[{"data":1,"prerenderedAt":892},["ShallowReactive",2],{"\u002Fzh\u002Fglossary\u002Fwhat-is-trend-following":3,"\u002Fzh\u002Fglossary\u002Fwhat-is-trend-following-related":233,"\u002Fzh\u002Fglossary\u002Fwhat-is-trend-following-langs":889},{"id":4,"title":5,"body":6,"category":216,"date":217,"description":218,"draft":219,"extension":220,"meta":221,"navigation":222,"path":223,"related":224,"seo":228,"stem":229,"term":230,"updated":231,"__hash__":232},"glossary\u002Fzh\u002F5.glossary\u002Fwhat-is-trend-following.md","什么是趋势跟踪？定义与真实数据",{"type":7,"value":8,"toc":208},"minimark",[9,13,17,26,29,40,43,51,54,57,65,68,101,104,111,115,147,168,171,200],[10,11,12],"p",{},"趋势跟踪是一类策略：顺着已经展开的价格走势入场，一直持有到走势结束。它通常赢不到一半的交易，收益靠少数几笔大赚。决定它的是盈亏的形态，而不是入场信号。",[14,15,16],"h2",{"id":16},"工作原理",[10,18,19,20,25],{},"入场故意做得很普通：突破 N 日区间、",[21,22,24],"a",{"href":23},"\u002Fglossary\u002Fwhat-is-a-moving-average","移动平均线","交叉、在排过序的品种池里创新高。真正起作用的是离场：用追踪止损，行情延续就一直拿着，行情停下就回吐固定比例的利润出场。",[10,27,28],{},"之后一切都取决于一个算术等式，平台计算期望值用的也是它：",[30,31,36],"pre",{"className":32,"code":34,"language":35},[33],"language-text","期望值 = (胜率 × 平均盈利) − (亏损率 × 平均亏损)\n\n趋势示例：  0.35 × 3.0R − 0.65 × 1.0R = 每笔交易 +0.40R\n剥头皮示例：0.85 × 0.4R − 0.15 × 2.0R = 每笔交易 +0.04R\n","text",[37,38,34],"code",{"__ignoreMap":39},"",[10,41,42],{},"两个系统都赚钱。第一个差不多每三笔亏两笔，却是更强的那个。只看胜率，不看旁边的平均盈利和平均亏损，这个数字就毫无信息量。",[10,44,45,46,50],{},"收益集中在少数几笔交易里，所以",[21,47,49],{"href":48},"\u002Fglossary\u002Fwhat-is-position-sizing","仓位管理","在这里不是锦上添花，它本身就是策略。在走出大行情的那几笔上仓位太小，负责赚回一切的那段尾部就被砍掉了。",[14,52,53],{"id":53},"为什么重要",[10,55,56],{},"这种结构的代价，是大量时间都在亏损。趋势系统会长时间横着走，中间偶尔猛涨一段；人们恰恰是在横着走的阶段放弃它，而且往往就在那波本可以把一切赚回来的行情到来之前。拿一个季度去评判它，看到的只是那个季度的市场，而不是这个系统。",[10,58,59,60,64],{},"这也意味着",[21,61,63],{"href":62},"\u002Fglossary\u002Fwhat-is-drawdown","回撤","不是故障。它是入场费，每一次都得提前付。",[14,66,67],{"id":67},"数据显示了什么",[10,69,70,71,75,76,80,81,84,85,88,89,75,92,96,97,100],{},"以 ShowMyTrades 上有交易历史的公开账户为样本（2026 年 8 月），胜率中位数为 ",[72,73,74],"strong",{},"68.8%","，",[21,77,79],{"href":78},"\u002Fglossary\u002Fwhat-is-a-profit-factor","盈利因子","中位数为 ",[72,82,83],{},"1.28","，单笔持仓时间中位数为 ",[72,86,87],{},"2.4 小时","。账户已平仓交易数的中位数为 ",[72,90,91],{},"171 笔",[21,93,95],{"href":94},"\u002Fglossary\u002Fwhat-is-the-sharpe-ratio","夏普比率","为 ",[72,98,99],{},"0.05","。",[10,102,103],{},"这几乎正好是趋势跟踪的反面：胜率高、持仓短、盈利因子刚过盈亏平衡，典型的靠许多笔小赚去填偶尔一笔大亏。您被拿来比较的这群人实际是什么样子，值得心里有数。",[10,105,106,107,100],{},"样本量也很关键。按 171 笔已平仓交易的中位数算，胜率 35% 的趋势系统大约只有 60 笔盈利单，大部分收益可能就落在其中八到十笔上。拿掉这几笔，记录就是负的。这不是账户脆弱，而是这种盈亏结构注定的结果，也正因如此，这类系统",[21,108,110],{"href":109},"\u002Fguides\u002Fhow-much-history-a-track-record-needs","要看几年而不是几个月，数字才有意义",[14,112,114],{"id":113},"在-showmytrades-上的位置","在 ShowMyTrades 上的位置",[10,116,117,118,121,122,125,126,125,129,132,133,139,140,121,143,146],{},"这种结构在高级统计模块里看得最清楚。把",[72,119,120],{},"胜率","和",[72,123,124],{},"平均盈利","、",[72,127,128],{},"平均亏损",[72,130,131],{},"期望值","放在一起看：胜率低、平均盈利大，是一个运转正常的趋势系统；胜率高、平均亏损却大得多，则正好相反。",[21,134,136],{"href":135},"\u002Fglossary\u002Fwhat-is-average-trade-length",[72,137,138],{},"平均持仓时间","能看出持仓周期和宣称的风格是否相符，一个自称“趋势跟踪”、平均每笔却只持仓九分钟的账户，显然是别的东西。",[72,141,142],{},"做多胜率",[72,144,145],{},"做空胜率","能看出这份记录是真正的双向系统，还是押中了某个恰好走出趋势的市场的单边赌注。",[10,148,149,150,152,153,159,160,163,164,167],{},"图表模块里的",[72,151,63],{},"视图就是水下图：回撤有多深、多频繁，尤其是持续多久，都在上面，而持续时间正是决定一个趋势系统值不值得投资的那个数字。",[21,154,156],{"href":155},"\u002Fguides\u002Fmonthly-returns-table-explained",[72,157,158],{},"月度收益表","会把净值曲线抹平的横盘期和亏损期摆出来；在混合账户上，用",[72,161,162],{},"自定义分析","面板里的",[72,165,166],{},"交易持续时间（秒）"," 筛选，就能把长线持仓和噪音分开。",[14,169,170],{"id":170},"常见误解",[172,173,174,182,188,194],"ul",{},[175,176,177,178,181],"li",{},"“",[72,179,180],{},"胜率 35% 说明系统坏了。","”单看这个数字说明不了任何问题。把它和平均盈利、平均亏损放在一起，再看期望值。",[175,183,177,184,187],{},[72,185,186],{},"我把止损收紧一点，胜率就上去了。","”止损越紧，命中率越高，但产生收益的那段尾部也被砍掉了。要把一个赚钱的趋势系统变成亏钱的，这是最可靠的办法。",[175,189,177,190,193],{},[72,191,192],{},"连续八个月表现不佳，说明它失效了。","”对这种结构来说，八个月横盘是正常的间隔，不是判决。",[175,195,177,196,199],{},[72,197,198],{},"胜率 68.8% 的账户更安全。","”高胜率是最容易做出来、也最没有预测力的指标。下任何结论之前，先看平均亏损。",[10,201,202,203,207],{},"上面提到的每个指标，在",[21,204,206],{"href":205},"\u002Fguides\u002Freading-a-trading-account-dashboard","如何读懂交易账户控制台","中都有按模块的逐一讲解。",{"title":39,"searchDepth":209,"depth":209,"links":210},2,[211,212,213,214,215],{"id":16,"depth":209,"text":16},{"id":53,"depth":209,"text":53},{"id":67,"depth":209,"text":67},{"id":113,"depth":209,"text":114},{"id":170,"depth":209,"text":170},"策略","2026-08-19T00:00:00.000Z","趋势跟踪就是买强卖弱，用许多笔小亏损换取少数几笔大盈利。为什么盈亏结构比胜率更重要。",false,"md",{},true,"\u002Fzh\u002Fglossary\u002Fwhat-is-trend-following",[225,226,227],"what-is-a-moving-average","what-is-a-profit-factor","what-is-drawdown",{"title":5,"description":218},"zh\u002F5.glossary\u002Fwhat-is-trend-following","趋势跟踪 (Trend Following)","2026-09-24T00:00:00.000Z","ngLD5n2AzFsg76f56aGMBB6kilasnaxpqrdaUrm6vWM",[234,422,690],{"id":235,"title":236,"body":237,"category":216,"date":217,"description":411,"draft":219,"extension":220,"meta":412,"navigation":222,"path":413,"related":414,"seo":418,"stem":419,"term":420,"updated":231,"__hash__":421},"glossary\u002Fzh\u002F5.glossary\u002Fwhat-is-a-moving-average.md","什么是移动平均线？SMA 与 EMA 详解",{"type":7,"value":238,"toc":404},[239,242,244,247,253,256,259,262,264,272,275,277,296,307,310,312,315,359,361,397],[10,240,241],{},"移动平均线是某个品种最近固定若干周期的平均价格，每走出一根新 K 线就重算一次。它把起伏杂乱的价格变成一条线，跟着市场走，但总是慢半拍。这种滞后不是调调参数就能去掉的缺陷，计算方式本身就决定了它。",[14,243,16],{"id":16},[10,245,246],{},"两种常见算法的唯一区别，在于怎样给过去的价格分配权重。",[30,248,251],{"className":249,"code":250,"language":35},[33],"SMA(n) = (P1 + P2 + ... + Pn) \u002F n\n\nEMA(n): k = 2 \u002F (n + 1)\n        EMA_今日 = (今日价格 × k) + (EMA_昨日 × (1 − k))\n",[37,252,250],{"__ignoreMap":39},[10,254,255],{},"简单移动平均线（SMA）给每个周期同样的权重，新 K 线一进来，最旧的那根就整根移出窗口。这一点很少有人留意：SMA 拐头，可能只是因为一根老 K 线离开了窗口，跟眼下的行情毫无关系。",[10,257,258],{},"指数移动平均线（EMA）更看重近期价格，旧数据也从不完全丢掉，每根历史 K 线的影响按几何级数递减。所以它对方向变化反应更快，也正因为这一点，假拐点更多。",[10,260,261],{},"滞后可以量化。如果价格沿直线运行，n 周期 SMA 大约落后 (n−1)\u002F2 个周期。日线上的 200 周期均线，从算法上就决定了它反映的是大约一百天前的价格位置。周期缩短，滞后变小，但均线被噪音带着来回掉头的次数也会变多。没有哪个参数能两头兼顾。",[14,263,53],{"id":53},[10,265,266,267,271],{},"凡是建立在移动平均线上的规则，都带着这份滞后。行情启动后才入场，行情结束后才离场，每一次进出都要付",[21,268,270],{"href":269},"\u002Fglossary\u002Fwhat-is-a-spread","点差","、佣金，可能还有滑点。在趋势行情里，这些成本和吃到的行情相比不算什么；在震荡行情里，结果就全由它们决定：均线来回穿越，每穿一次就是一笔要付费的交易，账户没有哪一笔亏得惊人，却一点点失血。",[10,273,274],{},"所以移动平均线只是一个部件，不是一套策略。它只反映近期的平均价格。仓位多大、止损放哪、眼下的市场适不适合这条规则，它一概不管。",[14,276,67],{"id":67},[10,278,279,280,282,283,75,285,287,288,290,291,295],{},"以 ShowMyTrades 上有交易历史的公开账户为样本（2026 年 8 月），账户已平仓交易数的中位数为 ",[72,281,91],{},"，单笔持仓时间的中位数为 ",[72,284,87],{},[21,286,79],{"href":78},"的中位数为 ",[72,289,83],{},"。日内周期上的均线交叉规则正是这个样子：进场频繁，持仓短，超出盈亏平衡的余量很薄，结果往往由",[21,292,294],{"href":293},"\u002Fglossary\u002Fwhat-is-cost-percentage","交易成本","决定。",[10,297,298,299,302,303,306],{},"因此，评估任何均线交叉系统，都要拿交易频率来对照。这些账户总共交易了 ",[72,300,301],{},"1,724,575 手","，这笔交易量产生的佣金达 ",[72,304,305],{},"$4,782,670","。信号触发得越频繁，光是更常做对还不够：每次做对赚到的，还得足以覆盖每次交叉都要交的那笔费用。",[10,308,309],{},"有一点必须如实说明：ShowMyTrades 不会推断某笔交易出自哪个指标或 EA（智能交易系统），所以这些数字都不能单独算到均线系统头上。它们描述的是全部已发布账户这个整体，也只能如实描述这个整体。",[14,311,114],{"id":113},[10,313,314],{},"图表模块画的是净值、增长、盈利和回撤，展示的是账户，而不是价格走势。这里没有指标叠加，因为平台要做的是展示策略实际做了什么，而不是帮您设计策略。",[10,316,317,318,323,324,327,328,125,330,121,333,125,340,121,342,345,346,349,350,353,354,358],{},"您能做的，是检验一个均线系统的真实表现。在",[21,319,321],{"href":320},"\u002Fguides\u002Fcustom-analysis-and-filter-presets",[72,322,162],{},"中按",[72,325,326],{},"魔术号","筛选，就能把共用账户上的某一个 EA 单独拎出来；再用",[72,329,166],{},[72,331,332],{},"星期",[72,334,335,336,339],{},"交易时段（服务器时间）",[72,337,338],{},"缩小范围，看这条规则实际在什么时候触发。在高级统计中，把","总交易数",[72,341,138],{},[72,343,344],{},"佣金合计","放在一起看，就知道这么高的交易频率赚不赚得回成本。",[72,347,348],{},"分类统计","面板的",[72,351,352],{},"按品种","标签页能看出盈利是不是只来自某个恰好走出趋势的品种，对任何",[21,355,357],{"href":356},"\u002Fglossary\u002Fwhat-is-trend-following","趋势跟踪","规则来说，这都是最关键的问题。",[14,360,170],{"id":170},[172,362,363,374,380,386],{},[175,364,177,365,368,369,373],{},[72,366,367],{},"交叉就是系统。","”交叉只是一个入场信号。没有仓位规则、离场规则和成本假设，它算不上策略，拿它做的",[21,370,372],{"href":371},"\u002Fglossary\u002Fwhat-is-backtesting","回测","测的也是市场，而不是方法。",[175,375,177,376,379],{},[72,377,378],{},"200 日均线很重要，因为人人都在看。","”看的人多，换来的是关注，不是优势。我们的数据既不能证明、也不能否定任何特定周期，任何诚实的来源也给不了您这样的周期。",[175,381,177,382,385],{},[72,383,384],{},"EMA 更好，因为反应更快。","”反应快，意味着入场更早，假信号也更多。这是取舍，不是升级。",[175,387,177,388,391,392,396],{},[72,389,390],{},"我优化过周期，34 效果最好。","”在已经看过的历史数据上挑出表现最好的周期，这就是",[21,393,395],{"href":394},"\u002Fglossary\u002Fwhat-is-curve-fitting","曲线拟合","，不管最后挑出的数字是多少。",[10,398,399,400,100],{},"想知道一套规则化的系统到底能不能赚回成本，该看哪些指标，请参阅",[21,401,403],{"href":402},"\u002Fguides\u002Ftracking-trading-performance","追踪您的交易表现",{"title":39,"searchDepth":209,"depth":209,"links":405},[406,407,408,409,410],{"id":16,"depth":209,"text":16},{"id":53,"depth":209,"text":53},{"id":67,"depth":209,"text":67},{"id":113,"depth":209,"text":114},{"id":170,"depth":209,"text":170},"移动平均线就是最近 N 个周期的平均价格，每根 K 线更新一次。SMA 和 EMA 各自怎么算，为什么滞后是它的本性，而不是毛病。",{},"\u002Fzh\u002Fglossary\u002Fwhat-is-a-moving-average",[415,416,417],"what-is-trend-following","what-is-backtesting","what-is-curve-fitting",{"title":236,"description":411},"zh\u002F5.glossary\u002Fwhat-is-a-moving-average","移动平均线 (Moving Average)","YIgcHEcTFyB8f6E0MrzklyQbv0SNHGc7yED76RZ3vWM",{"id":423,"title":424,"body":425,"category":679,"date":217,"description":680,"draft":219,"extension":220,"meta":681,"navigation":222,"path":682,"related":683,"seo":686,"stem":687,"term":688,"updated":217,"__hash__":689},"glossary\u002Fzh\u002F5.glossary\u002Fwhat-is-a-profit-factor.md","什么是盈利因子？公式、合理数值与真实数据",{"type":7,"value":426,"toc":672},[427,430,432,438,441,444,450,453,520,522,525,533,538,540,554,561,564,571,581,583,606,631,633,666],[10,428,429],{},"盈利因子就是总盈利除以总亏损：所有盈利交易赚的钱加起来，除以所有亏损交易亏的钱。盈利因子 1.50，意思是账户每亏 $1.00，就赚回 $1.50。低于 1.00，不管胜率多高，账户从结构上就注定亏钱。",[14,431,16],{"id":16},[30,433,436],{"className":434,"code":435,"language":35},[33],"盈利因子 = 总盈利 \u002F |总亏损|\n",[37,437,435],{"__ignoreMap":39},[10,439,440],{},"注意这个公式没有考虑什么：账户交易了多久、用了多少资金、中途跌得有多深。",[10,442,443],{},"它也可以用胜率、平均盈利和平均亏损来写，这样写就能拿来诊断问题：",[30,445,448],{"className":446,"code":447,"language":35},[33],"盈利因子 = (胜率 × 平均盈利) \u002F ((1 − 胜率) × 平均亏损)\n",[37,449,447],{"__ignoreMap":39},[10,451,452],{},"两个截然不同的系统，可能算出同一个数值。胜率 30%、盈利单平均是亏损单的四倍，结果是 1.71。胜率 80%、盈利单只有亏损单的一半，结果是 2.00。两者都能赚钱，但出问题的方式不同，做起来的感受也完全两样。",[454,455,456,468],"table",{},[457,458,459],"thead",{},[460,461,462,465],"tr",{},[463,464,79],"th",{},[463,466,467],{},"解读",[469,470,471,480,488,496,504,512],"tbody",{},[460,472,473,477],{},[474,475,476],"td",{},"低于 1.00",[474,478,479],{},"亏钱，而且是结构性的，不是差一点",[460,481,482,485],{},[474,483,484],{},"1.00 – 1.10",[474,486,487],{},"还在噪声范围内，一周行情不好就没了",[460,489,490,493],{},[474,491,492],{},"1.10 – 1.30",[474,494,495],{},"优势很薄但真实存在，前提是样本够大",[460,497,498,501],{},[474,499,500],{},"1.30 – 2.00",[474,502,503],{},"策略有效",[460,505,506,509],{},[474,507,508],{},"2.00 – 3.00",[474,510,511],{},"很强。请核对交易笔数，以及成本是否算全",[460,513,514,517],{},[474,515,516],{},"高于 3.00",[474,518,519],{},"在长期历史里很少见：通常是样本太短、靠几笔超大盈利撑着，或者还有亏损单没平仓",[14,521,53],{"id":53},[10,523,524],{},"要回答“这个系统到底赚不赚钱”，盈利因子是最干净的单一指标。它也比胜率难包装得多，因为每一笔亏损都按全额计入分母。",[10,526,527,528,532],{},"但它不是风险指标。盈利因子不管顺序：一个账户一个月赚够全部利润、另外十一个月一直失血，和一个每周稳步上涨的账户，数值可以一模一样。所以要和",[21,529,531],{"href":530},"\u002Fglossary\u002Fwhat-is-maximum-drawdown","最大回撤","一起看，否则只看到了终点，没看到过程。",[10,534,535,537],{},[21,536,372],{"href":371},"给出的盈利因子同样要小心：优化器追求的就是把这个比值做到最大，所以回测里的高数值，往往反映的是拟合做得多好，而不是策略本身多好。",[14,539,67],{"id":67},[10,541,542,543,75,546,553],{},"以下数字只描述在 ShowMyTrades 上发布的账户，不代表全体交易者。截至 2026 年 8 月，有交易历史的已发布账户中，",[72,544,545],{},"盈利因子中位数为 1.28",[72,547,548,552],{},[21,549,551],{"href":550},"\u002Fguides\u002Fwin-rate-is-not-an-edge","胜率中位数","为 68.8%","。两个数字都是扣除成本后的：每笔交易先计入库存费和佣金，再判定是盈利单还是亏损单，所以 1.28 是净值，不是毛值。",[10,555,556],{},[557,558],"img",{"alt":559,"src":560},"ShowMyTrades 公开账户的中位数：68.8% 的胜率与 1.28 的盈利因子并列","\u002Fimages\u002Farticles\u002Fwin-rate-vs-payoff.svg",[10,562,563],{},"同一组数据的两个中位数：胜率看着很高，背后的优势却很薄。",[10,565,566,567,570],{},"要看的正是这两个数放在一起。中位数账户里，将近七成的交易以盈利平仓，可整体优势也只是每亏 $1.00 赚 $1.28。拿一个两项都正好处在中位数的账户，把公式倒推回去：它的平均盈利只有平均亏损的约 ",[72,568,569],{},"0.58 倍","。赢得多和赚得多是两回事，前者要容易制造得多：亏损单扛着不平，盈利单早早落袋，胜率就往上走，盈利因子却往下掉。",[10,572,573,574,576,577,580],{},"同一组账户的夏普比率中位数是 ",[72,575,99],{},"，中位数账户有 ",[72,578,579],{},"171 笔已平仓交易","。样本这么小，盈利因子高于 3.00 并不能证明系统更强，只能说明样本还太少，最差的那笔交易还没出现。",[14,582,114],{"id":113},[10,584,585,586,589,590,592,593,125,595,125,597,125,599,125,601,121,603,605],{},"每个已发布账户页面的",[72,587,588],{},"高级统计","模块里都有",[72,591,79],{},"，旁边是能解释它的几个数字：",[72,594,120],{},[72,596,124],{},[72,598,128],{},[72,600,131],{},[72,602,339],{},[72,604,95],{},"。这六个数字合起来，能说清单一比值说不清的事。",[10,607,608,609,619,620,121,622,625,626,630],{},"有两个产品细节值得注意。第一，已经算进比值的成本会另外逐项列出：交易表格有",[72,610,611,612,618],{},"盈利（毛额）",[72,613,614,615,121],{},"列，",[72,616,617],{},"库存费","佣金","各占一列；高级统计还汇总了",[72,621,344],{},[72,623,624],{},"已付库存费合计","，可以看出净值数字里扣掉了多少成本。在所有已发布账户中，这两项合计分别是佣金 $4,782,670、库存费 $862,547。第二，",[21,627,628],{"href":320},[72,629,162],{},"可以先筛选（按日期范围、品种、魔术号、方向或手数），再对子集重算整个统计区块。用它可以检验：去掉表现最好的那个季度，或者去掉最赚钱的那个品种，盈利因子是否依然站得住。",[14,632,170],{"id":170},[172,634,635,645,654,660],{},[175,636,177,637,640,641,644],{},[72,638,639],{},"盈利因子高于 1，就说明我在赚钱。","”前提是成本已经算进比值。如果没有，一个毛值 1.05 的高频系统，",[21,642,643],{"href":293},"计入佣金和库存费之后","可能持平甚至亏损。",[175,646,177,647,650,651,653],{},[72,648,649],{},"盈利因子高，风险就低。","”它反映不了回撤、仓位大小和盈亏顺序。马丁格尔式逐级加仓的",[21,652,49],{"href":48},"方案，可以一直显示 4.00，直到哪天突然失灵。",[175,655,177,656,659],{},[72,657,658],{},"不同时间周期之间可以直接比。","”不能，它没有年化。剥头皮交易者 20,000 笔交易得出的 1.20，和波段交易者 60 笔交易得出的 1.20，说的不是一回事。",[175,661,177,662,665],{},[72,663,664],{},"未平仓的交易不影响它。","”会影响，恰恰因为它们不在里面。一直不平的浮动亏损完全不计入总亏损，在平仓之前会一直把比值抬高。",[10,667,668,669,100],{},"该长期追踪哪些指标、按什么顺序看，请参阅",[21,670,671],{"href":402},"交易表现追踪完全指南",{"title":39,"searchDepth":209,"depth":209,"links":673},[674,675,676,677,678],{"id":16,"depth":209,"text":16},{"id":53,"depth":209,"text":53},{"id":67,"depth":209,"text":67},{"id":113,"depth":209,"text":114},{"id":170,"depth":209,"text":170},"指标","盈利因子是总盈利除以总亏损，低于 1.0 的账户注定亏损。本文给出各数值区间的解读，以及数千个账户的中位数 1.28。",{},"\u002Fzh\u002Fglossary\u002Fwhat-is-a-profit-factor",[227,684,685,416],"what-is-maximum-drawdown","what-is-position-sizing",{"title":424,"description":680},"zh\u002F5.glossary\u002Fwhat-is-a-profit-factor","盈利因子 (Profit Factor)","OLz2UdKZBQ-s3NM0RQISYNe6Jrhtmy0tM84z39S0Thk",{"id":691,"title":692,"body":693,"category":679,"date":217,"description":880,"draft":219,"extension":220,"meta":881,"navigation":222,"path":882,"related":883,"seo":885,"stem":886,"term":887,"updated":217,"__hash__":888},"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":694,"toc":873},[695,698,700,703,709,712,715,747,749,752,758,761,764,774,776,782,796,799,801,822,837,839,868],[10,696,697],{},"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。",[14,699,16],{"id":16},[10,701,702],{},"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。",[30,704,707],{"className":705,"code":706,"language":35},[33],"回撤 % = (峰值 − 当前值) \u002F 峰值 × 100\n",[37,708,706],{"__ignoreMap":39},[10,710,711],{},"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。",[10,713,714],{},"这个数字到底是什么意思，要看三组区分：",[172,716,717,727,733],{},[175,718,719,722,723,726],{},[72,720,721],{},"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",[72,724,725],{},"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%。",[175,728,729,732],{},[72,730,731],{},"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。",[175,734,735,738,739,741,742,746],{},[72,736,737],{},"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",[21,740,531],{"href":530},"，大多数",[21,743,745],{"href":744},"\u002Fglossary\u002Fwhat-is-a-track-record","业绩记录","引用的就是它。",[14,748,53],{"id":53},[10,750,751],{},"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%。",[10,753,754],{},[557,755],{"alt":756,"src":757},"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",[10,759,760],{},"跌幅一旦超过 50%，回本所需的涨幅就比亏损本身还大。",[10,762,763],{},"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。",[10,765,766,767,121,769,773],{},"另外，回撤主要取决于",[21,768,49],{"href":48},[21,770,772],{"href":771},"\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。",[14,775,67],{"id":67},[10,777,778,779,100],{},"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 ",[72,780,781],{},"9.7%",[10,783,784,785,80,789,75,792,795],{},"再看同一批账户赚了多少。",[21,786,788],{"href":787},"\u002Fglossary\u002Fwhat-is-time-weighted-return","时间加权收益率",[72,790,791],{},"+3.2%",[72,793,794],{},"63.0%"," 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。",[10,797,798],{},"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。",[14,800,114],{"id":113},[10,802,803,804,807,808,811,812,814,815,821],{},"在每个已发布账户页面上，",[72,805,806],{},"账户统计","面板在",[72,809,810],{},"月均 %"," 下方连续两行显示两个数字：",[72,813,63],{},"，基于净值计算，包含持仓的浮动亏损；以及",[21,816,818],{"href":817},"\u002Fglossary\u002Fwhat-is-drawdown-on-balance",[72,819,820],{},"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。",[10,823,824,825,827,828,831,832,836],{},"图表模块有一个",[72,826,63],{},"视图，用柱状图画出账户整个生命周期的每日回撤。下跌",[72,829,830],{},"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。",[21,833,835],{"href":834},"\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。",[14,838,170],{"id":170},[172,840,841,847,853,862],{},[175,842,177,843,846],{},[72,844,845],{},"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。",[175,848,177,849,852],{},[72,850,851],{},"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。",[175,854,177,855,858,859,100],{},[72,856,857],{},"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，",[21,860,861],{"href":109},"回撤小反映的是样本量，而不是风险特征",[175,863,177,864,867],{},[72,865,866],{},"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。",[10,869,870,871,100],{},"回撤如何和账户页面上的其他数字对照着看，请参阅",[21,872,206],{"href":205},{"title":39,"searchDepth":209,"depth":209,"links":874},[875,876,877,878,879],{"id":16,"depth":209,"text":16},{"id":53,"depth":209,"text":53},{"id":67,"depth":209,"text":67},{"id":113,"depth":209,"text":114},{"id":170,"depth":209,"text":170},"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",[684,685,884,226],"what-is-a-stop-loss",{"title":692,"description":880},"zh\u002F5.glossary\u002Fwhat-is-drawdown","回撤 (Drawdown)","tI3oDk8FJPfBVP8sboa1RV13uVvVqDQQ3BpyQp8BpuI",[890,891],{"path":356},{"path":223},1793898060732]