[{"data":1,"prerenderedAt":1177},["ShallowReactive",2],{"\u002Fzh\u002Fglossary\u002Fwhat-is-average-trade-length":3,"\u002Fzh\u002Fglossary\u002Fwhat-is-average-trade-length-related":252,"\u002Fzh\u002Fglossary\u002Fwhat-is-average-trade-length-langs":1174},{"id":4,"title":5,"body":6,"category":234,"date":235,"description":236,"draft":237,"extension":238,"meta":239,"navigation":240,"path":241,"related":242,"seo":247,"stem":248,"term":249,"updated":250,"__hash__":251},"glossary\u002Fzh\u002F5.glossary\u002Fwhat-is-average-trade-length.md","什么是平均持仓时间？公式与真实数据",{"type":7,"value":8,"toc":226},"minimark",[9,13,17,20,31,34,42,49,52,55,58,77,90,93,103,114,125,129,143,166,193,196,219],[10,11,12],"p",{},"平均持仓时间是仓位从开仓到平仓的平均时长，统计范围是账户上的每一笔已平仓交易。它反映的是策略实际的持仓周期（以经纪商的记录为准），而不是策略设计时设想的周期。要检验一套宣称的策略和公开的业绩记录是不是一回事，这是最快的一项检查。",[14,15,16],"h2",{"id":16},"工作原理",[10,18,19],{},"每笔已平仓交易都有开仓时间和平仓时间。这项指标就是两者间隔的算术平均值，由我们的服务器根据账户的交易记录算好，发送到页面直接显示。",[21,22,27],"pre",{"className":23,"code":25,"language":26},[24],"language-text","平均持仓时间 = Σ (平仓时间 − 开仓时间) \u002F 同时带有两个时间戳的已平仓交易笔数\n","text",[28,29,25],"code",{"__ignoreMap":30},"",[10,32,33],{},"这个公式有三个特点，决定了该怎么看它。",[10,35,36,37,41],{},"它是",[38,39,40],"strong",{},"均值，不是中位数","。一笔开了三个月的仓位，和四百笔只持续九十秒的交易算在同一个平均值里，它对结果的影响比其中任何一笔都大得多。所以平均值短，基本可以说明交易确实短；平均值长，却可能只是几个离群值造成的。",[10,43,44,45,48],{},"它",[38,46,47],{},"只算已平仓交易","，而且只算买入和卖出订单：入金、出金和已取消的挂单都不算。一笔浮亏扛了一年的仓位，平仓之前完全不计入，等到终于平仓那天，平均值会突然跳升。",[10,50,51],{},"仓位逐笔计算，不合并。一个同时开着二十个订单的网格会记下二十个持续时间，所以这个数字回答的是“一笔交易开了多久”，而不是“账户在风险里暴露了多久”。",[14,53,54],{"id":54},"为什么重要",[10,56,57],{},"持仓周期是交易记录里藏不住的结构性事实。一份号称波段交易的记录，平均持仓却只有四分钟，那就不是波段交易，介绍怎么写都没用。两种风格本身没有高下之分，问题在于说的和做的对不上。",[10,59,60,61,66,67,71,72,76],{},"它还决定了",[62,63,65],"a",{"href":64},"\u002Fglossary\u002Fwhat-is-cost-percentage","哪类成本占大头","。",[62,68,70],{"href":69},"\u002Fglossary\u002Fwhat-is-a-spread","点差","和佣金每笔交易付一次，所以对持仓最短的交易负担最重。",[62,73,75],{"href":74},"\u002Fglossary\u002Fwhat-is-a-swap-in-forex","库存费","按夜计算，只针对跨过展期（rollover）时点的仓位。两个毛利相同、持仓周期不同的系统，钱花在了完全不同的地方。",[10,78,79,80,84,85,89],{},"它也决定了",[62,81,83],{"href":82},"\u002Fguides\u002Fhow-much-history-a-track-record-needs","一份业绩记录在单位时间内能证明多少","。平均持仓两小时的账户很快就能积累证据；平均持仓三周的账户，要好几年才能攒够同样数量的独立结果。所以持仓周期必须和",[62,86,88],{"href":87},"\u002Fglossary\u002Fwhat-is-a-profit-factor","盈利因子","一起看。",[14,91,92],{"id":92},"数据显示了什么",[10,94,95,96,99,100,66],{},"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，",[38,97,98],{},"持仓时间中位数为 2.4 小时","，账户已平仓交易数中位数为 ",[38,101,102],{},"171 笔",[10,104,105,106,109,110,113],{},"这是一个以日内交易为主的群体。中位数账户在同一个交易日内开仓、平仓，大部分仓位从没跨过展期时点，可这些记录里仍有 ",[38,107,108],{},"$862,547 的库存费","，而且",[38,111,112],{},"产生过库存费的账户中，86.3% 的净库存费为负","。隔夜成本集中在少数持仓过夜的账户上，对它们来说，这是一笔实实在在的支出，不是可以忽略的零头。",[10,115,116,117,120,121,124],{},"自动化的数字解释了中位数为什么这么短。",[38,118,119],{},"自动交易占比中位数为 99%","，",[38,122,123],{},"53.9% 的账户自动化程度超过 90%","。程序在规则触发时平仓，不会等到一天工作结束。",[14,126,128],{"id":127},"在-showmytrades-上的位置","在 ShowMyTrades 上的位置",[10,130,131,134,135,138,139,142],{},[38,132,133],{},"平均持仓时间","是",[38,136,137],{},"高级统计","表格视图中",[38,140,141],{},"交易","一栏的最后一行，账户所有者的控制台和公开账户页面上都一样。单位会自动调整：不到一小时显示分钟，不到一天显示保留一位小数的小时，再长就显示天和小时。",[10,144,145,146,149,150,153,154,157,158,161,162,165],{},"同一面板还能把持仓时间展开成分布。",[38,147,148],{},"持续时间","按钮（表格 \u002F 星期 \u002F 小时 \u002F 持续时间切换中的第四个）会打开一张名为",[38,151,152],{},"持仓时间与盈利能力","的散点图：横轴是持仓小时数，纵轴是盈亏，气泡大小代表以手计的交易量，分为",[38,155,156],{},"盈利交易","和",[38,159,160],{},"亏损交易","两组。有个限制要知道：它只画",[38,163,164],{},"最近 200 笔已平仓交易","，不是全部历史，反映的是当前的交易方式，而不是整个生命周期。图上每个结果都已扣除佣金和库存费。鼠标悬停可以看到订单号、品种、交易量、持续时间、盈亏，以及精确到分钟的开仓和平仓时间。",[10,167,168,169,172,173,175,176,179,180,179,183,179,186,189,190,66],{},"在单笔交易层面，",[38,170,171],{},"交易历史与余额变化","表格有一个可选的",[38,174,148],{},"列（仅限已平仓交易，默认关闭，可在列菜单中开启），格式为 ",[28,177,178],{},"2d 5h","、",[28,181,182],{},"3h 14m",[28,184,185],{},"12m",[28,187,188],{},"4m 20s"," 或 ",[28,191,192],{},"42s",[14,194,195],{"id":195},"常见误解",[197,198,199,207,213],"ul",{},[200,201,202,203,206],"li",{},"“",[38,204,205],{},"平均值就代表典型的交易。","”它是均值，而持仓时间有一条很长的右尾。先看看持续时间散点图，再判断平均值能不能代表任何一笔真实的交易。",[200,208,202,209,212],{},[38,210,211],{},"平均值短，就证明是剥头皮。","”也可能是一个系统几分钟就平掉盈利单，亏损单却扛上几周：平均值变短了，风险却拉长了。",[200,214,202,215,218],{},[38,216,217],{},"持仓也算在内。","”平仓之前不算。一笔长期持有的浮动仓位，在这里从设计上就是看不见的。",[10,220,221,222,66],{},"持仓周期如何和账户页面上的其他数字对照着看，请参阅",[62,223,225],{"href":224},"\u002Fguides\u002Freading-a-trading-account-dashboard","如何读懂交易账户控制台",{"title":30,"searchDepth":227,"depth":227,"links":228},2,[229,230,231,232,233],{"id":16,"depth":227,"text":16},{"id":54,"depth":227,"text":54},{"id":92,"depth":227,"text":92},{"id":127,"depth":227,"text":128},{"id":195,"depth":227,"text":195},"指标","2026-08-19T00:00:00.000Z","平均持仓时间是已平仓仓位从开仓到平仓的平均时长。本文介绍公式、为什么它是均值而不代表典型交易，以及数千个账户的数据。",false,"md",{},true,"\u002Fzh\u002Fglossary\u002Fwhat-is-average-trade-length",[243,244,245,246],"what-is-expectancy","what-is-a-profit-factor","what-is-a-lot-in-forex","what-is-terminal-ping",{"title":5,"description":236},"zh\u002F5.glossary\u002Fwhat-is-average-trade-length","平均持仓时间 (Avg. Trade Length)","2026-09-17T00:00:00.000Z","cNRr16YihC40iUp7Oj1a2qe2HNL7B77C1i1hmhGTHlg",[253,450,724,944],{"id":254,"title":255,"body":256,"category":436,"date":235,"description":437,"draft":237,"extension":238,"meta":438,"navigation":240,"path":439,"related":440,"seo":445,"stem":446,"term":447,"updated":448,"__hash__":449},"glossary\u002Fzh\u002F5.glossary\u002Fwhat-is-a-lot-in-forex.md","外汇中的手（Lot）是什么？标准手、迷你手与微型手",{"type":7,"value":257,"toc":429},[258,261,263,266,272,275,281,289,291,294,301,317,319,334,351,353,372,386,392,400,402,422],[10,259,260],{},"手是衡量交易规模的单位。一个标准手等于该货币对基础货币的 100,000 单位，迷你手是 10,000，微型手是 1,000。所以在 EUR\u002FUSD 上开 0.10 手，控制的就是 10,000 欧元。业绩记录上的所有交易量数字，以及经纪商收取的几乎所有成本，都是按手来计的。",[14,262,16],{"id":16},[10,264,265],{},"手是一个乘数，乘在经纪商为每个品种设定的合约规模上。",[21,267,270],{"className":268,"code":269,"language":26},[24],"交易单位 = 手数 × 合约规模\n\n标准手   1.00  = 100,000 单位\n迷你手   0.10  =  10,000 单位\n微型手   0.01  =   1,000 单位\n",[28,271,269],{"__ignoreMap":30},[10,273,274],{},"合约规模的作用，是把价格变动换算成金额。五位小数报价的外汇货币对，一个点是 0.0001；日元货币对是 0.01。点值由此可以直接算出：",[21,276,279],{"className":277,"code":278,"language":26},[24],"点值 = 手数 × 合约规模 × 点的大小，再换算为账户货币\n　　 = 1.00 × 100,000 × 0.0001 = EUR\u002FUSD 每点 $10\n",[28,280,278],{"__ignoreMap":30},[10,282,283,284,288],{},"100,000 只是外汇的惯例，并不是硬性规定。黄金、指数、能源和加密货币 CFD 各有各的合约规模，由经纪商设定，在品种规格里可以查到。所以 XAUUSD 上的“一手”和 EUR\u002FUSD 上的“一手”不是一回事。",[62,285,287],{"href":286},"\u002Fglossary\u002Fwhat-is-a-cent-account","美分账户","则是把账户货币缩小，同样的名义手数，实际规模又小了一截。",[14,290,54],{"id":54},[10,292,293],{},"经纪商就是按手计费的。佣金按每手报价，分单边收取或按来回收取；库存费按每手每晚报价；点差成本等于点差点数乘以点值，而点值本身又取决于合约规模。这些费用没有一项按交易笔数增长，全都随交易量增长。",[10,295,296,297,66],{},"所以，要比较两个账户，分母应该是总交易量。只看美元成本说明不了什么，大账户做大仓位，本来就该付得多。每手成本把账户规模这个因素去掉了，也只有用这种形式，才能",[62,298,300],{"href":299},"\u002Fguides\u002Fbroker-cost-audit","把一家经纪商的收费和另一家放在一起比",[10,302,303,304,308,309,311,312,316],{},"交易量还是交易者唯一完全由自己决定的数字。",[62,305,307],{"href":306},"\u002Fguides\u002Fwin-rate-is-not-an-edge","胜率","、回撤和",[62,310,88],{"href":87},"都是结果；手数则是一道指令，在市场还没开口之前就已经下达。所以单看交易量这一列，比看任何表现指标都能更快看出",[62,313,315],{"href":314},"\u002Fglossary\u002Fwhat-is-position-sizing","仓位管理","有没有纪律。",[14,318,92],{"id":92},[10,320,321,322,325,326,329,330,333],{},"以下数字只代表在 ShowMyTrades 上发布的账户，并非全体交易者。截至 2026 年 8 月，这里的公开账户共交易了 ",[38,323,324],{},"1,724,575 手","，平均每个有交易历史的已发布账户约 ",[38,327,328],{},"770 手","，分布在 ",[38,331,332],{},"703 个不同的经纪商服务器","上。",[10,335,336,337,340,341,346,347,350],{},"这些交易量对应的成本：在此同步的交易共产生 ",[38,338,339],{},"$4,782,670 佣金","和 ",[38,342,343,344],{},"$862,547 ",[62,345,75],{"href":74},"，在付过库存费的账户中，",[38,348,349],{},"86.3% 的净库存费为负","。别急着拿一个数去除另一个数，算出一个抢眼的费率：两个总额对应的账户群体并不相同，里面还混着合约规模不同的品种、按单边报价和按来回报价的经纪商，以及根本不收佣金的纯点差账户。每手成本确实是正确的比较方式，只不过要放在单个账户上算，那里的两个数字来自同一批交易。",[14,352,128],{"id":127},[10,354,355,358,359,361,362,364,365,157,368,371],{},[38,356,357],{},"总手数","在",[38,360,137],{},"的",[38,363,141],{},"栏里，保留两位小数，正下方就是",[38,366,367],{},"佣金合计",[38,369,370],{},"已付库存费合计","。这三项都是对账户已平仓买卖交易的汇总，由我们的服务器计算，不含入金和出金。两项成本都沿用经纪商记录的正负号，所以佣金通常显示为负数；账户付出的库存费多于收到的，库存费就显示为红色。三行放在一起看，就能为单个账户算出上面说的每手成本，依据的是经纪商的真实记录，而不是公开的价目表。",[10,373,374,375,377,378,381,382,385],{},"具体到每笔交易，",[38,376,171],{},"中的",[38,379,380],{},"交易量","列列出每个订单的规模，精确到两位小数。账户所有者可以在账户隐私设置里用",[38,383,384],{},"手数","开关，把这一列从公开页面和小组件中隐藏。所以公开账户页面上没有交易量列，是所有者有意为之，并不是数据缺失。",[10,387,388,389,391],{},"同一个高级统计面板里的",[38,390,148],{},"视图用气泡大小表示交易量，涵盖最近 200 笔已平仓交易。手数一步步加码，在这里看到的就是越来越大的圆圈，而不只是一列数字。",[10,393,394,395,399],{},"下单前想算好仓位，可以用",[62,396,398],{"href":397},"\u002Ftools\u002Fposition-size-calculator","仓位计算器","：输入账户余额、风险和止损距离，就能得出手数，合约规模可选微型手（1,000）、迷你手（10,000）或标准手（100,000）。",[14,401,195],{"id":195},[197,403,404,410,416],{},[200,405,202,406,409],{},[38,407,408],{},"0.01 手是小仓位。","”那要看跟什么比。EUR\u002FUSD 上一个微型手每点风险约 $0.10；在 $200 的账户上设 50 点止损，就是净值的 2.5%。",[200,411,202,412,415],{},[38,413,414],{},"总手数代表我的交易规模。","”它只是个合计。同样的总数，可能来自一万笔微型手交易，也可能来自一百笔标准手交易：分布情况要看交易表格。",[200,417,202,418,421],{},[38,419,420],{},"交易量越大，交易者越厉害。","”交易量反映的是活跃程度和成本，完全看不出这些交易有没有赚钱。",[10,423,424,425,66],{},"想把交易量列当作风险指纹来解读，请参阅",[62,426,428],{"href":427},"\u002Fguides\u002Ftrading-volume-analysis","交易量分析指南",{"title":30,"searchDepth":227,"depth":227,"links":430},[431,432,433,434,435],{"id":16,"depth":227,"text":16},{"id":54,"depth":227,"text":54},{"id":92,"depth":227,"text":92},{"id":127,"depth":227,"text":128},{"id":195,"depth":227,"text":195},"成本","手是衡量交易规模的单位，一个标准手等于 100,000 单位基础货币。本文讲清合约规模的算法，并附上 170 万手真实交易量的数据。",{},"\u002Fzh\u002Fglossary\u002Fwhat-is-a-lot-in-forex",[441,442,443,444],"what-is-position-sizing","what-is-a-spread","what-is-a-swap-in-forex","what-is-average-trade-length",{"title":255,"description":437},"zh\u002F5.glossary\u002Fwhat-is-a-lot-in-forex","手 (Lot)","2026-08-29T00:00:00.000Z","fij9ydCEZ30hDoAqII7e_JOwXlNKxdpqFWWeD1XoHSI",{"id":451,"title":452,"body":453,"category":234,"date":235,"description":713,"draft":237,"extension":238,"meta":714,"navigation":240,"path":715,"related":716,"seo":720,"stem":721,"term":722,"updated":235,"__hash__":723},"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":454,"toc":706},[455,458,460,466,469,472,478,481,548,550,553,561,568,570,582,589,592,599,610,612,639,664,666,699],[10,456,457],{},"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。",[14,459,16],{"id":16},[21,461,464],{"className":462,"code":463,"language":26},[24],"盈利因子 = 总盈利 \u002F |总亏损|\n",[28,465,463],{"__ignoreMap":30},[10,467,468],{},"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。",[10,470,471],{},"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：",[21,473,476],{"className":474,"code":475,"language":26},[24],"盈利因子 = (胜率 × 平均盈利) \u002F ((1 − 胜率) × 平均亏损)\n",[28,477,475],{"__ignoreMap":30},[10,479,480],{},"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。",[482,483,484,496],"table",{},[485,486,487],"thead",{},[488,489,490,493],"tr",{},[491,492,88],"th",{},[491,494,495],{},"解读",[497,498,499,508,516,524,532,540],"tbody",{},[488,500,501,505],{},[502,503,504],"td",{},"低于 1.00",[502,506,507],{},"亏钱，而且是结构性的，不是差一点",[488,509,510,513],{},[502,511,512],{},"1.00 – 1.10",[502,514,515],{},"还在噪声范围内，一周行情不好就没了",[488,517,518,521],{},[502,519,520],{},"1.10 – 1.30",[502,522,523],{},"优势很薄但真实存在，前提是样本够大",[488,525,526,529],{},[502,527,528],{},"1.30 – 2.00",[502,530,531],{},"策略有效",[488,533,534,537],{},[502,535,536],{},"2.00 – 3.00",[502,538,539],{},"很强。请核对交易笔数，以及成本是否算全",[488,541,542,545],{},[502,543,544],{},"高于 3.00",[502,546,547],{},"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",[14,549,54],{"id":54},[10,551,552],{},"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。",[10,554,555,556,560],{},"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",[62,557,559],{"href":558},"\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。",[10,562,563,567],{},[62,564,566],{"href":565},"\u002Fglossary\u002Fwhat-is-backtesting","回测","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。",[14,569,92],{"id":92},[10,571,95,572,120,575,581],{},[38,573,574],{},"盈利因子中位数为 1.28",[38,576,577,580],{},[62,578,579],{"href":306},"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。",[10,583,584],{},[585,586],"img",{"alt":587,"src":588},"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",[10,590,591],{},"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。",[10,593,594,595,598],{},"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 ",[38,596,597],{},"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。",[10,600,601,602,605,606,609],{},"同一组账户的夏普比率中位数是 ",[38,603,604],{},"0.05","，中位数账户有 ",[38,607,608],{},"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。",[14,611,128],{"id":127},[10,613,614,615,617,618,620,621,179,623,179,626,179,629,179,632,157,635,638],{},"每个已发布账户页面的",[38,616,137],{},"模块里都有",[38,619,88],{},"，旁边是能解释它的几个数字：",[38,622,307],{},[38,624,625],{},"平均盈利",[38,627,628],{},"平均亏损",[38,630,631],{},"期望值",[38,633,634],{},"总交易数",[38,636,637],{},"夏普比率","。这六个数字合起来，能说清单一比值说不清的事。",[10,640,641,642,651,652,157,654,656,657,663],{},"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",[38,643,644,645,650],{},"盈利（毛额）",[38,646,647,648,157],{},"列，",[38,649,75],{},"佣金","各占一列；高级统计还汇总了",[38,653,367],{},[38,655,370],{},"，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，",[62,658,660],{"href":659},"\u002Fguides\u002Fcustom-analysis-and-filter-presets",[38,661,662],{},"自定义分析","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。",[14,665,195],{"id":195},[197,667,668,678,687,693],{},[200,669,202,670,673,674,677],{},[38,671,672],{},"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，",[62,675,676],{"href":64},"计入佣金和库存费之后","可能持平甚至亏损。",[200,679,202,680,683,684,686],{},[38,681,682],{},"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",[62,685,315],{"href":314},"方案，可以一直显示 4.00，直到哪天突然失灵。",[200,688,202,689,692],{},[38,690,691],{},"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。",[200,694,202,695,698],{},[38,696,697],{},"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。",[10,700,701,702,66],{},"该长期追踪哪些指标、按什么顺序看，请参阅",[62,703,705],{"href":704},"\u002Fguides\u002Ftracking-trading-performance","交易表现追踪完全指南",{"title":30,"searchDepth":227,"depth":227,"links":707},[708,709,710,711,712],{"id":16,"depth":227,"text":16},{"id":54,"depth":227,"text":54},{"id":92,"depth":227,"text":92},{"id":127,"depth":227,"text":128},{"id":195,"depth":227,"text":195},"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",[717,718,441,719],"what-is-drawdown","what-is-maximum-drawdown","what-is-backtesting",{"title":452,"description":713},"zh\u002F5.glossary\u002Fwhat-is-a-profit-factor","盈利因子 (Profit Factor)","OLz2UdKZBQ-s3NM0RQISYNe6Jrhtmy0tM84z39S0Thk",{"id":725,"title":726,"body":727,"category":234,"date":235,"description":935,"draft":237,"extension":238,"meta":936,"navigation":240,"path":937,"related":938,"seo":940,"stem":941,"term":942,"updated":250,"__hash__":943},"glossary\u002Fzh\u002F5.glossary\u002Fwhat-is-expectancy.md","什么是期望值？公式、示例与真实数据",{"type":7,"value":728,"toc":928},[729,732,734,740,747,750,756,759,761,767,775,781,783,795,812,829,831,842,875,892,894,923],[10,730,731],{},"期望值就是一笔交易的平均结果，可以用货币或点数表示：账户每开一笔仓位，平均能赚多少。它由胜率、平均盈利、亏损率、平均亏损四个数字算出，回答的是“这套系统每做一笔到底赚不赚钱”，而且结果可以直接乘以交易笔数。正期望值重复得足够多，就是一门生意；负期望值则是在倒数计时。",[14,733,16],{"id":16},[21,735,738],{"className":736,"code":737,"language":26},[24],"期望值 = (胜率 × 平均盈利) − (亏损率 × 平均亏损)\n\n亏损率 = 1 − 胜率\n",[28,739,737],{"__ignoreMap":30},[10,741,742,743,746],{},"假设一个系统胜率 40%，每笔盈利单赚 $300，每笔亏损单亏 $100：(0.40 × $300) − (0.60 × $100) = $120 − $60 = ",[38,744,745],{},"每笔交易 +$60","。一年做四百笔，预期结果就是 $24,000，这还没考虑仓位大小的变化，也没算上任何意外。",[10,748,749],{},"把同一个公式改用 R 倍数来写（R 即一笔平均亏损），账户货币的影响就去掉了，不同系统也能放在一起比较：",[21,751,754],{"className":752,"code":753,"language":26},[24],"期望值 (R) = (胜率 × 盈亏比) − (1 − 胜率)\n\n盈亏比 = 平均盈利 \u002F 平均亏损\n",[28,755,753],{"__ignoreMap":30},[10,757,758],{},"算出来的数字有没有意义，取决于两点。第一，输入里必须已经扣除成本：点差、佣金和库存费每笔都要付，盈利单因此变小，亏损单因此变大。第二，平均值背后要有足够多的交易，不能让某一笔特别大的结果单独左右它。",[14,760,54],{"id":54},[10,762,763,764,766],{},"统计数字要变成交易计划，靠的就是期望值。",[62,765,88],{"href":87},"告诉您赚的钱和亏的钱之比；期望值告诉您再多做一笔交易值多少钱，把它乘以交易频率，就能得出一个月的预期结果。",[10,768,769,770,774],{},"胜率掩盖的取舍，也会在期望值里暴露出来。胜率只是优势的一半，另一半是",[62,771,773],{"href":772},"\u002Fglossary\u002Fwhat-is-risk-reward-ratio","风险回报比","，期望值正是两者的交汇点。动其中任何一个，结果都会跟着变；两个都不动，故事讲得再好，结果也不会变。",[10,776,777,778,780],{},"期望值还划出了成本的底线。如果一个策略每笔的期望是 +$8，而一个来回的",[62,779,70],{"href":69},"加佣金就要 $7，剩下的那点东西已经算不上策略了。",[14,782,92],{"id":92},[10,784,785,786,120,789,120,792,66],{},"以下数字描述的是 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）取中间水平：",[38,787,788],{},"胜率中位数 68.8%",[38,790,791],{},"盈利因子中位数 1.28",[38,793,794],{},"已平仓交易数中位数 171 笔",[10,796,797,798,800,801,804,805,66],{},"有了这几个数字，这个“中间账户”的期望值就能算出来。把盈利因子公式反推回去，平均盈利约为平均亏损的 ",[38,799,597],{},"；剩下的用 R 形式一算便知：(0.688 × 0.58) − 0.312 = ",[38,802,803],{},"+0.087R","。也就是说，每做一笔交易，只赚到一笔平均亏损的百分之九。171 笔交易下来，全部利润加起来大约只相当于十五笔平均亏损，所以尽管近七成交易以盈利平仓，",[38,806,807,811],{},[62,808,810],{"href":809},"\u002Fglossary\u002Fwhat-is-time-weighted-return","时间加权收益率","的中位数也只有 +3.2%",[10,813,814,815,340,817,820,821,824,825,828],{},"优势薄到这个程度，成本就不是可以忽略的零头了。已发布账户累计支付了 ",[38,816,339],{},[38,818,819],{},"$862,547 库存费","；在产生过库存费的账户中，",[38,822,823],{},"86.3%"," 的净库存费为负。持仓时间中位数只有 ",[38,826,827],{},"2.4 小时","，在这种节奏下，0.087R 的优势要想保住，一个来回的成本相对平均亏损就必须足够小。",[14,830,128],{"id":127},[10,832,614,833,617,835,837,838,841],{},[38,834,137],{},[38,836,631],{},"这一行，位于",[38,839,840],{},"表现指标","分组。它在同一行里给出两种单位（先点数，后账户货币）：点数便于在不同品种之间比较，货币金额便于在不同仓位大小之间比较，单看哪一个都不完整。",[10,843,844,845,847,848,179,850,179,852,157,854,856,857,862,863,157,865,867,868,179,870,157,872,874],{},"它的四个输入在同一模块的",[38,846,141],{},"分组里：",[38,849,307],{},[38,851,625],{},[38,853,628],{},[38,855,634],{},"，旁边还有",[62,858,860],{"href":859},"\u002Fglossary\u002Fwhat-is-average-trade-length",[38,861,133],{},"，能看出这点薄薄的单笔优势是几小时赚到的，还是几周才赚到的。已经扣除的成本显示在",[38,864,367],{},[38,866,370],{},"两行，每笔交易的明细则在交易表格的",[38,869,644],{},[38,871,75],{},[38,873,650],{},"列。",[10,876,877,878,882,883,179,886,157,889,891],{},"想检验一个期望值到底是一种优势，还是几种优势平均出来的，可以打开",[62,879,880],{"href":659},[38,881,662],{},"侧边面板：按品种、魔术号、方向、日期范围、星期或小时筛选出子集，整个区块就会重新计算。高级统计还提供",[38,884,885],{},"星期",[38,887,888],{},"小时",[38,890,148],{},"视图，不用筛选就能把同一结果拆开来看。",[14,893,195],{"id":195},[197,895,896,905,911,917],{},[200,897,202,898,901,902,904],{},[38,899,900],{},"正期望值就意味着我会赚钱。","”它只说明平均每笔交易是盈利的。仓位定错了，正期望值的系统照样会亏光：能不能活下来，是",[62,903,315],{"href":314},"的问题，不是期望值的问题。",[200,906,202,907,910],{},[38,908,909],{},"期望值就是我的平均每笔交易。","”前提是平均值按净额计算。按毛利润算就漏掉了佣金和库存费，而薄优势恰恰是在这里消失的。",[200,912,202,913,916],{},[38,914,915],{},"高胜率就意味着正期望值。","”单看胜率说明不了任何问题。盈亏比为 0.58 时，60% 的胜率是个亏钱的系统。",[200,918,202,919,922],{},[38,920,921],{},"这个数字很稳定。","”只有几百笔交易时并不稳定。一笔特别大的盈利单就能把整个记录的期望值抬上去，把它去掉再算一遍，是个公平的检验。",[10,924,925,926,66],{},"哪些指标值得长期追踪、该按什么顺序看，请参阅",[62,927,705],{"href":704},{"title":30,"searchDepth":227,"depth":227,"links":929},[930,931,932,933,934],{"id":16,"depth":227,"text":16},{"id":54,"depth":227,"text":54},{"id":92,"depth":227,"text":92},{"id":127,"depth":227,"text":128},{"id":195,"depth":227,"text":195},"期望值是单笔交易的平均盈亏：(胜率 × 平均盈利) − (亏损率 × 平均亏损)。本文给出公式、完整算例和公开账户的真实数据。",{},"\u002Fzh\u002Fglossary\u002Fwhat-is-expectancy",[244,939,441,717],"what-is-risk-reward-ratio",{"title":726,"description":935},"zh\u002F5.glossary\u002Fwhat-is-expectancy","期望值 (Expectancy)","4gLLgYBgHrrvWKTKFixk2tUfM9kXODaMx4LcyeirnhQ",{"id":945,"title":946,"body":947,"category":1161,"date":235,"description":1162,"draft":237,"extension":238,"meta":1163,"navigation":240,"path":1164,"related":1165,"seo":1170,"stem":1171,"term":1172,"updated":250,"__hash__":1173},"glossary\u002Fzh\u002F5.glossary\u002Fwhat-is-terminal-ping.md","什么是终端延迟（Ping）？与交易服务器之间的往返时间",{"type":7,"value":948,"toc":1154},[949,955,957,960,966,969,972,1006,1009,1011,1014,1017,1020,1022,1034,1046,1055,1057,1089,1103,1116,1118,1142],[10,950,951,952,66],{},"终端延迟（ping）是 MetaTrader 终端与经纪商交易服务器之间的往返时间，也就是一个数据包发出去再回来要多久，单位是毫秒。它取决于网络路径，与上面跑的策略无关：在这段空档里，您请求的价格只是一个愿望。MetaTrader 以微秒报告这个值，ShowMyTrades 换算成毫秒并保留一位小数，所以 12,245 µs 显示为 ",[38,953,954],{},"12.2 ms",[14,956,16],{"id":16},[10,958,959],{},"终端会记下最近一次测到的、与交易服务器之间的往返时间。订单的每一步，都叠加在这个时间之上。",[21,961,964],{"className":962,"code":963,"language":26},[24],"Ping (ms) = 终端 ↔ 交易服务器之间最近一次测得的往返时间\n\n订单延迟 ≈ ping\u002F2   （请求发出）\n          + 经纪商撮合与校验\n          + ping\u002F2   （确认返回）\n",[28,965,963],{"__ignoreMap":30},[10,967,968],{},"三项里只有中间那一项归经纪商。第一项和第三项取决于地理位置和路由，所以同一家经纪商，同一数据中心里的终端读数是 3 ms，地球另一端的家庭网络则是 280 ms。",[10,970,971],{},"参考区间：",[197,973,974,984,990],{},[200,975,976,979,980,66],{},[38,977,978],{},"低于 50 ms","：优秀，卡片的提示框里用的就是这个词。通常是离经纪商数据中心很近的 ",[62,981,983],{"href":982},"\u002Fglossary\u002Fwhat-is-a-vps","VPS",[200,985,986,989],{},[38,987,988],{},"50 到 150 ms","：家庭或办公室网络连接区域服务器的正常水平。",[200,991,992,995,996,1000,1001,1005],{},[38,993,994],{},"高于 200 ms","（提示框的警告阈值）：",[62,997,999],{"href":998},"\u002Fglossary\u002Fwhat-is-order-execution","订单执行","明显变慢，实际",[62,1002,1004],{"href":1003},"\u002Fglossary\u002Fwhat-is-slippage","滑点","扩大。",[10,1007,1008],{},"这些到底有多要紧，取决于一个区别。止损单和止盈单存放在交易服务器上，不在终端里，所以连接再差也会照常触发。Ping 只会拖慢由终端发起的操作：开仓、手动平仓、改单，以及所有写在 EA（智能交易系统）内部、而不是作为服务器端订单存在的追踪逻辑。",[14,1010,54],{"id":54},[10,1012,1013],{},"在市价执行下，延迟会直接变成成本：成交价是请求到达那一刻的价格，而不是请求发出时屏幕上的那个价格。200 ms 的往返就是五分之一秒，这段时间里市场可以随意变动，而且偏偏在订单最要紧的时候动得最厉害。",[10,1015,1016],{},"这五分之一秒贵不贵，完全看目标。持仓几个小时、瞄准五十点行情的单子，根本感觉不到。目标只有三到五点的系统，同样的时间里可能丢掉相当一部分优势，而且笔笔如此。回测对此只字不提：回测里永远按请求价格瞬间成交。",[10,1018,1019],{},"怪经纪商之前，先看 ping。读数 300 ms 时成交差，其实是网络问题，只是顶着经纪商的名字。",[14,1021,92],{"id":92},[10,1023,1024,1025,1028,1029,1033],{},"以下数字只描述在 ShowMyTrades 上发布的账户，不代表交易者整体。只有所有者自己的终端运行我们的 EA，才会有终端读数；而 ",[38,1026,1027],{},"1 万+ 个已连接的账户","中，绝大多数用的是 ",[62,1030,1032],{"href":1031},"\u002Fglossary\u002Fwhat-is-metatrader","MetaTrader","。cTrader 和 TradeLocker 账户通过经纪商的 API 连接，没有本地终端，也就没有 ping 可报。",[10,1035,1036,1037,120,1040,1042,1043,1045],{},"这批账户对延迟非常敏感。",[38,1038,1039],{},"自动交易占比的中位数为 99%",[38,1041,123],{},"，几乎每个订单都是程序发出的，一路往返，没有人停下来核对价格。这些账户连接着 ",[38,1044,332],{},"，并不存在统一的延迟环境。",[10,1047,1048,1054],{},[38,1049,1050,1053],{},[62,1051,1052],{"href":859},"持仓时间中位数","为 2.4 小时","，处于中位数的账户，根本到不了毫秒定胜负的程度。以秒为单位交易的那部分账户，才是这个数字值得放在页面上的原因。",[14,1056,128],{"id":127},[10,1058,1059,1062,1063,1066,1067,1070,1071,1073,1074,179,1077,179,1080,179,1083,340,1086,66],{},[38,1060,1061],{},"Ping"," 是",[38,1064,1065],{},"终端","面板六个卡片中的第一个，在账户所有者自己的控制台里、该账户的页面上。MT5 上标题为",[38,1068,1069],{},"终端与已打开的图表","，因为终端还会报告已打开的图表；MT4 上标题只是",[38,1072,1065],{},"，因为 MT4 的 EA 不收集这份列表。其余卡片是",[38,1075,1076],{},"重传",[38,1078,1079],{},"终端版本号",[38,1081,1082],{},"内存",[38,1084,1085],{},"自动交易",[38,1087,1088],{},"DLL 导入",[10,1090,1091,1092,179,1095,1098,1099,1102],{},"标题旁边有状态徽章（",[38,1093,1094],{},"已连接",[38,1096,1097],{},"未连接交易服务器","或",[38,1100,1101],{},"离线","），以及读数是多久以前的。数据新不新说了算，标志说了不算：超过 11 分钟没有更新，账户就会显示为离线，不管终端最后报告的是什么。",[10,1104,1105,1106,1110,1111,1115],{},"有两个限制要注意。这个面板只有所有者能看到，提供数据的接口会拒绝访问别人的账户，所以终端读数，包括已打开图表上的 ",[62,1107,1109],{"href":1108},"\u002Fglossary\u002Fwhat-is-an-expert-advisor","EA"," 名称，从不出现在公开页面或",[62,1112,1114],{"href":1113},"\u002Fguides\u002Ftrading-widgets-and-embeds","小组件中","。另外，它只适用于 MetaTrader：cTrader 和 TradeLocker 账户没有终端；终端报告过一次之前，这个面板也根本不会显示。",[14,1117,195],{"id":195},[197,1119,1120,1126,1136],{},[200,1121,202,1122,1125],{},[38,1123,1124],{},"Ping 低，经纪商就好。","”它衡量的是路径，不是交易对手。读数 2 ms、成交却很差的服务器，不过是一条通往坏价格的快车道。",[200,1127,202,1128,1131,1132,1135],{},[38,1129,1130],{},"一次读数就能说明问题。","”它只是最近一次同步时的值。稳定比数值本身更重要，而稳定性正是",[62,1133,1076],{"href":1134},"\u002Fglossary\u002Fwhat-is-packet-retransmission","补上的信息。",[200,1137,202,1138,1141],{},[38,1139,1140],{},"有了 VPS，执行就没问题了。","”VPS 解决的是网络路径。撮合速度、报价源和点差政策，一样都没变。",[10,1143,1144,1145,1149,1150,66],{},"同一段往返，如果从网络的角度而不是从这个卡片的角度来讲，就是",[62,1146,1148],{"href":1147},"\u002Fglossary\u002Fwhat-is-server-latency","服务器延迟","。同一策略在两家经纪商运行时，怎样把延迟和成本分开看，见",[62,1151,1153],{"href":1152},"\u002Fguides\u002Fsame-strategy-different-brokers","同一策略、两家经纪商、不同结果",{"title":30,"searchDepth":227,"depth":227,"links":1155},[1156,1157,1158,1159,1160],{"id":16,"depth":227,"text":16},{"id":54,"depth":227,"text":54},{"id":92,"depth":227,"text":92},{"id":127,"depth":227,"text":128},{"id":195,"depth":227,"text":195},"平台","终端延迟（ping）是 MetaTrader 终端与经纪商交易服务器之间的往返时间。多少毫秒算好、多少算差，以及它怎样变成滑点。",{},"\u002Fzh\u002Fglossary\u002Fwhat-is-terminal-ping",[1166,1167,1168,1169],"what-is-packet-retransmission","what-is-slippage","what-is-a-vps","what-is-order-execution",{"title":946,"description":1162},"zh\u002F5.glossary\u002Fwhat-is-terminal-ping","终端延迟 (Terminal Ping)","Sq2425Hy2tRnONhjF0wJUWkFr2nMzONcmogzSLOF7A8",[1175,1176],{"path":859},{"path":241},1793898057949]