元仓位拆到了几个代码上。",[38,618,619],{},"图表","视图会画出",[38,622,623],{},"按货币对统计交易",[38,625,626],{},"按货币对统计盈利","，不读表格也能看出交易是否集中在最大的几个品种上。",[10,629,630,637],{},[49,631,633,636],{"href":632},"\u002Fguides\u002Fmagic-numbers-and-strategy-labels",[38,634,635],{},"按魔术号","标签页","对策略做同样的检查：这些相关的行是同一个 EA（智能交易系统）做出来的，还是几个 EA 撞在了同一种货币上，一看便知。",[10,639,640,641,644,645,648,649,652,653,656,657,660,661,122,663,665],{},"在",[38,642,643],{},"交易表格","里看",[38,646,647],{},"开仓时间","列，留意扎堆的地方：同一分钟内在不同品种上开出好几张订单，这是一次相关的入场，不是分散的入场。",[38,650,651],{},"自定义分析","侧边面板可以按品种、魔术号以及",[38,654,655],{},"买入（做多）"," \u002F ",[38,658,659],{},"卖出（做空）"," 筛选，排除某一组货币后重新计算统计数据，看看这份记录有多少是靠它撑着的。再看",[38,662,176],{},[38,664,60],{},"视图，效果就出来了：相关的持仓组合是一下子跌下去的，而不是分好几步。",[14,667,197],{"id":197},[199,669,670,676,685,695],{},[202,671,204,672,675],{},[38,673,674],{},"一个货币对就是一个交易品种。","”它是两种货币。EUR\u002FUSD 同时是一个欧元仓位和一个美元仓位，所以凡是报价里含有这两种货币之一的品种，它都会跟着一起动。",[202,677,204,678,681,682,684],{},[38,679,680],{},"相关系数能告诉我仓位该开多大。","”它只说明两个仓位怎样一起波动，不说明波动多远。定仓位要把相关性和",[49,683,531],{"href":530},"结合起来看：两个都很平静的品种即使高度相关，问题也不大。",[202,686,204,687,690,691,694],{},[38,688,689],{},"每个货币对只有一个相关系数。","”每个货币对在",[38,692,693],{},"每个时间窗口","里各有一个数：两个品种按小时看可能毫不相干，按周看却同进同退。",[202,696,204,697,700],{},[38,698,699],{},"这只适用于主要外汇货币对。","”以美元报价的黄金、指数和加密货币对，和 EUR\u002FUSD 归在同一组的情况，比交易者想象的更常见。",[10,702,703,704,231],{},"如果想从交易量而不是盈利的角度看集中度，请参阅",[49,705,707],{"href":706},"\u002Fguides\u002Ftrading-volume-analysis","交易量分析",{"title":30,"searchDepth":240,"depth":240,"links":709},[710,711,712,713,714],{"id":16,"depth":240,"text":16},{"id":44,"depth":240,"text":44},{"id":67,"depth":240,"text":67},{"id":115,"depth":240,"text":116},{"id":197,"depth":240,"text":197},"策略","含有同一种货币的货币对会一起涨跌。相关性衡量的是这种联动有多紧，三笔看似各自独立的交易，就是因为它变成了同一个仓位。",{},"\u002Fzh\u002Fglossary\u002Fwhat-is-forex-pair-correlation",[720,721,257,722],"what-is-correlation-risk","what-is-volatility","what-is-leverage",{"title":468,"description":716},"zh\u002F5.glossary\u002Fwhat-is-forex-pair-correlation","外汇货币对相关性 (Forex Pair Correlation)","EO9V84b99m2LdKbsdx3ArhjOdhSDLW4oTuHtQ1aHh5M",{"id":728,"title":729,"body":730,"category":247,"date":248,"description":1028,"draft":250,"extension":251,"meta":1029,"navigation":253,"path":1030,"related":1031,"seo":1033,"stem":1034,"term":1035,"updated":248,"__hash__":1036},"glossary\u002Fzh\u002F5.glossary\u002Fwhat-is-maximum-drawdown.md","什么是最大回撤？公式与回本计算",{"type":7,"value":731,"toc":1021},[732,739,741,744,750,753,808,813,830,832,835,896,899,906,908,914,920,923,933,959,961,975,988,990,1016],[10,733,734,735,738],{},"最大回撤是账户整个历史上出现过的最大一次",[49,736,737],{"href":59},"从峰值到谷底的跌幅","，以峰值的百分比表示。它是策略实际造成过的最大亏损，而不是账户主人心里预期的最大亏损。它只增不减：一旦出现 34% 的最大回撤，此后无论再创多少次新高，它都停在 34%。",[14,740,16],{"id":16},[10,742,743],{},"沿着净值曲线逐点往前看，同时记下到目前为止的最高值。每到一个点，就量一量它比这个最高值低了多少。全程找到的最深跌幅，就是最大回撤。",[21,745,748],{"className":746,"code":747,"language":26},[24],"最大回撤 % = 对所有 t 取最大值 [ (峰值(0..t) − 数值(t)) \u002F 峰值(0..t) ] × 100\n",[28,749,747],{"__ignoreMap":30},[10,751,752],{},"结果完全取决于用的是哪条曲线，ShowMyTrades 两条都公布。",[754,755,756,774],"table",{},[757,758,759],"thead",{},[760,761,762,765,768,771],"tr",{},[763,764,247],"th",{},[763,766,767],{},"计算基础",[763,769,770],{},"是否计入持仓？",[763,772,773],{},"说明了什么",[775,776,777,793],"tbody",{},[760,778,779,784,787,790],{},[780,781,782],"td",{},[38,783,60],{},[780,785,786],{},"净值（余额 + 浮动盈亏）",[780,788,789],{},"是",[780,791,792],{},"账户在那一刻实际价值的真实跌幅",[760,794,795,799,802,805],{},[780,796,797],{},[38,798,393],{},[780,800,801],{},"仅已平仓余额",[780,803,804],{},"否",[780,806,807],{},"已实现结果的跌幅",[10,809,810,812],{},[49,811,393],{"href":390},"的数字更好看，而且这是计算方式决定的：未实现的亏损它根本看不到。网格、马丁格尔和摊平加仓类系统正是利用了这一点：亏损单永远不平，余额曲线自然一直平滑。",[10,814,815,816,594,819,822,823,826,827,829],{},"所以，一个账户显示",[38,817,818],{},"回撤 41%",[38,820,821],{},"余额回撤 6%","，并不是数据出了问题。它曾经比峰值低 41%，已平仓的结果却只承认有过一段 6% 的低谷。净值那个数字才是诚实的，而",[38,824,825],{},"两者之间的差距","本身就说明了策略是怎么做的。如果两个数字很接近，说明仓位基本是在亏损出现的地方附近就平掉了，这正是",[49,828,351],{"href":350},"的用处。",[14,831,44],{"id":44},[10,833,834],{},"回本是不对称的：亏损是在较大的基数上亏掉的，盈利却得在变小了的基数上靠复利赚回来。",[754,836,837,846],{},[757,838,839],{},[760,840,841,843],{},[763,842,318],{},[763,844,845],{},"回到原峰值所需涨幅",[775,847,848,856,864,872,880,888],{},[760,849,850,853],{},[780,851,852],{},"10%",[780,854,855],{},"11.1%",[760,857,858,861],{},[780,859,860],{},"20%",[780,862,863],{},"25.0%",[760,865,866,869],{},[780,867,868],{},"30%",[780,870,871],{},"42.9%",[760,873,874,877],{},[780,875,876],{},"50%",[780,878,879],{},"100.0%",[760,881,882,885],{},[780,883,884],{},"70%",[780,886,887],{},"233.3%",[760,889,890,893],{},[780,891,892],{},"90%",[780,894,895],{},"900.0%",[10,897,898],{},"回撤在 20% 以内，这种不对称只是有点麻烦。超过 50%，它就成了这个账户压倒一切的事实：您得靠刚刚让资金腰斩的同一套策略，把资金翻倍。",[10,900,901,902,905],{},"最大回撤还决定了您实际能用的杠杆上限。一个系统历史上回撤过 30%，用两倍的",[49,903,904],{"href":51},"仓位大小","去跑，就意味着 60% 的回撤，而没有任何证据表明您扛得住。",[14,907,67],{"id":67},[10,909,910,911,913],{},"这里的数字来自 ShowMyTrades 上已发布的账户，不代表全体交易者。在有交易历史的已发布账户中（2026 年 8 月），最深回撤的中位数是 ",[38,912,88],{},"，但无论往高还是往低，分布都很散。",[10,915,916],{},[333,917],{"alt":918,"src":919},"ShowMyTrades 公开账户的最深回撤：38.5% 低于 5%，23.3% 在 5% 到 20% 之间，20.6% 在 20% 到 50% 之间，17.6% 超过 50%","\u002Fimages\u002Farticles\u002Fdrawdown-distribution.svg",[10,921,922],{},"中位数 9.7%，尾部却比多数公开宣传估计的更长。",[10,924,925,926,89,929,932],{},"在最差的一端，",[38,927,928],{},"17.6% 的账户曾比峰值低 50% 以上",[38,930,931],{},"38.2% 的账户曾回撤超过 20%","。换句话说，大约每六个账户就有一个碰上过上面回本表里 50% 那一行：光是回到原来的水平，就需要 100% 的涨幅。",[10,934,935,936,939,940,89,942,946,947,949,950,953,954,958],{},"最好的一端同样值得怀疑：",[38,937,938],{},"38.5%"," 的账户最大回撤不到 5%。其中有些确实稳健保守，但很多只是资历太浅。中位数水平的已发布账户有 ",[38,941,92],{},[49,943,945],{"href":944},"\u002Fglossary\u002Fwhat-is-average-trade-length","持仓时间","中位数 ",[38,948,96],{},"；样本就这么大，最大回撤小，记录的只是“还没发生”，而不是“不会发生”。最大回撤说的是分布的尾部，而",[49,951,952],{"href":434},"尾部需要足够长的历史来支撑","，所以这个数字只有在",[49,955,957],{"href":956},"\u002Fglossary\u002Fwhat-is-a-verified-track-record","经验证的业绩记录","上才真正有分量。",[14,960,116],{"id":115},[10,962,963,964,966,967,133,969,971,972,974],{},"每个已发布账户页面的",[38,965,148],{},"面板里，",[38,968,60],{},[38,970,393],{},"紧挨着排在",[38,973,163],{}," 下方，两者都来自经纪商的数据，而非自行填报。先把这两个数字对比一下：它们之间的差距，是整个页面上最快的判断依据。",[10,976,977,978,980,981,984,985,987],{},"图表模块里有一个专门的",[38,979,60],{},"视图，用柱状图画出每日回撤，回答了醒目百分比回答不了的问题：账户在高水位线下方待了",[38,982,983],{},"多久","。七周内就收复的 25% 回撤，和十四个月后仍未收复的 25% 回撤，显示的是同一个数字，却完全不是同一种账户。",[49,986,408],{"href":407},"可以拿您自己的余额套用上面的回本表。",[14,989,197],{"id":197},[199,991,992,998,1004,1010],{},[202,993,204,994,997],{},[38,995,996],{},"我的最大回撤今年改善了。","”它不可能改善。它是历史最大值，一年业绩再好也抹不掉。",[202,999,204,1000,1003],{},[38,1001,1002],{},"两个回撤数字应该一致。","”只有仓位在亏损附近就平掉时，两者才会一致。差距大是死扛亏损单的标志，不是数据出错。",[202,1005,204,1006,1009],{},[38,1007,1008],{},"最大回撤小，风险就低。","”历史短的话就不一定。把它当风险指标之前，先问问它是多少笔交易、多少个月得出来的。",[202,1011,204,1012,1015],{},[38,1013,1014],{},"百分比回撤和金额回撤可以互相替代。","”$2,000 的账户和 $200,000 的账户各跌 30%，风险特征相同，感受却天差地别；但能在账户之间比较的，只有百分比。",[10,1017,1018,1019,231],{},"想在实时账户页面上把最大回撤和其他指标放在一起看，请阅读",[49,1020,448],{"href":447},{"title":30,"searchDepth":240,"depth":240,"links":1022},[1023,1024,1025,1026,1027],{"id":16,"depth":240,"text":16},{"id":44,"depth":240,"text":44},{"id":67,"depth":240,"text":67},{"id":115,"depth":240,"text":116},{"id":197,"depth":240,"text":197},"最大回撤是账户有史以来从峰值到谷底的最深跌幅。本文介绍计算公式、回本对照表，以及数千个账户的真实分布。",{},"\u002Fzh\u002Fglossary\u002Fwhat-is-maximum-drawdown",[256,257,1032,461],"what-is-a-verified-track-record",{"title":729,"description":1028},"zh\u002F5.glossary\u002Fwhat-is-maximum-drawdown","最大回撤 (Maximum Drawdown)","H1pMHY5DcM__vxubkgaLVDhrp8FNU0EbYexdvdzrV1o",{"id":1038,"title":1039,"body":1040,"category":1230,"date":248,"description":1231,"draft":250,"extension":251,"meta":1232,"navigation":253,"path":1233,"related":1234,"seo":1235,"stem":1236,"term":1237,"updated":248,"__hash__":1238},"glossary\u002Fzh\u002F5.glossary\u002Fwhat-is-position-sizing.md","什么是仓位管理？公式与手数计算",{"type":7,"value":1041,"toc":1223},[1042,1048,1050,1053,1059,1062,1068,1074,1077,1097,1103,1106,1108,1111,1118,1120,1132,1135,1149,1151,1158,1184,1186,1216],[10,1043,1044,1045,1047],{},"仓位管理（position sizing）决定一笔交易开多大，依据是这笔交易失败时您愿意亏掉账户的多少。它用两个数把风险百分比换算成手数：到止损的距离，以及这个手数下一个点值多少钱。账户的",[49,1046,60],{"href":59},"特征就由这一个变量决定，而且它是在交易前选定的，不是交易后才发现的。",[14,1049,16],{"id":16},[10,1051,1052],{},"分三步，按顺序来。",[21,1054,1057],{"className":1055,"code":1056,"language":26},[24],"1. 风险金额   = 账户余额 × 风险 %\n2. 点值　　   = 合约规模 × 以报价货币计的一个点 × 换算为账户货币的汇率\n3. 仓位大小   = 风险金额 \u002F (止损距离点数 × 每手点值)\n",[28,1058,1056],{"__ignoreMap":30},[10,1060,1061],{},"举个例子：$10,000 的账户，每笔交易风险 1%，EUR\u002FUSD 止损 50 点，一标准手每点 $10：",[21,1063,1066],{"className":1064,"code":1065,"language":26},[24],"风险金额   = 10,000 × 0.01      = $100\n仓位大小   = 100 \u002F (50 × 10)    = 0.20 手\n",[28,1067,1065],{"__ignoreMap":30},[10,1069,1070,1071,1073],{},"止损放宽到 100 点，同样 $100 的风险就只能开 0.10 手。风险不变，手数跟着调整。整个思路就在这种反向关系里，也正因为如此，仓位管理和",[49,1072,351],{"href":350},"位置不能分开决定。",[10,1075,1076],{},"实践中最常用的有三种方法：",[199,1078,1079,1085,1091],{},[202,1080,1081,1084],{},[38,1082,1083],{},"固定手数","。不管止损多远、余额多少，永远 0.10 手。简单，但止损和余额一变，单笔风险就跟着漂移。",[202,1086,1087,1090],{},[38,1088,1089],{},"百分比风险（固定比例）","。也就是上面的公式。风险始终与账户成比例，账户缩水时，按金额计的亏损也随之变小，回本在算术上才有可能。",[202,1092,1093,1096],{},[38,1094,1095],{},"马丁格尔递增","。亏损后加大手数，想把亏损扳回来。盈利频繁而且每笔都小，净值曲线漂亮得无可挑剔，但结果的分布有一条很厚的左尾，而且会一下子全部兑现。",[10,1098,1099],{},[333,1100],{"alt":1101,"src":1102},"每次亏损后手数加倍，对比按净值固定比例下单：胜率相同，市场相同，结局相反","\u002Fimages\u002Farticles\u002Flot-progression.svg",[10,1104,1105],{},"一个账户用的是哪一种，光看手数就能看出来。",[14,1107,44],{"id":44},[10,1109,1110],{},"入场决定您能不能赢，仓位管理决定您还能不能留下来做下一笔。两个交易者用同样的信号、同样的止损，胜率一模一样，账户却天差地别，因为一个每笔只冒 0.5% 的风险，另一个冒 8%。",[10,1112,1113,1117],{},[49,1114,1116],{"href":1115},"\u002Fglossary\u002Fwhat-is-leverage","杠杆","真正造成伤害，也是通过仓位管理。高杠杆本身不是风险，只是允许您去冒险。风险来自用这份许可开出的仓位：按它的止损距离，一笔亏损大到账户连亏两次都扛不住。",[14,1119,67],{"id":67},[10,1121,1122,1123,1125,1126,89,1129,1131],{},"仓位管理会在回撤的尾部留下痕迹。在 ShowMyTrades 上有交易历史的已发布账户中（2026 年 8 月；统计对象是在本站公开的账户，并不代表所有交易者），",[38,1124,558],{}," 曾",[49,1127,1128],{"href":317},"回撤超过 20%",[38,1130,562],{}," 亏掉了峰值的一半以上。",[10,1133,1134],{},"规模定得合理的固定比例账户，很少会落到最后这一档。按每笔 1% 的风险亏掉 50%，大约要连亏 69 次。落在这条尾部里的账户，一般另有原因：亏损后不断加码，或者定手数时根本没考虑止损。",[10,1136,1137,1138,89,1141,1144,1145,1148],{},"再看看仓位大小是谁定的：这些账户自动交易占比的中位数为 ",[38,1139,1140],{},"99%",[38,1142,1143],{},"53.9%"," 的账户自动化程度超过 90%，低于 10% 的则有 ",[38,1146,1147],{},"42.2%","。在一半以上的已发布账户上，仓位大小只是 EA 设置里的一行，设好一次，余额怎么变都不再调整。",[14,1150,116],{"id":115},[10,1152,1153,1157],{},[49,1154,1156],{"href":1155},"\u002Ftools\u002Fposition-size-calculator","仓位计算器","用实时价格、以您的账户货币套用上面的公式，风险可以按百分比或固定金额设定，止损可以用价格或点数表示。",[10,1159,1160,1161,1163,1164,1167,1168,1170,1171,1174,1175,1177,1178,1183],{},"在已发布账户的页面上，仓位管理可以实际核查，不必凭空猜测。",[38,1162,643],{},"有一列",[38,1165,1166],{},"交易量","，列出每笔交易的手数，每亏一次就加一级的手数一眼就能看出来。",[38,1169,651],{},"提供",[38,1172,1173],{},"手数","的最小\u002F最大筛选：把下限设在账户正常手数之上，看看剩下的交易是不是都紧跟在亏损之后出现，这就是马丁格尔的特征。",[38,1176,121],{},"模块里的",[49,1179,1180],{"href":706},[38,1181,1182],{},"总手数","给出汇总值；所有已发布账户累计交易了 1,724,575 手。",[14,1185,197],{"id":197},[199,1187,1188,1194,1204,1210],{},[202,1189,204,1190,1193],{},[38,1191,1192],{},"风险 2%，就是动用账户的 2%。","”不是。2% 是止损被触发时的亏损。仓位占用的保证金是另一个完全不同的数字，由杠杆决定。",[202,1195,204,1196,1199,1200,1203],{},[38,1197,1198],{},"我每笔只冒 1% 的风险，所以最坏也就亏 1%。","”只有同时只开一个仓位时才成立。",[49,1201,1202],{"href":476},"五个相关的仓位","各占 1%，其实是一笔换了装扮的 5% 交易，而相关的货币对偏偏在最伤人的时候同涨同跌。",[202,1205,204,1206,1209],{},[38,1207,1208],{},"不设止损，就没有风险要算。","”没有止损，仓位大小公式就没有分母：风险不是很小，而是根本无法定义，它的真实数值就是追加保证金通知。",[202,1211,204,1212,1215],{},[38,1213,1214],{},"复利就是不断加仓。","”固定比例的仓位管理是双向调整的。赢了加手数、亏了却不减，这不是复利，而是一个慢慢收紧的棘轮，把账户推向一场您定仓位时从未考虑过的回撤。",[10,1217,1218,1219,231],{},"全套风险与仓位管理工具，以及各自适用的场景，请参阅",[49,1220,1222],{"href":1221},"\u002Fguides\u002Fforex-calculators-guide","外汇计算器指南",{"title":30,"searchDepth":240,"depth":240,"links":1224},[1225,1226,1227,1228,1229],{"id":16,"depth":240,"text":16},{"id":44,"depth":240,"text":44},{"id":67,"depth":240,"text":67},{"id":115,"depth":240,"text":116},{"id":197,"depth":240,"text":197},"风险","仓位管理就是根据止损距离和点值，把风险百分比换算成手数。本文介绍公式、三种常见方法，以及仓位定错的代价。",{},"\u002Fzh\u002Fglossary\u002Fwhat-is-position-sizing",[256,258,460,722],{"title":1039,"description":1231},"zh\u002F5.glossary\u002Fwhat-is-position-sizing","仓位管理 (Position Sizing)","5hi2OmhGtFDhESyalituW-Nh51a9wIit5SKZxEMEQyk",[1240,1241],{"path":530},{"path":254},1793898070000]