[{"data":1,"prerenderedAt":1208},["ShallowReactive",2],{"\u002Fglossary\u002Fwhat-is-the-sharpe-ratio":3,"\u002Fglossary\u002Fwhat-is-the-sharpe-ratio-related":255},{"id":4,"title":5,"body":6,"category":240,"date":241,"description":242,"draft":243,"extension":244,"meta":245,"navigation":243,"path":246,"related":247,"seo":252,"stem":253,"term":126,"updated":241,"__hash__":254},"glossary\u002F5.glossary\u002Fwhat-is-the-sharpe-ratio.md","What Is the Sharpe Ratio? Formula, Limits and Real Data",{"type":7,"value":8,"toc":232},"minimark",[9,13,18,21,32,35,41,50,53,57,60,68,76,80,99,115,118,122,162,165,188,192,225],[10,11,12],"p",{},"The Sharpe ratio is average return divided by the standard deviation of those returns: how much return an account produced per unit of volatility. Two accounts that finish a year at the same gain can carry very different Sharpe ratios, and the higher one got there on a smoother path.",[14,15,17],"h2",{"id":16},"how-it-works","How it works",[10,19,20],{},"The textbook definition uses periodic returns and subtracts a risk-free rate:",[22,23,28],"pre",{"className":24,"code":26,"language":27},[25],"language-text","Sharpe Ratio = (Mean return − Risk-free rate) \u002F Standard deviation of returns\n","text",[29,30,26],"code",{"__ignoreMap":31},"",[10,33,34],{},"ShowMyTrades computes it on closed trades rather than on calendar periods, and does not subtract a risk-free rate:",[22,36,39],{"className":37,"code":38,"language":27},[25],"Sharpe Ratio = Mean per-trade result \u002F Standard deviation of per-trade results\n",[29,40,38],{"__ignoreMap":31},[10,42,43,44,49],{},"Each trade enters the series as profit + commission + swap, so the inputs are already net of costs. Only closed buy and sell trades count: deposits, withdrawals and balance operations are filtered out before the series is built. The denominator is the population standard deviation of that same series — squared distances divided by the number of trades, not by that number minus one — which is why ",[45,46,48],"a",{"href":47},"\u002Fglossary\u002Fwhat-is-standard-deviation-in-trading","standard deviation"," sits immediately above the ratio in our statistics table: it is literally the divisor. An account with fewer than two closed trades reads 0.00.",[10,51,52],{},"Measuring per trade rather than per month means the figure is not annualised, so the usual conventions do not transfer. \"Above 1.0 is good, above 2.0 is excellent\" describes an annualised Sharpe on periodic returns. On a per-trade series, 1.00 would mean the average trade is worth a full standard deviation of trade outcomes — almost no account with a meaningful trade count reaches it.",[14,54,56],{"id":55},"why-it-matters","Why it matters",[10,58,59],{},"Gain alone says where an account ended. Sharpe says something about how it travelled. A strategy with a small, dependable edge repeated often will show a modest but positive ratio; a strategy carried by two lucky trades will show a large mean sitting on an even larger dispersion, and the ratio collapses.",[10,61,62,63,67],{},"It is also the honest counter to a chart that merely looks smooth, built from the same closed trades that produce ",[45,64,66],{"href":65},"\u002Fglossary\u002Fwhat-is-a-profit-factor","profit factor",".",[10,69,70,71,75],{},"What it cannot do is distinguish good volatility from bad. A trade that overshoots to the upside raises the denominator exactly as much as one that overshoots to the downside. That objection is the reason the ",[45,72,74],{"href":73},"\u002Fglossary\u002Fwhat-is-the-sortino-ratio","Sortino ratio"," exists.",[14,77,79],{"id":78},"what-the-data-shows","What the data shows",[10,81,82,83,87,88,91,92,67],{},"The figures below describe accounts published on ShowMyTrades, not traders in general. Across the published accounts that have trading history (August 2026), the ",[84,85,86],"strong",{},"median Sharpe ratio is 0.05",", on a ",[84,89,90],{},"median of 171 closed trades"," and a ",[84,93,94,98],{},[45,95,97],{"href":96},"\u002Fglossary\u002Fwhat-is-average-trade-length","median trade length"," of 2.4 hours",[10,100,101,102,106,107,110,111,114],{},"Read through our own formula, 0.05 says something concrete: on the median account the average trade is worth about one twentieth of a standard deviation of trade results. Roughly twenty parts noise to one part signal. That is not a failing set of accounts — the median ",[45,103,105],{"href":104},"\u002Fglossary\u002Fwhat-is-time-weighted-return","time-weighted return"," is ",[84,108,109],{},"+3.2%"," and ",[84,112,113],{},"63.0%"," of them are positive over time. It is what a real edge looks like at the level of a single trade.",[10,116,117],{},"It also explains why short samples prove nothing. At that signal-to-noise ratio, a run of twenty green trades is an ordinary outcome for an account with no edge at all, and a week of red says nothing about an account that has one.",[14,119,121],{"id":120},"where-you-see-it-on-showmytrades","Where you see it on ShowMyTrades",[10,123,124,127,128,131,132,136,137,140,141,144,145,144,148,144,151,110,154,157,158,67],{},[84,125,126],{},"Sharpe Ratio"," is the exact label, in the ",[84,129,130],{},"Advanced Statistics"," module of the account page, inside the ",[133,134,135],"em",{},"Performance Metrics"," column and carrying a tooltip. It sits directly under ",[84,138,139],{},"Standard Deviation"," — its own denominator — and beside ",[84,142,143],{},"Profit Factor",", ",[84,146,147],{},"Z-Score (Probability)",[84,149,150],{},"Expectancy",[84,152,153],{},"AHPR",[84,155,156],{},"GHPR",". It is calculated on our servers from the stored closed trades, printed to two decimals, and coloured green from 1.00 up, yellow from 0 to 1, red below zero. The module is shown by default on published account pages, though the owner can hide it, and it is not part of the ",[45,159,161],{"href":160},"\u002Fguides\u002Ftrading-widgets-and-embeds","embeddable widgets",[10,163,164],{},"One warning about our own page: the tooltip on that label still quotes the conventional bands, above 1.0 good and above 2.0 excellent. Those bands belong to an annualised Sharpe on periodic returns, not to the per-trade figure printed beside them.",[10,166,167,168,144,171,144,174,110,177,180,181,187],{},"The same module offers ",[84,169,170],{},"Table",[84,172,173],{},"Weekday",[84,175,176],{},"Hourly",[84,178,179],{},"Duration"," views, and ",[84,182,183],{},[45,184,186],{"href":185},"\u002Fguides\u002Fcustom-analysis-and-filter-presets","Custom Analysis"," recomputes the whole block on a filtered subset — Date Range, Symbols, Magic Numbers, Trade Type, Trading Hours, Lot Size and more. Recomputing Sharpe without an account's best symbol shows whether the ratio describes a system or an episode.",[14,189,191],{"id":190},"common-misunderstandings","Common misunderstandings",[193,194,195,202,213,219],"ul",{},[196,197,198,201],"li",{},[84,199,200],{},"\"Above 1.0 is good.\""," True of an annualised Sharpe on periodic returns. Ours is per trade, so the threshold does not apply.",[196,203,204,207,208,212],{},[84,205,206],{},"\"Sharpe measures risk.\""," It measures dispersion of closed trade results. It knows nothing about how deep the account fell — that is ",[45,209,211],{"href":210},"\u002Fglossary\u002Fwhat-is-drawdown","drawdown",", reported separately.",[196,214,215,218],{},[84,216,217],{},"\"A high Sharpe means a safe account.\""," Grid and averaging systems post tight, uniform closed trades precisely because they refuse to realise losers. Floating losses never enter the series.",[196,220,221,224],{},[84,222,223],{},"\"I can compare my Sharpe with any account.\""," Only at comparable trade frequency and holding period: a scalper closing 200 trades a month and a swing trader closing 8 are measured in different units.",[10,226,227,228,67],{},"For how Sharpe reads next to every other number on an account page, see the guide to ",[45,229,231],{"href":230},"\u002Fguides\u002Freading-a-trading-account-dashboard","reading a trading account dashboard",{"title":31,"searchDepth":233,"depth":233,"links":234},2,[235,236,237,238,239],{"id":16,"depth":233,"text":17},{"id":55,"depth":233,"text":56},{"id":78,"depth":233,"text":79},{"id":120,"depth":233,"text":121},{"id":190,"depth":233,"text":191},"Metrics","2026-08-19T00:00:00.000Z","The Sharpe ratio is average return divided by the volatility of those returns. Here is the formula, how we compute it per trade, and why our median is 0.05.",true,"md",{},"\u002Fglossary\u002Fwhat-is-the-sharpe-ratio",[248,249,250,251],"what-is-standard-deviation-in-trading","what-is-the-sortino-ratio","what-is-recovery-factor","what-is-a-profit-factor",{"title":5,"description":242},"5.glossary\u002Fwhat-is-the-sharpe-ratio","ba1NCAQnPXiXHA3mQUH2kyWz1dokcPnUnzXgxtvFftA",[256,537,767,1008],{"id":257,"title":258,"body":259,"category":240,"date":241,"description":526,"draft":527,"extension":244,"meta":528,"navigation":243,"path":65,"related":529,"seo":534,"stem":535,"term":143,"updated":241,"__hash__":536},"glossary\u002F5.glossary\u002Fwhat-is-a-profit-factor.md","What Is Profit Factor? Formula, Good Values, Real Data",{"type":7,"value":260,"toc":519},[261,264,266,272,275,278,284,287,355,357,360,368,376,378,392,399,404,411,422,424,449,475,477,512],[10,262,263],{},"Profit factor is gross profit divided by gross loss: the total won across all winning trades, divided by the total lost across all losing trades. A profit factor of 1.50 means the account made $1.50 for every $1.00 it gave back. Below 1.00 the account loses money by construction, regardless of how often it wins.",[14,265,17],{"id":16},[22,267,270],{"className":268,"code":269,"language":27},[25],"Profit Factor = Gross Profit \u002F |Gross Loss|\n",[29,271,269],{"__ignoreMap":31},[10,273,274],{},"Note what the formula ignores: how long the account traded, how much capital it used, and how deep it fell along the way.",[10,276,277],{},"It can also be written in terms of win rate and average trade size, which is where it becomes diagnostic:",[22,279,282],{"className":280,"code":281,"language":27},[25],"Profit Factor = (Win Rate × Avg Win) \u002F ((1 − Win Rate) × Avg Loss)\n",[29,283,281],{"__ignoreMap":31},[10,285,286],{},"Two very different systems can land on the same value. A 30% win rate with winners four times the size of losers gives 1.71. An 80% win rate with winners half the size of losers gives 2.00. Both work; they fail differently and feel completely different to trade.",[288,289,290,303],"table",{},[291,292,293],"thead",{},[294,295,296,300],"tr",{},[297,298,299],"th",{},"Profit factor",[297,301,302],{},"Reading",[304,305,306,315,323,331,339,347],"tbody",{},[294,307,308,312],{},[309,310,311],"td",{},"Below 1.00",[309,313,314],{},"Loses money — structurally, not marginally",[294,316,317,320],{},[309,318,319],{},"1.00 – 1.10",[309,321,322],{},"Inside the noise. One bad week erases it",[294,324,325,328],{},[309,326,327],{},"1.10 – 1.30",[309,329,330],{},"Thin but real edge, if the sample is large",[294,332,333,336],{},[309,334,335],{},"1.30 – 2.00",[309,337,338],{},"A working strategy",[294,340,341,344],{},[309,342,343],{},"2.00 – 3.00",[309,345,346],{},"Strong. Check trade count and cost accounting",[294,348,349,352],{},[309,350,351],{},"Above 3.00",[309,353,354],{},"Rare on long histories: usually a short sample, a few outsized winners, or open losers not yet realised",[14,356,56],{"id":55},[10,358,359],{},"Profit factor is the cleanest single answer to \"does this system make money\", and it is much harder to dress up than a win rate, because every loss enters the denominator at full size.",[10,361,362,363,367],{},"It is not, however, a risk measure. Profit factor knows nothing about sequence: an account that made its money in one month and bled for eleven shows the same value as one that ground upward every week. Read it next to ",[45,364,366],{"href":365},"\u002Fglossary\u002Fwhat-is-maximum-drawdown","maximum drawdown",", or you are measuring the destination without the trip.",[10,369,370,371,375],{},"The same caution applies to a profit factor quoted from a ",[45,372,374],{"href":373},"\u002Fglossary\u002Fwhat-is-backtesting","backtest",": optimisers maximise exactly this ratio, so a high backtested value often measures the fitting process rather than the strategy.",[14,377,79],{"id":78},[10,379,82,380,383,384,391],{},[84,381,382],{},"median profit factor is 1.28"," and the ",[84,385,386,390],{},[45,387,389],{"href":388},"\u002Fguides\u002Fwin-rate-is-not-an-edge","median win rate"," is 68.8%",". Both are calculated after costs here: swap and commission are folded into each trade before it is classed as a winner or a loser, so 1.28 is net, not gross.",[10,393,394],{},[395,396],"img",{"alt":397,"src":398},"Medians across public ShowMyTrades accounts: a 68.8% win rate beside a 1.28 profit factor","\u002Fimages\u002Farticles\u002Fwin-rate-vs-payoff.svg",[10,400,401],{},[133,402,403],{},"Two medians from the same set: the win rate looks strong, the edge behind it is thin.",[10,405,406,407,410],{},"That pairing is the point. Nearly seven trades in ten close green on the median account, and the whole edge still amounts to $1.28 earned for every $1.00 lost. Take an account sitting on both medians and invert the formula: its average winner is worth roughly ",[84,408,409],{},"0.58×"," its average loser. Winning often and earning well are separate achievements, and the first is far easier to manufacture — hold losers, cut winners, and the win rate climbs while the profit factor falls.",[10,412,413,414,417,418,421],{},"The median Sharpe ratio on the same set is ",[84,415,416],{},"0.05",", and the median account has ",[84,419,420],{},"171 closed trades",". At that sample size a profit factor above 3.00 is not evidence of a superior system; it is evidence that the sample is too small to have met its worst trade yet.",[14,423,121],{"id":120},[10,425,426,428,429,431,432,144,435,144,438,144,441,144,443,110,446,448],{},[84,427,143],{}," sits in the ",[84,430,130],{}," module on every published account page, alongside the numbers that explain it: ",[84,433,434],{},"Win Rate",[84,436,437],{},"Avg. Win",[84,439,440],{},"Avg. Loss",[84,442,150],{},[84,444,445],{},"Total Trades",[84,447,126],{},". The six together say what the single ratio cannot.",[10,450,451,452,455,456,110,459,462,463,110,466,469,470,474],{},"Two product details matter. First, the costs already inside the ratio are itemised separately: the trades table carries ",[84,453,454],{},"Profit (Gross)"," with ",[84,457,458],{},"Swap",[84,460,461],{},"Commission"," as their own columns, and Advanced Statistics totals ",[84,464,465],{},"Total Commissions",[84,467,468],{},"Total Swap Paid",", so you can see how much the net figure absorbed. Across published accounts those totals stand at $4,782,670 in commissions and $862,547 in swap. Second, ",[45,471,472],{"href":185},[84,473,186],{}," recomputes the whole statistics block on a filtered subset — by date range, symbol, magic number, direction or lot size — which is how you check whether a profit factor holds up outside its best quarter or without its best symbol.",[14,476,191],{"id":190},[193,478,479,489,500,506],{},[196,480,481,484,485,67],{},[84,482,483],{},"\"Profit factor above 1 means I am profitable.\""," Only if costs are inside the ratio. Where they are not, a high-frequency system at 1.05 gross can be flat or negative ",[45,486,488],{"href":487},"\u002Fglossary\u002Fwhat-is-cost-percentage","once commission and swap are applied",[196,490,491,494,495,499],{},[84,492,493],{},"\"A high profit factor means low risk.\""," It says nothing about drawdown, position size or sequence. A martingale ",[45,496,498],{"href":497},"\u002Fglossary\u002Fwhat-is-position-sizing","position-sizing"," progression can post 4.00 right up to the day it does not.",[196,501,502,505],{},[84,503,504],{},"\"It is comparable across timeframes.\""," It is not annualised. A scalper's 1.20 over 20,000 trades and a swing trader's 1.20 over 60 are not the same statement.",[196,507,508,511],{},[84,509,510],{},"\"Open trades do not affect it.\""," They do, by absence. Floating losses left open are excluded from gross loss entirely, which inflates the ratio until the position is closed.",[10,513,514,515,67],{},"For which metrics to track over time and in what order, see ",[45,516,518],{"href":517},"\u002Fguides\u002Ftracking-trading-performance","the guide to tracking trading performance",{"title":31,"searchDepth":233,"depth":233,"links":520},[521,522,523,524,525],{"id":16,"depth":233,"text":17},{"id":55,"depth":233,"text":56},{"id":78,"depth":233,"text":79},{"id":120,"depth":233,"text":121},{"id":190,"depth":233,"text":191},"Profit factor is gross profit divided by gross loss. Below 1.0 an account loses by construction. Here are the bands and the 1.28 median across thousands of accounts.",false,{},[530,531,532,533],"what-is-drawdown","what-is-maximum-drawdown","what-is-position-sizing","what-is-backtesting",{"title":258,"description":526},"5.glossary\u002Fwhat-is-a-profit-factor","0loLRuqRU0mx9nEOfjdEFW6U1Fb_JAC1jN2JQSkIjuE",{"id":538,"title":539,"body":540,"category":240,"date":241,"description":758,"draft":243,"extension":244,"meta":759,"navigation":243,"path":760,"related":761,"seo":763,"stem":764,"term":765,"updated":241,"__hash__":766},"glossary\u002F5.glossary\u002Fwhat-is-recovery-factor.md","What Is Recovery Factor? Formula, Bands and Real Data",{"type":7,"value":541,"toc":751},[542,545,547,553,556,562,565,610,618,620,628,638,640,651,658,668,670,676,697,704,714,716,745],[10,543,544],{},"Recovery factor is net profit divided by maximum drawdown: how much return an account produced for each unit of decline it put its owner through. A value of 2.0 means the account earned twice what it gave up at its worst point. Below 1.0 it has not yet earned back the size of the hole it dug.",[14,546,17],{"id":16},[22,548,551],{"className":549,"code":550,"language":27},[25],"Recovery Factor = Net Profit \u002F Maximum Drawdown\n",[29,552,550],{"__ignoreMap":31},[10,554,555],{},"The formula is trivial; the units are not. Two versions circulate and they do not agree:",[22,557,560],{"className":558,"code":559,"language":27},[25],"Currency form    = Net profit ($) \u002F Maximum drawdown ($)\nPercentage form  = Total return (%) \u002F Maximum drawdown (%)\n",[29,561,559],{"__ignoreMap":31},[10,563,564],{},"Both sides must be measured the same way. Mixing a currency numerator with a percentage denominator produces a figure that looks like a ratio and means nothing.",[288,566,567,576],{},[291,568,569],{},[294,570,571,574],{},[297,572,573],{},"Recovery factor",[297,575,302],{},[304,577,578,586,594,602],{},[294,579,580,583],{},[309,581,582],{},"Below 1.0",[309,584,585],{},"The account has not out-earned its worst decline",[294,587,588,591],{},[309,589,590],{},"1.0 – 2.0",[309,592,593],{},"Working, but the reward is close to the risk absorbed",[294,595,596,599],{},[309,597,598],{},"2.0 – 3.0",[309,600,601],{},"Solid, if the history is long enough to have found the real maximum",[294,603,604,607],{},[309,605,606],{},"Above 3.0",[309,608,609],{},"Strong — check trade count and account age before believing it",[10,611,612,613,617],{},"One caveat outranks the whole table. Recovery factor rises with time on its own: the numerator accumulates as long as the account keeps earning, while the denominator is a historical maximum that only moves when things get worse. ",[45,614,616],{"href":615},"\u002Fguides\u002Fhow-much-history-a-track-record-needs","A five-year account and a six-month account are not comparable here"," — the older one has simply had longer to divide by a fixed number.",[14,619,56],{"id":55},[10,621,622,623,627],{},"Profit and drawdown are usually quoted apart, and separately both are easy to present well. Recovery factor refuses to let them be separate. It is the compact form of the question every ",[45,624,626],{"href":625},"\u002Fglossary\u002Fwhat-is-a-track-record","track record"," has to answer: was the pain worth it?",[10,629,630,631,633,634,637],{},"It also fills the gaps left by its neighbours. ",[45,632,299],{"href":65}," compares gross wins to gross losses and ignores sequence, so an account that made everything in one month scores like one that ground upward every week. The ",[45,635,636],{"href":246},"Sharpe ratio"," measures dispersion but not depth. Recovery factor is the only common metric where the worst moment appears explicitly in the denominator.",[14,639,79],{"id":78},[10,641,82,642,383,648,67],{},[84,643,644,645,647],{},"median ",[45,646,105],{"href":104}," is +3.2%",[84,649,650],{},"median deepest drawdown is 9.7%",[10,652,653,654,657],{},"Those medians come from different accounts, so they do not give a population median for the ratio. But an account sitting on both would carry a recovery factor of about ",[84,655,656],{},"0.33"," in the percentage form: roughly one unit of return for every three units of decline endured to get it — worth holding next to the convention that anything under 3.0 is unremarkable.",[10,659,660,661,664,665,667],{},"The tail matters too. ",[84,662,663],{},"38.2%"," of these accounts have been down more than 20%, and ",[84,666,113],{}," are positive over time — so more than a third have a negative numerator. For those the ratio is not low, it is meaningless: dividing a loss by a drawdown produces a number no band interprets.",[14,669,121],{"id":120},[10,671,672,675],{},[84,673,674],{},"Recovery factor is not a metric on ShowMyTrades."," No panel, chart or widget displays it under that name or any other, and our statistics engine does not calculate one. You compute it in one step from figures the account page does publish, all derived from the platform feed rather than typed in by the owner.",[10,677,678,679,682,683,686,687,690,691,693,694,696],{},"For the percentage form, take ",[84,680,681],{},"Gain"," from the ",[84,684,685],{},"Account Stats"," panel and divide it by ",[84,688,689],{},"Drawdown",", four rows below it in the same panel. Both are percentages, so the division is valid. Note what each is: ",[84,692,681],{}," is time-weighted return, which strips out deposits and withdrawals, so it cannot be inflated by adding capital to a losing account; ",[84,695,689],{}," is the worst equity decline recorded.",[10,698,699,700,703],{},"For the currency form, ",[84,701,702],{},"Profit"," in the same panel gives net closed profit and loss, after commission and swap, but we do not publish maximum drawdown as a currency figure, so there is no second number to divide by. Use the percentage form, or rebuild the equity path from the closed trade list, which the trades table offers as a CSV download.",[10,705,706,707,709,710,713],{},"Either way, open the ",[84,708,689],{}," chart view beside that panel first. It plots drawdown day by day as bars, showing ",[133,711,712],{},"when"," the maximum was set and how long the account spent below its peak — the duration the ratio cannot express.",[14,715,191],{"id":190},[193,717,718,724,730,739],{},[196,719,720,723],{},[84,721,722],{},"\"Higher is always better.\""," Not across different histories. The ratio grows with age at constant risk, so compare only comparable track record lengths.",[196,725,726,729],{},[84,727,728],{},"\"It is the same as profit factor.\""," Profit factor divides gross wins by gross losses. Recovery factor divides the final result by the single worst peak-to-trough decline.",[196,731,732,735,736,738],{},[84,733,734],{},"\"It shows how fast I recover.\""," It shows magnitude, not time — see ",[45,737,211],{"href":210}," for why duration removes more traders than depth.",[196,740,741,744],{},[84,742,743],{},"\"Use DD on Balance for a cleaner denominator.\""," That is the flattering denominator: strategies that never close losers keep balance drawdown small by construction. Prefer the equity figure, and say which you used.",[10,746,747,748,67],{},"For how the denominator is measured, and why the maximum never falls, see the guide to ",[45,749,366],{"href":750},"\u002Fguides\u002Fmaximum-drawdown-explained",{"title":31,"searchDepth":233,"depth":233,"links":752},[753,754,755,756,757],{"id":16,"depth":233,"text":17},{"id":55,"depth":233,"text":56},{"id":78,"depth":233,"text":79},{"id":120,"depth":233,"text":121},{"id":190,"depth":233,"text":191},"Recovery factor is net profit divided by maximum drawdown: return earned per unit of decline. Here is the formula, the bands, and where to find the inputs.",{},"\u002Fglossary\u002Fwhat-is-recovery-factor",[531,530,251,762],"what-is-the-sharpe-ratio",{"title":539,"description":758},"5.glossary\u002Fwhat-is-recovery-factor","Recovery Factor","PEuiM5pQPFw3EDZ1EoqyqYWRrmfWCLdpkwGvVfRHmqY",{"id":768,"title":769,"body":770,"category":240,"date":241,"description":1000,"draft":243,"extension":244,"meta":1001,"navigation":243,"path":47,"related":1002,"seo":1005,"stem":1006,"term":139,"updated":241,"__hash__":1007},"glossary\u002F5.glossary\u002Fwhat-is-standard-deviation-in-trading.md","What Is Standard Deviation in Trading? Formula and Data",{"type":7,"value":771,"toc":993},[772,775,777,783,801,837,840,842,845,856,859,886,888,895,898,906,908,928,954,956,987],[10,773,774],{},"Standard deviation is the typical distance between a single trade's result and the account's average trade result. On ShowMyTrades it is reported in the account currency, so it answers a direct question: how much does one trade usually differ from the ordinary trade on this account? A small figure means results cluster; a large one means the history is made of outliers.",[14,776,17],{"id":16},[22,778,781],{"className":779,"code":780,"language":27},[25],"Standard Deviation = √( Σ (rᵢ − mean)² \u002F n )\n",[29,782,780],{"__ignoreMap":31},[10,784,785,786,789,790,144,793,796,797,800],{},"Where ",[29,787,788],{},"rᵢ"," is the net result of closed trade ",[133,791,792],{},"i",[29,794,795],{},"mean"," is the average of those results, and ",[29,798,799],{},"n"," is the number of closed trades. Four details decide what the number means:",[193,802,803,812,818,831],{},[196,804,805,808,809,811],{},[84,806,807],{},"Each trade enters net of costs."," ",[29,810,788],{}," is profit + commission + swap, so commission-heavy and swap-heavy trades widen the dispersion exactly as much as they narrow the profit.",[196,813,814,817],{},[84,815,816],{},"Only closed buy and sell trades count."," Deposits and withdrawals are filtered out before the series is built, so funding the account is not a giant outlier.",[196,819,820,823,824,826,827,830],{},[84,821,822],{},"It is the population form",", dividing by ",[29,825,799],{}," rather than ",[29,828,829],{},"n − 1",". With fewer than two closed trades the metric reads 0.00.",[196,832,833,836],{},[84,834,835],{},"It is in currency, not percent."," A $50 standard deviation on a $2,000 account and on a $200,000 account describe completely different behaviour. Always read it against the account balance.",[10,838,839],{},"Because distances are squared before averaging, one exceptional trade moves the figure far more than a dozen ordinary ones — which is what makes it catch the trade that does not belong.",[14,841,56],{"id":55},[10,843,844],{},"Standard deviation is the honest test of the word \"consistent\": anyone can call an equity curve steady, and this puts a currency value on how steady the trades were.",[10,846,847,848,850,851,855],{},"It is also the denominator of our ",[45,849,636],{"href":246},": Sharpe is the average trade expressed in units of this figure. And it is the natural companion to ",[45,852,854],{"href":853},"\u002Fglossary\u002Fwhat-is-expectancy","expectancy",", which gives the average outcome per trade with no indication of how reliably that average is delivered.",[10,857,858],{},"The most useful reading is comparative, against the two averages sitting a few rows away:",[193,860,861,867,880],{},[196,862,863,866],{},[84,864,865],{},"Close to Avg. Win"," — results are tight. The account behaves the same way most of the time.",[196,868,869,872,873,110,876,879],{},[84,870,871],{},"Several times Avg. Win"," — a handful of trades carries the history. Check ",[84,874,875],{},"Best Trade (P&L)",[84,877,878],{},"Worst Trade (P&L)"," before trusting the totals.",[196,881,882,885],{},[84,883,884],{},"Much larger than Avg. Loss"," — at least one loss sits far outside the routine.",[14,887,79],{"id":78},[10,889,82,890,892,893,67],{},[84,891,86],{}," on a ",[84,894,90],{},[10,896,897],{},"Run our own definition backwards and that pairing is precise about dispersion: since Sharpe is the mean trade divided by this standard deviation, a median of 0.05 means the average trade on the middle account is worth roughly one twentieth of the standard deviation of its trades. Routine variation between one trade and the next is about twenty times larger than the edge extracted from them.",[10,899,900,901,905],{},"On a ",[84,902,903,98],{},[45,904,97],{"href":96},", that is why a full day of results teaches a trader nothing: a day sits well inside the dispersion. It is also why a low figure on a young account is not evidence of control — the sample may simply not contain the outlier the strategy can produce, and squared distances mean that when it arrives, it arrives loudly.",[14,907,121],{"id":120},[10,909,910,127,912,131,914,916,917,919,920,922,923,927],{},[84,911,139],{},[84,913,130],{},[133,915,135],{}," column, formatted in the account currency and carrying a tooltip. It sits directly above ",[84,918,126],{},", the ratio it divides, below ",[84,921,143],{},", and is calculated on our servers from the stored closed trades. The module is shown by default on published account pages, though ",[45,924,926],{"href":925},"\u002Fguides\u002Fpublic-profile-and-custom-slugs","the owner can hide it",", and it is not part of the embeddable widgets.",[10,929,930,931,934,935,144,937,144,939,144,941,110,943,945,946,948,949,953],{},"Read it across to the ",[133,932,933],{},"Trades"," column of the same module, where ",[84,936,437],{},[84,938,440],{},[84,940,875],{},[84,942,878],{},[84,944,445],{}," give the figure its context. The module's ",[84,947,179],{}," view makes the dispersion visual: every closed trade is one point, holding time in hours across the bottom and profit or loss up the side, so the spread becomes a shape rather than a number. ",[84,950,951],{},[45,952,186],{"href":185}," recomputes the whole block on a filtered subset — Date Range, Symbols, Magic Numbers, Trade Type, Lot Size and more — which shows whether the dispersion belongs to the strategy or to one instrument.",[14,955,191],{"id":190},[193,957,958,964,975,981],{},[196,959,960,963],{},[84,961,962],{},"\"It is a percentage.\""," Ours is an amount of money per trade. Comparing it between accounts of different size is meaningless without the balance beside it.",[196,965,966,969,970,974],{},[84,967,968],{},"\"It measures market volatility.\""," It measures the ",[45,971,973],{"href":972},"\u002Fglossary\u002Fwhat-is-volatility","volatility"," of your results. A calm market traded with escalating size produces a large figure.",[196,976,977,980],{},[84,978,979],{},"\"Low standard deviation means low risk.\""," Only closed trades enter it. Systems that hold losers open post uniformly small closed results while the real exposure sits in floating positions.",[196,982,983,986],{},[84,984,985],{},"\"A low figure on 30 trades is reassuring.\""," Below a few hundred trades, treat it as provisional.",[10,988,989,990,67],{},"For how many trades a set of statistics needs before it means anything, see the guide on ",[45,991,992],{"href":615},"how much history a track record needs",{"title":31,"searchDepth":233,"depth":233,"links":994},[995,996,997,998,999],{"id":16,"depth":233,"text":17},{"id":55,"depth":233,"text":56},{"id":78,"depth":233,"text":79},{"id":120,"depth":233,"text":121},{"id":190,"depth":233,"text":191},"Standard deviation measures how far a single trade lands from your average trade. Here is the formula we use, net of costs, and how to read it on thousands of accounts.",{},[762,1003,1004,249],"what-is-volatility","what-is-expectancy",{"title":769,"description":1000},"5.glossary\u002Fwhat-is-standard-deviation-in-trading","erd9RmXvwGmqWy07DwcKxdN6CBwW8ySYP01l-ftXMHQ",{"id":1009,"title":1010,"body":1011,"category":240,"date":241,"description":1201,"draft":243,"extension":244,"meta":1202,"navigation":243,"path":73,"related":1203,"seo":1204,"stem":1205,"term":1206,"updated":241,"__hash__":1207},"glossary\u002F5.glossary\u002Fwhat-is-the-sortino-ratio.md","What Is the Sortino Ratio? Downside Risk Explained",{"type":7,"value":1012,"toc":1194},[1013,1016,1018,1024,1030,1035,1038,1052,1055,1057,1069,1072,1078,1080,1096,1099,1115,1117,1148,1161,1163,1189],[10,1014,1015],{},"The Sortino ratio is the Sharpe ratio with a narrower denominator: instead of dividing return by the volatility of all results, it divides by the volatility of the results that fell below a chosen target. Upside swings do not count as risk. It answers the objection that a trader punished for large winners is being punished for the wrong thing.",[14,1017,17],{"id":16},[22,1019,1022],{"className":1020,"code":1021,"language":27},[25],"Sortino Ratio = (Mean return − Target return) \u002F Downside deviation\n",[29,1023,1021],{"__ignoreMap":31},[22,1025,1028],{"className":1026,"code":1027,"language":27},[25],"Downside deviation = √( Σ min(rᵢ − Target, 0)² \u002F n )\n",[29,1029,1027],{"__ignoreMap":31},[10,1031,1032,1033,67],{},"Every observation above the target contributes zero to the denominator. Everything below contributes its squared shortfall, exactly as in ",[45,1034,48],{"href":47},[10,1036,1037],{},"Two choices sit inside that formula, and both change the number:",[193,1039,1040,1046],{},[196,1041,1042,1045],{},[84,1043,1044],{},"The target."," Zero, the risk-free rate, or a required return. A Sortino computed against 0% and one computed against 4% a year are different statistics wearing the same name.",[196,1047,1048,1051],{},[84,1049,1050],{},"The divisor."," Some implementations divide by the total count of observations, others only by the count of below-target ones. The second always produces a larger, friendlier ratio.",[10,1053,1054],{},"Because neither choice is standardised, a Sortino ratio quoted without its target and its convention is not a comparable figure. That is a practical problem, not a pedantic one: it is the reason two platforms can report different Sortino ratios for the same account and both be correct.",[14,1056,56],{"id":55},[10,1058,1059,1060,1064,1065,1068],{},"The idea behind Sortino is sound. Investors do not experience a 5% up-month as risk, and a measure that treats it as such misprices strategies with positive skew — ",[45,1061,1063],{"href":1062},"\u002Fglossary\u002Fwhat-is-trend-following","trend following",", for instance, where a few very large winners carry the whole result and inflate the ",[45,1066,1067],{"href":246},"Sharpe"," denominator.",[10,1070,1071],{},"But treating upside volatility as free is a modelling choice, not a law, and it cuts the wrong way for a whole family of trading systems. Grid, martingale and averaging-down strategies produce their up-moves and their down-moves from the same mechanism: adding size against the position. Score them with Sortino and the ordinary small profits stay in the numerator while most of the dispersion drops out of the denominator, until the day the mechanism fails. Sortino flatters exactly the strategies whose risk is hardest to see in closed trades.",[10,1073,1074,1075,1077],{},"The safe reading is that Sortino is a useful second opinion on a return series, never a verdict, and never a substitute for ",[45,1076,366],{"href":365},", which measures the thing an investor actually lives through.",[14,1079,79],{"id":78},[10,1081,82,1082,1084,1085,1088,1089,1091,1092,1095],{},[84,1083,650],{},", but the distribution is where the argument sits: ",[84,1086,1087],{},"38.5%"," never fell more than 5%, while ",[84,1090,663],{}," have been down more than 20% and ",[84,1093,1094],{},"17.6%"," have lost more than half their peak value.",[10,1097,1098],{},"Two nearly equal groups, at opposite ends of the range. That shape is why downside-only statistics are attractive and why they are dangerous on the same data: the accounts in the deep tail did not get there through symmetric wobble, they got there through one direction. Any measure that discards half the distribution has to be right about which half.",[10,1100,1101,1102,1105,1106,1109,1110,1114],{},"One more asymmetry is already built into the raw numbers. Published accounts have paid ",[84,1103,1104],{},"$862,547"," in swap, and ",[84,1107,1108],{},"86.3%"," of the accounts that carry swap at all ",[45,1111,1113],{"href":1112},"\u002Fguides\u002Fswap-and-carry-trade","pay net negative swap",". The drift on the average holding cost points down, before any strategy makes a decision.",[14,1116,121],{"id":120},[10,1118,1119,1120,1123,1124,1126,1127,110,1129,1131,1132,1134,1135,1137,1138,1140,1141,1147],{},"Nowhere, and it is worth being exact about that. ",[84,1121,1122],{},"No panel, chart, widget or export on ShowMyTrades displays a Sortino ratio",", and our statistics engine does not calculate one either — there is no stored value waiting behind the interface. The ",[84,1125,130],{}," module publishes ",[84,1128,126],{},[84,1130,139],{}," in its ",[133,1133,135],{}," column; the ",[84,1136,685],{}," panel publishes ",[84,1139,689],{},", the worst equity drawdown, and ",[84,1142,1143],{},[45,1144,1146],{"href":1145},"\u002Fglossary\u002Fwhat-is-drawdown-on-balance","DD on Balance",", the worst drawdown measured on closed results alone. Those are the downside figures our pages carry today.",[10,1149,1150,1151,144,1153,144,1155,110,1157,1160],{},"If you want a Sortino on a published account, you can build one. The trades table on an account page has a CSV download of the closed trade list, one row per trade with ",[84,1152,702],{},[84,1154,458],{},[84,1156,461],{},[84,1158,1159],{},"Net profit"," columns — that last column is the same series our Sharpe ratio is computed on. Pick your target, state it, and the number is reproducible by anyone else with the same file. That is a better outcome than an unlabelled ratio on a page, and it is the standard we would have to meet before publishing one.",[14,1162,191],{"id":190},[193,1164,1165,1171,1177,1183],{},[196,1166,1167,1170],{},[84,1168,1169],{},"\"Sortino is a better Sharpe.\""," It is a different question, not a strictly better answer. It measures shortfall against a target; Sharpe measures dispersion. Both ignore drawdown depth.",[196,1172,1173,1176],{},[84,1174,1175],{},"\"A high Sortino means small losses.\""," It means few or small results below the target. A single catastrophic trade that has not happened yet is absent from both ratios.",[196,1178,1179,1182],{},[84,1180,1181],{},"\"Sortino ratios are comparable across sites.\""," Only when the target and the divisor convention match, and they usually are not published.",[196,1184,1185,1188],{},[84,1186,1187],{},"\"ShowMyTrades hides its Sortino.\""," There is nothing to hide: we do not compute one. The closed-trade export lets you compute your own.",[10,1190,1191,1192,67],{},"For the measure we do publish on the downside, and how to read it, see the guide to ",[45,1193,366],{"href":750},{"title":31,"searchDepth":233,"depth":233,"links":1195},[1196,1197,1198,1199,1200],{"id":16,"depth":233,"text":17},{"id":55,"depth":233,"text":56},{"id":78,"depth":233,"text":79},{"id":120,"depth":233,"text":121},{"id":190,"depth":233,"text":191},"The Sortino ratio replaces total volatility with downside deviation. Here is the formula, the modelling choice inside it, and why we do not publish one.",{},[762,248,531,250],{"title":1010,"description":1201},"5.glossary\u002Fwhat-is-the-sortino-ratio","Sortino Ratio","McckUz-i7wK1j0jeiRBNXe71AOkmVU22WIMkpahZ1N4",1787415693480]