[{"data":1,"prerenderedAt":1939},["ShallowReactive",2],{"article-ai-stock-analysis-tools-compared-what-each-one-reads":3,"related-blogs-ai-stock-analysis-tools-compared-what-each-one-reads":693},{"id":4,"title":5,"author":6,"body":7,"coverImg":679,"description":680,"extension":681,"head":6,"meta":682,"navigation":683,"path":684,"publishedAt":685,"schemaOrg":6,"seo":686,"sitemap":687,"stem":691,"updatedAt":685,"__hash__":692},"blog\u002Fblog\u002F2026\u002F10\u002Fai-stock-analysis-tools-compared-what-each-one-reads.md","AI Stock Analysis Tools Compared: What Each One Reads",null,{"type":8,"value":9,"toc":660},"minimark",[10,14,17,20,25,28,117,120,124,127,266,269,273,276,288,295,298,348,351,358,362,371,374,377,383,389,393,396,399,404,407,411,414,417,422,427,431,434,437,445,450,455,459,462,465,470,475,479,482,485,490,493,497,500,503,517,521,524,532,536,544,547,566,569,573,582,591,597,601,604,637,641,644,647,650,654,657],[11,12,13],"p",{},"Seven tabs open. Seven AI stock analysis tools, and every one of them says its AI is the clever one. So which is best?",[11,15,16],{},"Wrong question, it turns out. These tools read completely different things, from hundreds of indicators across the whole market to a single photo of a chart. They hand back completely different things too, from a score of 1 to 10 to a paragraph in plain English. Pick by the question you bring and the choice gets easy.",[11,18,19],{},"Below is the verdict first, then what each tool reads, what it hands back and the catch in each. One of the seven, Primo Trade, is our own product. It gets the same treatment, gaps included.",[21,22,24],"h2",{"id":23},"the-best-ai-for-stock-analysis-depends-on-your-question","The Best AI for Stock Analysis Depends on Your Question",[11,26,27],{},"Find your question in the left column. That is most of the decision.",[29,30,31,47],"table",{},[32,33,34],"thead",{},[35,36,37,41,44],"tr",{},[38,39,40],"th",{},"Your question",[38,42,43],{},"Start with",[38,45,46],{},"Because it",[48,49,50,62,73,84,95,106],"tbody",{},[35,51,52,56,59],{},[53,54,55],"td",{},"\"Which of thousands of stocks should I even look at?\"",[53,57,58],{},"Danelfin or Kavout",[53,60,61],{},"Ranks the whole market by a score",[35,63,64,67,70],{},[53,65,66],{},"\"Where is this chart pattern showing up?\"",[53,68,69],{},"Tickeron or TrendSpider",[53,71,72],{},"Searches charts across many symbols",[35,74,75,78,81],{},[53,76,77],{},"\"What is moving in the next hour?\"",[53,79,80],{},"Trade Ideas",[53,82,83],{},"Scans live during the session",[35,85,86,89,92],{},[53,87,88],{},"\"Can I get a signal feed on my phone?\"",[53,90,91],{},"Prospero.ai",[53,93,94],{},"Is a phone app built around signals",[35,96,97,100,103],{},[53,98,99],{},"\"What is going on with this one stock or chart?\"",[53,101,102],{},"Primo Trade",[53,104,105],{},"Reads one ticker or one chart photo and explains it",[35,107,108,111,114],{},[53,109,110],{},"\"Can I just ask anything, my way?\"",[53,112,113],{},"A general chatbot",[53,115,116],{},"Takes any question, but you design it and check it",[11,118,119],{},"Plenty of people use two of these: a ranker or scanner to find candidates, then a single-stock read to look closer before deciding for themselves.",[21,121,123],{"id":122},"ai-stock-analysis-tools-at-a-glance","AI Stock Analysis Tools at a Glance",[11,125,126],{},"Every detail below comes from each tool's own pages, checked from the United States on October 8, 2026, so treat it as dated to that day. Prices are left out; they change too often.",[29,128,129,148],{},[32,130,131],{},[35,132,133,136,139,142,145],{},[38,134,135],{},"Tool",[38,137,138],{},"What the AI reads",[38,140,141],{},"What you get back",[38,143,144],{},"Horizon",[38,146,147],{},"The catch",[48,149,150,167,184,201,218,234,250],{},[35,151,152,155,158,161,164],{},[53,153,154],{},"Danelfin",[53,156,157],{},"900+ technical, fundamental and sentiment indicators",[53,159,160],{},"AI Score 1 to 10, four subscores",[53,162,163],{},"3 months",[53,165,166],{},"Probabilities, not certainties, in its own words",[35,168,169,172,175,178,181],{},[53,170,171],{},"Kavout",[53,173,174],{},"Fundamental, technical and sentiment data on 9,000+ US stocks",[53,176,177],{},"Kai Score, Stock Rank, Technical Rating, strategy lists",[53,179,180],{},"Not stated (rankings refresh daily)",[53,182,183],{},"The picker sits behind the Pro tier",[35,185,186,189,192,195,198],{},[53,187,188],{},"TrendSpider",[53,190,191],{},"Your charts, market data, fundamentals, filings, news",[53,193,194],{},"Pattern recognition, an AI assistant, backtests, bots",[53,196,197],{},"Whatever you set",[53,199,200],{},"A lot of platform for one question",[35,202,203,206,209,212,215],{},[53,204,205],{},"Tickeron",[53,207,208],{},"Price patterns and trends",[53,210,211],{},"Pattern matches with confidence levels, signals, robots",[53,213,214],{},"Intraday to swing",[53,216,217],{},"Confidence describes past setups",[35,219,220,222,225,228,231],{},[53,221,80],{},[53,223,224],{},"Live market data",[53,226,227],{},"Holly buy and sell signals, scans",[53,229,230],{},"Intraday",[53,232,233],{},"The homepage examples are a highlight reel",[35,235,236,238,241,244,247],{},[53,237,91],{},[53,239,240],{},"Institutional data and market movements",[53,242,243],{},"Key signals, cumulative scores",[53,245,246],{},"Short and long term",[53,248,249],{},"Its fine print says the results are simulated",[35,251,252,254,257,260,263],{},[53,253,102],{},[53,255,256],{},"One ticker's prices, news and financials, or one chart photo",[53,258,259],{},"Signal, Primo Score, strategy idea, Key Insights",[53,261,262],{},"6 to 12 months for the Forecast Value",[53,264,265],{},"One stock at a time, no scans or alerts",[11,267,268],{},"Primo Trade gives AI-generated analysis for information and education. It is not financial advice. Markets carry risk; do your own research before you trade.",[21,270,272],{"id":271},"one-monday-three-honest-answers","One Monday, Three Honest Answers",[11,274,275],{},"Before the tools, one dated example of why the horizon column matters more than any brand name.",[11,277,278,279,287],{},"Monday, January 27, 2025. Nvidia closed at $118.42, down from $142.62 the Friday before, a drop of about 17% in one session, according to Yahoo Finance's daily ",[280,281,286],"a",{"href":282,"target":283,"rel":284},"https:\u002F\u002Ffinance.yahoo.com\u002Fquote\u002FNVDA\u002Fhistory\u002F?period1=1737676800&period2=1754006400","_blank",[285],"noopener","price history",". Reuters tied the sell-off to a low-cost AI model from the Chinese startup DeepSeek and called it a record one-day loss in market value for a Wall Street stock.",[289,290],"img",{"src":291,"alt":292,"width":293,"height":294},"https:\u002F\u002Fmedia.primotrade.app\u002Fblog\u002Fai-stock-analysis-tools-compared-what-each-one-reads\u002Flong-red-candle-macro-panel.jpg","Stylised chart art of one long red candlestick in sharp focus among smaller green and amber candles.",1067,600,[11,296,297],{},"Here is what followed, read long after the fact from the same Yahoo Finance data:",[29,299,300,313],{},[32,301,302],{},[35,303,304,307,310],{},[38,305,306],{},"Checkpoint after the January 27 close",[38,308,309],{},"Close",[38,311,312],{},"Versus January 27",[48,314,315,326,337],{},[35,316,317,320,323],{},[53,318,319],{},"Next day, January 28, 2025",[53,321,322],{},"$128.99",[53,324,325],{},"+8.9%",[35,327,328,331,334],{},[53,329,330],{},"Three months, April 28, 2025",[53,332,333],{},"$108.73",[53,335,336],{},"-8.2%",[35,338,339,342,345],{},[53,340,341],{},"Six months, July 28, 2025",[53,343,344],{},"$176.75",[53,346,347],{},"+49.3%",[11,349,350],{},"A same-day signal, a three-month score and a 6 to 12 month estimate were answering three different questions that Monday. Each could look wrong at one checkpoint and fine at the next. Nothing on the chart that day said which path came next, and no model could have known it either.",[11,352,353,357],{},[354,355,356],"strong",{},"So the first thing to match is not the AI but the horizon."," A tool built for the next hour and a tool built for the next quarter can both be honest and still disagree.",[21,359,361],{"id":360},"danelfin-one-number-per-stock","Danelfin: One Number per Stock",[11,363,364,365,370],{},"Danelfin boils each stock down to a single AI Score from 1 to 10. According to its ",[280,366,369],{"href":367,"target":283,"rel":368},"https:\u002F\u002Fdanelfin.com\u002Fhow-it-works",[285],"own explanation",", a higher score means a higher probability that the stock outperforms the market over the next three months. For US stocks and ETFs the benchmark is the S&P 500 total return index; for European stocks it is the STOXX Europe 600.",[11,372,373],{},"The inputs run long: 10,000 features per stock per day, built from more than 600 technical, 150 fundamental and 150 sentiment indicators. It covers US stocks, Europe's main markets and US-listed ETFs. Beside the headline score sit Fundamental, Technical, Sentiment and Low Risk subscores, and an explanation window shows which features pushed the score up (green) or down (red).",[11,375,376],{},"That window is the good bit. You can see that a high score leans on momentum rather than earnings, and decide whether you believe the reason. Danelfin also publishes how each score level has done historically, with the universe, date range and sample size printed beside the chart: more homework than most tools show.",[11,378,379,382],{},[354,380,381],{},"The catch:"," Danelfin says it plainly. Its AI \"calculates probabilities, not certainties,\" and the scores are \"useful but not infallible information.\" Some stock pages also show an AI forecast for set time frames, so check which horizon a number means. It runs on the web and in free-to-download iOS and Android apps with paid plans inside.",[11,384,385,388],{},[354,386,387],{},"Best for:"," sorting a big universe of stocks by one comparable number, with the reasons on show.",[21,390,392],{"id":391},"kavout-rankings-by-strategy","Kavout: Rankings by Strategy",[11,394,395],{},"Kavout's AI Stock Picker ranks more than 9,000 US stocks every day with three measures: the Kai Score, Stock Rank and a Technical Rating. Kavout says they combine fundamental, technical and sentiment data, and the rankings refresh daily.",[11,397,398],{},"What you browse is a menu of strategies rather than one score: Greenblatt Magic Formula, Lynch Growth Momentum, Dual Analysis and more, each with its own top list for large caps, small caps, the S&P 500 and so on. Around the picker sit tools named InvestGPT, Financial Agents, Smart Signals and Market Movers.",[11,400,401,403],{},[354,402,381],{}," a top-ten list tells you which stocks a model prefers under its rules today. It does not tell you why one of them suits you, so the research on each name starts after the list. The picker itself sits behind the Pro tier, and Kavout's disclaimer says nothing on the site \"constitutes investment, financial, trading, or any other form of advice.\"",[11,405,406],{},"If you like screening by a named strategy and want a ranked shortlist to dig into, this is your tool.",[21,408,410],{"id":409},"trendspider-the-whole-cockpit","TrendSpider: The Whole Cockpit",[11,412,413],{},"TrendSpider is the multi-tool of the group. It is a full charting and research platform with automated pattern recognition, scanners, a no-code strategy tester that runs over decades of market data, and an ML Quant Lab that trains a machine learning model on inputs you choose.",[11,415,416],{},"The newest layer is Sidekick, its AI assistant. TrendSpider says Sidekick can see your charts, pull market data, read SEC filings, insider transactions, news and earnings call transcripts, and work as an agent: building scanners, creating alerts across a watch list and researching several symbols from one prompt. You choose which large language model answers, from Google, Anthropic or OpenAI, and even a personality, from long-term investor to number-crunching quant.",[11,418,419,421],{},[354,420,381],{}," breadth. TrendSpider rewards people who already chart every day and know what they want to test. For one quick read on one stock, it is a lot of cockpit. Its no-code bots can also place trades for you, a bigger commitment than reading a chart. Its own footer says nothing in the platform \"should be construed as investment advice of any kind.\"",[11,423,424,426],{},[354,425,387],{}," active traders who want charting, scanning, backtesting and an AI assistant under one roof.",[21,428,430],{"id":429},"tickeron-pattern-hunting-at-scale","Tickeron: Pattern Hunting at Scale",[11,432,433],{},"Tickeron builds its AI around chart shapes. Its AI Pattern Search Engine finds end-of-day chart patterns with breakout and target prices, confidence levels and how the pattern behaved in the past. AI Real Time Patterns does the same during the session, and the AI Trend Prediction Engine gives, in Tickeron's words, \"AI's opinion on whether a stock may go up or down.\"",[11,435,436],{},"If you would rather not run the tools yourself, there are Daily Buy\u002FSell Signals aimed at swing traders, delivered end of day, and AI robots that generate several trade ideas a day, which its robots page sorts into 5, 15 and 60 minute timeframes. Several of the tools offer a free try, and there are apps for iOS and Android.",[11,438,439,440,444],{},"If you already spot ",[280,441,443],{"href":442},"https:\u002F\u002Fprimotrade.app\u002Fblog\u002Fbullish-and-bearish-chart-patterns","bullish and bearish chart patterns"," by eye, Tickeron is a way to hunt for them across many stocks at once.",[11,446,447,449],{},[354,448,381],{}," a \"confidence level\" on a pattern describes how similar setups behaved before, not what this one will do. Same shape, new market.",[11,451,452,454],{},[354,453,387],{}," traders who think in patterns and want a machine to do the hunting.",[21,456,458],{"id":457},"trade-ideas-built-for-the-next-hour","Trade Ideas: Built for the Next Hour",[11,460,461],{},"Trade Ideas is the most caffeinated tool on this list. Its AI, called Holly, produces buy and sell signals with entry and exit points during the session. Around it sit Money Machine for momentum setups, TI Wave for market pattern recognition, a scanner, backtesting and simulated trading on live data. Orders can go straight to the market through its TradeStation link.",[11,463,464],{},"There is a free account to explore the scans and a free weekly email of AI setups. The AI signals themselves are part of the paid plans.",[11,466,467,469],{},[354,468,381],{}," on October 8, 2026 its homepage led with four recent Holly buy signals and the big moves that followed each one. That is a highlight reel, not a track record. A list of the best calls says nothing about the calls that did not work.",[11,471,472,474],{},[354,473,387],{}," day traders who want live scans and signals and already have a plan for acting on them.",[21,476,478],{"id":477},"prosperoai-signals-in-your-pocket","Prospero.ai: Signals in Your Pocket",[11,480,481],{},"Prospero.ai is a phone-first app on iOS and Android, and it markets itself as free. Its homepage says it combines millions of data points into key signals, drawing on institutional data and market movements.",[11,483,484],{},"In the app, cumulative scores show which stocks are gaining strength or losing steam, and an in-app guide explains how to read the signals.",[11,486,487,489],{},[354,488,381],{}," the headline is confident; the fine print is careful. The homepage invites you to \"identify short-term and long-term winners\" and publishes performance figures for its newsletters. Its own disclaimer, on the same page, says those results relate to signals from its models, \"are not actual trading results,\" and that simulated results \"do not represent actual trading.\" Read that part first, here and on any tool's page.",[11,491,492],{},"It suits investors who want a signal feed on their phone and treat it as one input among several.",[21,494,496],{"id":495},"what-about-just-asking-a-chatbot","What About Just Asking a Chatbot?",[11,498,499],{},"Fair question. You can paste a chart screenshot into a general assistant such as ChatGPT and ask what it sees; OpenAI says image inputs work on both free and paid plans.",[11,501,502],{},"What it reads is the picture plus your question. What comes back is whatever shape you ask for, which is both the appeal and the problem: you are designing the analysis as you go, and nothing fixes which data it checks or how it scores anything.",[11,504,505,506,511,512,516],{},"OpenAI's own ",[280,507,510],{"href":508,"target":283,"rel":509},"https:\u002F\u002Fhelp.openai.com\u002Fen\u002Farticles\u002F8400551-image-inputs-for-chatgpt-faq",[285],"image inputs FAQ"," is frank about the weak spots. It says the model may struggle with graphs where colors or line styles vary, struggles with tasks that need precise positions, and \"may generate incorrect descriptions.\" A ",[280,513,515],{"href":514},"https:\u002F\u002Fprimotrade.app\u002Fblog\u002Fhow-to-read-candlestick-charts","candlestick chart"," is mostly colors, line styles and positions.",[21,518,520],{"id":519},"primo-trade-one-stock-or-one-chart-at-a-time","Primo Trade: One Stock or One Chart at a Time",[11,522,523],{},"Our own app, so weigh this section accordingly. It reads one thing at a time, through one of two doors.",[11,525,526,527,531],{},"Search a US-listed stock or ADR ticker, from about 16,000 symbols, and the AI reads its prices, recent news and financial metrics. Or photograph or screenshot any chart, and the AI reads the chart itself: trend, volatility, volume, sentiment and the pattern it sees, which our ",[280,528,530],{"href":529},"https:\u002F\u002Fprimotrade.app\u002Fchart-patterns-cheat-sheet","chart patterns cheat sheet"," lets you look up. If it spots a US ticker on the image, the full ticker analysis runs instead.",[289,533],{"src":534,"alt":535,"width":293,"height":294},"https:\u002F\u002Fmedia.primotrade.app\u002Fblog\u002Fai-stock-analysis-tools-compared-what-each-one-reads\u002Fphone-photographing-chart-2.jpg","Seen from behind, a person holds a phone up to photograph a sideways candlestick chart on a desk monitor in the evening.",[11,537,538,539,543],{},"Every result carries a Signal from Strong Buy to Strong Sell as the AI's directional read, a Primo Score radar of five scores from 0 to 100, a Trading Strategy idea with a risk level, and up to three Key Insights. Ticker analyses add a price chart, Recent News and a Forecast Value, the model's 6 to 12 month estimate. A chart photo has no ticker data behind it, so it gets none of those three. The ",[280,540,542],{"href":541},"https:\u002F\u002Fprimotrade.app\u002Fblog\u002Fhow-ai-reads-a-stock-chart","way AI reads a chart"," has real limits, and they apply here.",[11,545,546],{},"Its gaps, as plainly as the others':",[548,549,550,554,557,560,563],"ul",{},[551,552,553],"li",{},"No market-wide scan or ranking; you bring the stock.",[551,555,556],{},"No alerts, watchlist, portfolio, backtests, bots or order placement.",[551,558,559],{},"No real-time data, and only US-listed stocks and ADRs.",[551,561,562],{},"Mobile only, on iOS and Android.",[551,564,565],{},"A crop with no axis or timeframe limits what the AI can read.",[11,567,568],{},"It is free to download, with free analyses to start, and it fits best as a second read on one stock or chart you already have open.",[21,570,572],{"id":571},"red-flags-regulators-have-named","Red Flags Regulators Have Named",[11,574,575,576,581],{},"\"AI\" makes a handy sticker, and regulators have noticed. In March 2024 the SEC announced ",[280,577,580],{"href":578,"target":283,"rel":579},"https:\u002F\u002Fwww.sec.gov\u002Fnewsroom\u002Fpress-releases\u002F2024-36",[285],"settled charges against"," two investment advisers, Delphia and Global Predictions, for false and misleading statements about their use of AI. The agency called the practice \"AI washing,\" and the firms agreed to pay civil penalties of $225,000 and $175,000.",[11,583,584,585,590],{},"FINRA, the SEC and NASAA also published a joint alert on ",[280,586,589],{"href":587,"target":283,"rel":588},"https:\u002F\u002Fwww.finra.org\u002Finvestors\u002Finsights\u002Fartificial-intelligence-and-investment-fraud",[285],"AI and fraud",". It warns about platforms claiming their AI \"can't lose,\" and cautions that AI-generated information might rest on data that is inaccurate, incomplete or out of date. Even when the input is correct, it says, AI output \"can be faulty, or even completely made up.\"",[11,592,593,594],{},"None of the tools in this guide make \"can't lose\" claims, and several carry careful disclaimers. The alert is still a good filter for anything new you find. ",[354,595,596],{},"If a tool sells certainty, the problem is the tool, not your skepticism.",[21,598,600],{"id":599},"five-checks-for-any-ai-stock-analysis-app","Five Checks for Any AI Stock Analysis App",[11,602,603],{},"Once you know which kind of tool fits, these five questions sort a useful one from a confusing one:",[548,605,606,613,619,625,631],{},[551,607,608,612],{},[609,610,611],"em",{},"Can you see why?"," A score with no visible reasons is hard to act on. Danelfin's explanation window is one example of a tool showing its work.",[551,614,615,618],{},[609,616,617],{},"Which horizon does it mean?"," A three-month probability, a 6 to 12 month estimate and an intraday signal answer different questions, as the Nvidia Monday shows.",[551,620,621,624],{},[609,622,623],{},"What does it cover?"," US only, US and Europe, ETFs or not. Check before you search for a stock it never rates.",[551,626,627,630],{},[609,628,629],{},"How are past results framed?"," Backtests, simulated results and homepage highlight reels are not a record of real trades. The better tools say so on the same page.",[551,632,633,636],{},[609,634,635],{},"Does it fit your week?"," A platform you open once a month will not repay the hours a full charting suite asks for.",[21,638,640],{"id":639},"free-ai-stock-analysis-what-you-can-try-first","Free AI Stock Analysis: What You Can Try First",[11,642,643],{},"Almost everything here lets you look before you pay, though what is free differs a lot.",[11,645,646],{},"Prospero.ai presents its app as free. Danelfin is free to download with paid plans inside, and Primo Trade is free to download, with free analyses to start. Tickeron offers free tries of several tools, Trade Ideas has a free account and a free weekly email of AI setups, and TrendSpider includes a small monthly allowance of free Sidekick messages. Kavout keeps the AI Stock Picker in its Pro tier.",[11,648,649],{},"A free tier is a fair way to judge one thing: whether the output makes sense to you. Run it on a stock you already know well. If the AI's read matches what you understand about the company, or points out something you missed and shows why, that is a good sign. If it gives a firm signal with no visible reason, the free trial has already told you what you needed to know.",[21,651,653],{"id":652},"the-bottom-line","The Bottom Line",[11,655,656],{},"AI stock analysis tools are not one product. Rankers score the whole market, scanners hunt for setups, signal apps push a feed, chatbots answer whatever you ask, and single-stock readers explain one name or one chart. Each reads different data and answers a different question.",[11,658,659],{},"Compare them on four things: what the AI reads, what it hands back, the horizon it means and how honestly it describes its own limits. Treat every score, signal and backtest as one input to your own research, and be wary of any tool that promises more than that.",{"title":661,"searchDepth":662,"depth":662,"links":663},"",2,[664,665,666,667,668,669,670,671,672,673,674,675,676,677,678],{"id":23,"depth":662,"text":24},{"id":122,"depth":662,"text":123},{"id":271,"depth":662,"text":272},{"id":360,"depth":662,"text":361},{"id":391,"depth":662,"text":392},{"id":409,"depth":662,"text":410},{"id":429,"depth":662,"text":430},{"id":457,"depth":662,"text":458},{"id":477,"depth":662,"text":478},{"id":495,"depth":662,"text":496},{"id":519,"depth":662,"text":520},{"id":571,"depth":662,"text":572},{"id":599,"depth":662,"text":600},{"id":639,"depth":662,"text":640},{"id":652,"depth":662,"text":653},"https:\u002F\u002Fmedia.primotrade.app\u002Fblog\u002Fai-stock-analysis-tools-compared-what-each-one-reads\u002Fcover-panel.jpg","AI stock analysis tools compared by what the AI reads and hands back: scores, scanners, signal feeds and chart-photo reads, each with its catch named.","md",{},true,"\u002Fblog\u002F2026\u002F10\u002Fai-stock-analysis-tools-compared-what-each-one-reads","2026-10-08",{"title":5,"description":680},{"loc":684,"images":688,"lastmod":685},[689,690],{"loc":291},{"loc":534},"blog\u002F2026\u002F10\u002Fai-stock-analysis-tools-compared-what-each-one-reads","F1hIQYIwP49AnQ55GTdHotncKBnMRSjEVdiF4tkGVIM",[694,1220,1580],{"id":695,"title":696,"author":6,"body":697,"coverImg":1208,"description":1209,"extension":681,"head":6,"meta":1210,"navigation":683,"path":1211,"publishedAt":1212,"schemaOrg":6,"seo":1213,"sitemap":1214,"stem":1218,"updatedAt":685,"__hash__":1219},"blog\u002Fblog\u002F2026\u002F10\u002Fhow-ai-reads-a-stock-chart.md","AI Chart Analysis: What a Screenshot Can and Cannot Show",{"type":8,"value":698,"toc":1189},[699,702,705,712,716,719,722,725,729,732,736,845,848,852,857,866,869,872,875,878,887,891,894,898,901,910,914,923,926,929,933,936,951,955,958,1032,1041,1044,1048,1057,1066,1069,1073,1087,1095,1098,1102,1153,1157,1166,1174,1177,1179,1183,1186],[11,700,701],{},"It is late, a chart on your monitor looks interesting, and you grab a screenshot on your phone. You crop it so it looks tidy, drop it into an AI and ask what it sees. Back comes a confident paragraph: a steady uptrend, a short pause that looks like a bull flag, volume picking up into the right edge.",[11,703,704],{},"Here is the question nobody asks at that moment. You cropped off the price axis, the ticker and the dates. So what exactly was the AI reading?",[11,706,707,708,711],{},"The shape. ",[354,709,710],{},"AI chart analysis works from the pixels in front of it, and a screenshot keeps the shape of the price while throwing away most of what gives that shape a meaning."," This post walks through what an AI can read from a chart image, what it cannot know no matter how clever it is, where published research says these models slip, and a short routine for getting a read you can check.",[21,713,715],{"id":714},"what-an-ai-sees-when-it-reads-a-stock-chart","What an AI Sees When It Reads a Stock Chart",[11,717,718],{},"What the AI got from your crop was pixels, not prices. A multimodal model, the kind that takes in an image and text together, never sees the numbers behind a chart as a table. It sees lines, colours, candle bodies and maybe a few labels along the edges. From those it infers what the chart says, much as you would, and then puts it into words.",[11,720,721],{},"That has two consequences. The model can take in the overall shape of a chart fast, which is often exactly what you want from a second read. And everything it says depends on how well it read the picture, so a blurred photo, a cropped axis or an unusual chart style can change its answer.",[11,723,724],{},"Think of it as a very quick reader handed a page torn out of a book. It can tell you the story on the page. It cannot tell you which book it came from, or how the next chapter goes.",[21,726,728],{"id":727},"ai-chart-analysis-readable-versus-missing","AI Chart Analysis: Readable Versus Missing",[11,730,731],{},"Some things live in the pixels. Others live in labels that a crop removes, and a few were never on the screen at all.",[289,733],{"src":734,"alt":735,"width":293,"height":294},"https:\u002F\u002Fmedia.primotrade.app\u002Fblog\u002Fhow-ai-reads-a-stock-chart\u002Fscreenshot-readable-vs-missing-panel.jpg","Diagram of a cropped candlestick chart screenshot, with the rising trend, a candle, the sideways flag pattern and the volume marked as readable, and the ticker, timeframe, price axis and dates marked as missing.",[29,737,738,751],{},[32,739,740],{},[35,741,742,745,748],{},[38,743,744],{},"What you want to know",[38,746,747],{},"In the pixels?",[38,749,750],{},"Where it has to come from",[48,752,753,764,773,783,794,805,815,825,835],{},[35,754,755,758,761],{},[53,756,757],{},"Direction of the trend",[53,759,760],{},"Yes",[53,762,763],{},"The candles themselves",[35,765,766,769,771],{},[53,767,768],{},"Shape of a pattern, such as a flag or a double top",[53,770,760],{},[53,772,763],{},[35,774,775,778,781],{},[53,776,777],{},"Long wicks, small bodies, gaps between candles",[53,779,780],{},"Yes, if the image is sharp",[53,782,763],{},[35,784,785,788,791],{},[53,786,787],{},"Whether volume rose or fell",[53,789,790],{},"Yes, if the volume bars are in frame",[53,792,793],{},"The volume panel",[35,795,796,799,802],{},[53,797,798],{},"Which stock it is",[53,800,801],{},"No",[53,803,804],{},"The ticker label, or you",[35,806,807,810,812],{},[53,808,809],{},"The price level, and how big a move was in percent",[53,811,801],{},[53,813,814],{},"The price axis",[35,816,817,820,822],{},[53,818,819],{},"Whether each candle is five minutes or one week",[53,821,801],{},[53,823,824],{},"The timeframe label or the dates",[35,826,827,830,832],{},[53,828,829],{},"When it happened, and what news was out",[53,831,801],{},[53,833,834],{},"The dates, then a news check",[35,836,837,840,842],{},[53,838,839],{},"What happened after the right edge",[53,841,801],{},[53,843,844],{},"Nothing: no screenshot has it",[11,846,847],{},"The top half of that table is where an AI chart read pulls its weight. The bottom half is where it can only guess, and a guess written in fluent sentences looks a lot like knowledge.",[21,849,851],{"id":850},"three-things-a-cropped-screenshot-throws-away","Three Things a Cropped Screenshot Throws Away",[853,854,856],"h3",{"id":855},"the-timeframe","The timeframe",[11,858,859,860,865],{},"Traders have long said that charts look alike at every zoom level, and a famous mathematician took the folklore seriously. Benoit Mandelbrot, the man behind fractals, wrote that once a stock chart is enlarged or shrunk to fit the same frame, an observer ",[280,861,864],{"href":862,"target":283,"rel":863},"https:\u002F\u002Fwww.scientificamerican.com\u002Farticle\u002Fmultifractals-explain-wall-street\u002F",[285],"cannot tell which"," data change week to week, day to day or hour to hour. He published that in Scientific American in 1999.",[11,867,868],{},"For a screenshot, that means a bull flag on a five-minute chart and a bull flag on a weekly chart can be the same picture. One describes an afternoon, the other most of a year. Without a timeframe label or readable dates, the AI cannot tell them apart, and neither can you.",[853,870,814],{"id":871},"the-price-axis",[11,873,874],{},"Without the axis, the AI cannot know whether the stock trades at 3 dollars or 300, or whether a tall candle was a 1% wobble or a 10% drop. Those are not small details.",[11,876,877],{},"The scale matters too. On a linear chart, a move from 10 to 20 takes the same height as a move from 70 to 80; on a log chart, equal percentage moves take equal height. Over a long stretch, the same prices drawn both ways can give different trendlines, and a screenshot rarely says which scale it used.",[11,879,880,881,886],{},"Reading values off an axis is also one of the weakest skills these models have. FinChart-Bench, the first benchmark built only from real financial charts (1,200 charts from 2015 to 2024 and 7,016 questions, posted in 2025 and presented at ACL 2026), found that models did relatively well when numbers were printed next to the points and poorly when a value had to be inferred ",[280,882,885],{"href":883,"target":283,"rel":884},"https:\u002F\u002Farxiv.org\u002Fabs\u002F2507.14823",[285],"from the axes",". Its authors call spatial reasoning a significant bottleneck. A price chart is exactly that kind of chart: almost nothing is labelled, and everything is read against an axis.",[853,888,890],{"id":889},"the-ticker","The ticker",[11,892,893],{},"No ticker means no company, no earnings date and no news, so no way to know whether the gap on the right was a company announcement or a market-wide sell-off. An AI working from pixels alone reads the chart the way a technical analyst would with a sticky note over the name. That is an interesting exercise, and an incomplete one.",[21,895,897],{"id":896},"when-the-picture-lies-nvidias-split-in-june-2024","When the Picture Lies: Nvidia's Split in June 2024",[11,899,900],{},"Here is the failure case as one real, closed example from June 2024, read after the fact.",[11,902,903,904,909],{},"Nvidia split its stock ten for one that month. As ",[280,905,908],{"href":906,"target":283,"rel":907},"https:\u002F\u002Ffinance.yahoo.com\u002Fnews\u002Fnvidia-stock-rises-after-10-for-1-stock-split-204412528.html",[285],"Yahoo Finance reported"," on Monday, June 10, 2024, the shares began trading that day on the new basis, Friday's closing price of $1,208.88 was revised to $120.88, and the stock closed up nearly 1% on its first day after the split. Nothing bad happened to the company that weekend. Each shareholder simply held ten shares for every one.",[289,911],{"src":912,"alt":913,"width":293,"height":294},"https:\u002F\u002Fmedia.primotrade.app\u002Fblog\u002Fhow-ai-reads-a-stock-chart\u002Fsplit-cliff-chart-panel.jpg","Stylised candlestick chart that climbs steadily, then drops in one sheer vertical cliff and carries on calmly at a much lower level.",[11,915,916,917,922],{},"Now picture a chart that does not adjust its history for the split. It shows a stock that climbed for months and then lost about 90% over a weekend. StockCharts, which adjusts its own data for exactly this reason, warns that an unadjusted split ",[280,918,921],{"href":919,"target":283,"rel":920},"https:\u002F\u002Fhelp.stockcharts.com\u002Fdata-and-ticker-symbols\u002Fdata-availability\u002Fprice-data-adjustments",[285],"gives the impression"," that something bearish happened, and that most technical indicators on such a chart would give sell signals.",[11,924,925],{},"Show a screenshot of that cliff to an AI with no ticker and no dates, and \"a collapse\" is a perfectly reasonable description of the pixels. It is also completely wrong. The fix is not a smarter model. It is context: the ticker, the date and a quick news check.",[11,927,928],{},"The general rule holds well beyond splits: the same picture of a price can come from two very different stories, and the picture alone cannot say which.",[21,930,932],{"id":931},"how-well-ai-reads-charts-what-the-benchmarks-say","How Well AI Reads Charts: What the Benchmarks Say",[11,934,935],{},"So how often do these models get a chart right? Researchers measure it with benchmarks: sets of real charts with questions that have checkable answers.",[11,937,938,939,944,945,950],{},"Early ones such as the ",[280,940,943],{"href":941,"target":283,"rel":942},"https:\u002F\u002Farxiv.org\u002Fabs\u002F2203.10244",[285],"ChartQA benchmark",", published in 2022, asked thousands of questions about charts, including questions that need arithmetic over the values shown, and models improved on them quickly. The authors of CharXiv, a 2024 benchmark, argued that such datasets used simple, uniform charts and template questions, which gave an ",[280,946,949],{"href":947,"target":283,"rel":948},"https:\u002F\u002Farxiv.org\u002Fabs\u002F2406.18521",[285],"over-optimistic measure"," of progress. When they changed the charts or questions only slightly, scores fell by up to 34.5%.",[853,952,954],{"id":953},"harder-tests-wider-gap","Harder tests, wider gap",[11,956,957],{},"Harder tests show a clear gap between models and people:",[29,959,960,979],{},[32,961,962],{},[35,963,964,967,970,973,976],{},[38,965,966],{},"Benchmark",[38,968,969],{},"What it tests",[38,971,972],{},"People",[38,974,975],{},"Best model at release",[38,977,978],{},"Best open-source model",[48,980,981,998,1015],{},[35,982,983,986,989,992,995],{},[53,984,985],{},"CharXiv (2024), reasoning questions",[53,987,988],{},"2,323 real charts from scientific papers",[53,990,991],{},"80.5%",[53,993,994],{},"47.1% (GPT-4o)",[53,996,997],{},"29.2%",[35,999,1000,1003,1006,1009,1012],{},[53,1001,1002],{},"ChartMuseum (NeurIPS 2025)",[53,1004,1005],{},"1,162 questions on real charts from 184 sources",[53,1007,1008],{},"93%",[53,1010,1011],{},"63.0% (Gemini 2.5 Pro)",[53,1013,1014],{},"38.5%",[35,1016,1017,1020,1023,1026,1029],{},[53,1018,1019],{},"FinChart-Bench (2025), open questions",[53,1021,1022],{},"1,200 real financial charts",[53,1024,1025],{},"Not reported",[53,1027,1028],{},"63.59%",[53,1030,1031],{},"59.78%",[11,1033,1034,1035,1040],{},"The ",[280,1036,1039],{"href":1037,"target":283,"rel":1038},"https:\u002F\u002Farxiv.org\u002Fabs\u002F2505.13444",[285],"ChartMuseum authors"," also split questions that need mainly visual reasoning from questions that lean on text. On the visual ones, every model they tested dropped 35% to 55% from its score on the text-heavy ones. In a synthetic test that could only be solved by looking, models got worse as the picture got busier, while people held steady. A price chart is almost all visual, so the parts that need careful looking, such as exact levels, small candles and crowded stretches, are where a model is most likely to slip.",[11,1042,1043],{},"CharXiv and ChartMuseum use scientific and general charts, FinChart-Bench's 14 chart types are bars, lines, pies and the like rather than candlesticks, and newer models keep raising the scores. The direction is still clear: on charts that need real looking, the best models trail people by a wide margin.",[21,1045,1047],{"id":1046},"can-ai-predict-a-stock-from-its-chart","Can AI Predict a Stock From Its Chart?",[11,1049,1050,1051,1056],{},"The most cited research here is about text, not charts. Alejandro Lopez-Lira and Yuehua Tang had GPT-4 score news headlines published after its knowledge cutoff. The scores captured the market's first reaction to the news, which they describe as non-tradable, and also predicted some of the drift afterwards, especially for small stocks and bad news. The edge also shrank: in their data it ",[280,1052,1055],{"href":1053,"target":283,"rel":1054},"https:\u002F\u002Farxiv.org\u002Fabs\u002F2304.07619",[285],"faded as adoption"," of these models rose.",[11,1058,1059,1060,1065],{},"Testing such claims is tricky. Paul Glasserman and Caden Lin showed that backtests of AI news strategies can be ",[280,1061,1064],{"href":1062,"target":283,"rel":1063},"https:\u002F\u002Farxiv.org\u002Fabs\u002F2309.17322",[285],"quietly biased"," when the model already knows what happened to the stock after the news, or when its general knowledge of a company distracts it from the text in front of it.",[11,1067,1068],{},"None of this research shows that an AI reading a chart can tell you where a price will go. A chart is a record of past prices, and an AI read of it is one more reading of that same record: quick, often useful, and never a view of the future.",[21,1070,1072],{"id":1071},"confident-fluent-and-sometimes-wrong","Confident, Fluent and Sometimes Wrong",[11,1074,1075,1076,1080,1081,1086],{},"The SEC, NASAA and FINRA, in a joint alert on AI and investment fraud, warn that even when the input is correct, AI output ",[280,1077,1079],{"href":587,"target":283,"rel":1078},[285],"can be faulty",", or even completely made up. The same alert quotes pitches such as \"Our proprietary AI trading system can't lose!\" as the kind to run from, and it urges caution about ",[280,1082,1085],{"href":1083,"target":283,"rel":1084},"https:\u002F\u002Fwww.investor.gov\u002Fintroduction-investing\u002Fgeneral-resources\u002Fnews-alerts\u002Falerts-bulletins\u002Finvestor-alerts\u002Fartificial-intelligence-fraud",[285],"using AI-generated information"," to predict the market's direction or a stock's price, because it may rest on data that is incomplete, outdated or misleading.",[11,1088,1089,1090,1094],{},"The SEC has acted on overstated AI claims too. In 2024 it settled charges against two firms, Delphia and Global Predictions, for false and misleading statements about their use of AI, with ",[280,1091,1093],{"href":578,"target":283,"rel":1092},[285],"$400,000 in penalties"," between them.",[11,1096,1097],{},"A chart description reads just as smoothly whether it is right or wrong. So the useful habit is neither trusting nor distrusting the AI. It is checking it.",[21,1099,1101],{"id":1100},"how-to-get-an-ai-chart-read-you-can-check","How to Get an AI Chart Read You Can Check",[1103,1104,1105,1111,1117,1123,1136,1142],"ol",{},[551,1106,1107,1110],{},[354,1108,1109],{},"Keep the labels in."," Crop to the chart, but leave the ticker, the price axis and the dates in frame. They are the whole bottom half of the table above.",[551,1112,1113,1116],{},[354,1114,1115],{},"Name the timeframe if it is not shown."," \"Daily candles, about six months\" changes what a pattern means.",[551,1118,1119,1122],{},[354,1120,1121],{},"Shoot straight and sharp."," A photo of a screen at an angle, with glare, blurs the small candles a model already finds hardest.",[551,1124,1125,1128,1129,1132,1133,1135],{},[354,1126,1127],{},"Check every claim against the chart."," If it names a level or a pattern, find it yourself. You can only do that if you can read the chart, which our guides to ",[280,1130,1131],{"href":514},"reading candlestick charts"," and ",[280,1134,443],{"href":442}," cover.",[551,1137,1138,1141],{},[354,1139,1140],{},"Ask what it cannot see."," A good follow-up is \"what would change this read?\" Earnings next week, a split or a market-wide drop are all invisible in a cropped chart.",[551,1143,1144,1147,1148,1152],{},[354,1145,1146],{},"Plan for being wrong."," Decide in advance where you would get out. Our guide to ",[280,1149,1151],{"href":1150},"https:\u002F\u002Fprimotrade.app\u002Fblog\u002Fwhat-is-a-stop-loss","what a stop-loss is"," explains how.",[21,1154,1156],{"id":1155},"how-primo-trade-reads-a-chart-photo","How Primo Trade Reads a Chart Photo",[11,1158,1159,1162,1163,1165],{},[280,1160,102],{"href":1161},"https:\u002F\u002Fprimotrade.app\u002Fdownload?c=post_inline"," lets you take a photo of a chart or choose a screenshot from your gallery. Its AI reads the trend, volatility, volume and sentiment it sees, the chart pattern, and what that pattern usually signals. Our ",[280,1164,530],{"href":529}," shows what confirms each one. If the image is not a chart, the app says so instead of guessing.",[11,1167,1168,1169,1173],{},"The limits in this post apply to it as well, and our ",[280,1170,1172],{"href":1171},"https:\u002F\u002Fprimotrade.app\u002Fblog\u002Fai-stock-analysis-tools-compared-what-each-one-reads","comparison of AI stock analysis tools"," sets it beside six others. When the AI can read a US ticker on the image, the app runs the full ticker analysis instead, pulling prices, financial metrics and recent news before the AI writes, and the result adds a price chart and a Forecast Value, the model's 6 to 12 month estimate. When it cannot read a ticker, it works from the image alone, so the result has no price chart, no Forecast Value and no news. One more reason to keep the ticker in frame.",[11,1175,1176],{},"Either way, the result shows a signal from Strong Buy to Strong Sell as the AI's read, a Primo Score of five ratings from 0 to 100, a Trading Strategy idea with a risk level, and Key Insights. A Risk score of 80 is a rating on a scale, not an 80% chance of anything.",[11,1178,268],{},[21,1180,1182],{"id":1181},"what-an-ai-chart-read-is-good-for","What an AI Chart Read Is Good For",[11,1184,1185],{},"An AI reads a stock chart the way it reads any image: from pixels into words. From a screenshot it can describe the trend, the candles, the pattern and the volume in seconds. It cannot know the ticker, the price level, the timeframe or the date unless they are in the picture, and the best models still trail people on charts that need careful looking, most of all when a value has to be read off an axis.",[11,1187,1188],{},"So keep the labels in, check what it says against the chart, and treat it as a quick second opinion on past prices. Any tool that claims an AI can see the future in a chart is the warning sign regulators describe.",{"title":661,"searchDepth":662,"depth":662,"links":1190},[1191,1192,1193,1199,1200,1203,1204,1205,1206,1207],{"id":714,"depth":662,"text":715},{"id":727,"depth":662,"text":728},{"id":850,"depth":662,"text":851,"children":1194},[1195,1197,1198],{"id":855,"depth":1196,"text":856},3,{"id":871,"depth":1196,"text":814},{"id":889,"depth":1196,"text":890},{"id":896,"depth":662,"text":897},{"id":931,"depth":662,"text":932,"children":1201},[1202],{"id":953,"depth":1196,"text":954},{"id":1046,"depth":662,"text":1047},{"id":1071,"depth":662,"text":1072},{"id":1100,"depth":662,"text":1101},{"id":1155,"depth":662,"text":1156},{"id":1181,"depth":662,"text":1182},"https:\u002F\u002Fmedia.primotrade.app\u002Fblog\u002Fhow-ai-reads-a-stock-chart\u002Fcover-v3.jpg","AI chart analysis reads trend, candles and patterns from a screenshot, but not the ticker, the timeframe or the price scale. What it sees, where it slips.",{},"\u002Fblog\u002F2026\u002F10\u002Fhow-ai-reads-a-stock-chart","2026-10-06",{"title":696,"description":1209},{"loc":1211,"images":1215,"lastmod":685},[1216,1217],{"loc":734},{"loc":912},"blog\u002F2026\u002F10\u002Fhow-ai-reads-a-stock-chart","oH2xwoJGjea3JgXemVUfFbDP4tXnpO_lFqgvPsJTCXY",{"id":1221,"title":1222,"author":6,"body":1223,"coverImg":1569,"description":1570,"extension":681,"head":6,"meta":1571,"navigation":683,"path":1572,"publishedAt":1573,"schemaOrg":6,"seo":1574,"sitemap":1575,"stem":1578,"updatedAt":6,"__hash__":1579},"blog\u002Fblog\u002F2026\u002F11\u002Fbull-vs-bear-market-how-long-each-one-really-lasts.md","Bull vs Bear Market: How Long Each One Really Lasts",{"type":8,"value":1224,"toc":1552},[1225,1228,1235,1238,1242,1245,1262,1265,1269,1272,1275,1284,1288,1291,1295,1298,1389,1392,1395,1399,1408,1411,1414,1420,1424,1427,1436,1439,1443,1446,1449,1453,1456,1459,1462,1471,1475,1478,1481,1484,1488,1494,1498,1501,1510,1516,1519,1523,1526,1529,1531,1542,1546,1549],[11,1226,1227],{},"On April 8, 2025, the S&P 500 closed 18.9% below its February record, and the evening coverage said it was inches from a bear market. It never became one. By the closing-price rule, the index stopped 1.1 percentage points short, and the drop went into the record as a correction.",[11,1229,1230,1231,1234],{},"That near miss is the whole bull vs bear market question in one day. ",[354,1232,1233],{},"A bear market is a fall of 20% or more from a closing high, and a bull market is a rise of 20% or more from a closing low, but how long each one lasts depends on who is counting."," Since 1928, bear markets in the S&P 500 have lasted about 9.6 months on one count and close to 19 months on another. Both numbers are honest. They measure different things.",[11,1236,1237],{},"This post lays the counts side by side, explains why they disagree, and shows why neither label can be read off a chart in real time.",[21,1239,1241],{"id":1240},"bull-vs-bear-market-the-key-numbers","Bull vs bear market: the key numbers",[11,1243,1244],{},"Here is what the published research agrees on, before the detail:",[548,1246,1247,1250,1253,1256,1259],{},[551,1248,1249],{},"Bear markets are shorter than bull markets in every count, and usually by a wide margin.",[551,1251,1252],{},"The typical bear market fell about a third from its peak: 33% to 38%, depending on the source.",[551,1254,1255],{},"The shortest bear on record lasted 33 days, in 2020. The longest, measured from 1872, ran 113 months.",[551,1257,1258],{},"There have been more bear markets than recessions, so a falling market does not always mean a shrinking economy.",[551,1260,1261],{},"Both labels are only certain after the fact, because the threshold has to be crossed before anyone can date the turn.",[11,1263,1264],{},"The rest of this post is where those numbers come from, and why the same history can produce different ones.",[21,1266,1268],{"id":1267},"what-counts-as-a-bull-or-bear-market","What counts as a bull or bear market",[11,1270,1271],{},"The common rule is simple. A bear market starts when a broad index such as the S&P 500 closes at least 20% below its most recent high. A drop of 10% to 19.9% is a correction. A new bull market begins when the index closes 20% above its low.",[11,1273,1274],{},"The word that does the work is \"closes\". In April 2025 the S&P 500 traded below the 20% line during the session, but on the closing measure it never got there. The rule ignores intraday lows on purpose: a few panicked hours do not make a regime.",[11,1276,1277,1278,1283],{},"Bull markets have the looser definition of the two. According to ",[280,1279,1282],{"href":1280,"target":283,"rel":1281},"https:\u002F\u002Fwww.fidelity.com\u002Flearning-center\u002Fsmart-money\u002Fbullish-meaning",[285],"Fidelity's explainer",", the SEC describes a bull market as a broad index rising 20% or more over at least two months, while other sources set no exact threshold at all and simply look for new highs.",[853,1285,1287],{"id":1286},"the-1990-edge-case","The 1990 edge case",[11,1289,1290],{},"Thresholds create strange edges. In 1990 the S&P 500 fell 19.9%, a hair short of the bear line. Because of that missing tenth of a point, the bull market that began in December 1987 is technically the longest on record, running until the dot-com crash in March 2000.",[21,1292,1294],{"id":1293},"how-long-bear-markets-last-by-four-counts","How long bear markets last, by four counts",[11,1296,1297],{},"Ask four research houses how long a bear market lasts and you get four answers. The table puts them next to each other.",[29,1299,1300,1319],{},[32,1301,1302],{},[35,1303,1304,1307,1310,1313,1316],{},[38,1305,1306],{},"Source",[38,1308,1309],{},"Period",[38,1311,1312],{},"Bear markets counted",[38,1314,1315],{},"Typical length",[38,1317,1318],{},"Typical decline",[48,1320,1321,1338,1355,1372],{},[35,1322,1323,1326,1329,1332,1335],{},[53,1324,1325],{},"Ned Davis Research, via Hartford Funds",[53,1327,1328],{},"1928 to 2024",[53,1330,1331],{},"27",[53,1333,1334],{},"289 days on average (9.6 months)",[53,1336,1337],{},"35% on average",[35,1339,1340,1343,1346,1349,1352],{},[53,1341,1342],{},"Fidelity",[53,1344,1345],{},"Since 1872",[53,1347,1348],{},"26",[53,1350,1351],{},"19 months, median",[53,1353,1354],{},"33%, median",[35,1356,1357,1360,1363,1366,1369],{},[53,1358,1359],{},"S&P Dow Jones Indices, via CNBC (2022)",[53,1361,1362],{},"Since the late 1920s",[53,1364,1365],{},"Not given",[53,1367,1368],{},"Almost 19 months on average",[53,1370,1371],{},"38% on average",[35,1373,1374,1377,1380,1383,1386],{},[53,1375,1376],{},"Edward Jones",[53,1378,1379],{},"Since 1946",[53,1381,1382],{},"12",[53,1384,1385],{},"About 16 months on average",[53,1387,1388],{},"About 34% on average",[11,1390,1391],{},"A fifth count, from Yardeni Research and cited by Investopedia, finds 22 bear markets since the 1929 crash, with an average length of about 11.4 months.",[11,1393,1394],{},"The decline column barely moves: every source lands between a third and two fifths. The length column runs from under 10 months to 19. That spread is the real story.",[853,1396,1398],{"id":1397},"why-the-counts-disagree","Why the counts disagree",[11,1400,1401,1402,1407],{},"The main reason is how each source splits a long slide into pieces. The ",[280,1403,1406],{"href":1404,"target":283,"rel":1405},"https:\u002F\u002Fwww.hartfordfunds.com\u002Fpractice-management\u002Fclient-conversations\u002Fmanaging-volatility\u002Fbear-markets.html",[285],"Hartford Funds summary"," of Ned Davis data counts 12 separate bear markets between 1928 and 1945 alone. The Great Depression decline, for example, appears there as six distinct bears between 1929 and 1933, each with its own start and end date, because the index bounced enough in between to reset the clock.",[11,1409,1410],{},"Chop one long decline into six short ones and the average length falls fast. Count it as one decline from the 1929 peak to the 1932 trough, as many closing-basis lists do, and it becomes a single bear of almost three years. Same prices, different rule.",[11,1412,1413],{},"The second reason is the statistic itself. Fidelity reports medians, while Hartford and S&P report averages. A few very long bears pull an average up, so a median is usually the better guide to a typical one.",[11,1415,1416,1417],{},"The third reason is the start date. Edward Jones begins in 1946 and so leaves out the 1930s, the most violent decade in the record. ",[354,1418,1419],{},"When you read \"the average bear market lasts X months\", the first question is whose rule, and since when.",[21,1421,1423],{"id":1422},"how-long-bull-markets-last","How long bull markets last",[11,1425,1426],{},"The bull side shows the same pattern of agreement on direction and disagreement on size.",[11,1428,1429,1430,1435],{},"Ned Davis Research, through Hartford Funds, counts 28 bull markets since 1928, averaging 988 days, or about 2.7 years, with an average rise of 112%. Fidelity's count from 1872 finds 26 bull markets with a median length of 42 months, about 3.5 years, and a median rise of 87%. The ",[280,1431,1434],{"href":1432,"target":283,"rel":1433},"https:\u002F\u002Fwww.edwardjones.com\u002Fsites\u002Fdefault\u002Ffiles\u002Fdam\u002F\u002Fang-4011-a.pdf",[285],"Edward Jones study",", which starts in 1946, puts the average bull at about five and a half years with a rise of about 192%.",[11,1437,1438],{},"The Edward Jones figure is the largest because its window excludes the short, broken rallies of the 1930s. The Ned Davis figure is the smallest because its rule ends a bull at every 20% drop, and there were many of those before the Second World War.",[853,1440,1442],{"id":1441},"time-spent-in-each","Time spent in each",[11,1444,1445],{},"The ratio matters more than any single average. Hartford's summary notes that, of the 95 years it covers, bear markets took up about 21.4. The other 78% or so of the time, the index was in a bull market or a correction inside one.",[11,1447,1448],{},"That is why the two words feel lopsided in practice. Most of the years an investor lives through are bull years, and the bear years arrive faster and hurt more per month.",[21,1450,1452],{"id":1451},"the-shortest-bear-and-the-most-recent-one","The shortest bear and the most recent one",[11,1454,1455],{},"Two recent bear markets show how far apart the shapes can be.",[11,1457,1458],{},"The 2020 bear ran from February 19 to March 23, 33 days in total, and took the S&P 500 down 33.92%. It was the shortest bear market in the Ned Davis record and one of the steepest. A trader who blinked missed the whole thing.",[11,1460,1461],{},"The 2022 bear ran from January 3 to October 12, 282 days, with a decline of 25.43%. It was close to the long-run average in length and milder than average in depth. It also arrived without a recession, which surprised plenty of people who had learned the two words as a pair.",[11,1463,1464,1465,1470],{},"Then came 2025. On April 8, as ",[280,1466,1469],{"href":1467,"target":283,"rel":1468},"https:\u002F\u002Fwww.cnn.com\u002F2025\u002F04\u002F24\u002Finvesting\u002Fus-stock-market-bottom",[285],"CNN reported",", the S&P 500 posted its lowest close of the year, down 18.9% from its February peak. Under every source in the table above, that was a correction, not a bear market. It shows up in no bear market count, even though it came within 1.1 percentage points of the line.",[21,1472,1474],{"id":1473},"why-the-label-only-arrives-after-the-turn","Why the label only arrives after the turn",[11,1476,1477],{},"Here is the part most explainers skip. Neither label can be applied on the day the market turns.",[11,1479,1480],{},"A bear market is only declared once the index is already 20% down. By then most of the decline has usually happened: the median bear in Fidelity's data fell 33%, so the official label arrives with well over half of the typical drop already behind it.",[11,1482,1483],{},"A bull market is worse. The rule says it begins at the closing low, but nobody knows a low was the low until the index has climbed 20% from it. The bull market that began at the October 12, 2022 low was only recognised months later, and then dated back to that day. Hartford's summary adds that about 36% of the S&P 500's strongest days between 2003 and March 2025 came in the first two months of a bull market, before it was clear one had begun.",[289,1485],{"src":1486,"alt":1487,"width":293,"height":294},"https:\u002F\u002Fmedia.primotrade.app\u002Fblog\u002Fbull-vs-bear-market-how-long-each-one-really-lasts\u002Fcycle-labels-after-the-fact.jpg","Diagram of a price line that falls from a peak to a trough and climbs again, with the bear and bull labels placed only after each threshold is crossed",[11,1489,1490,1493],{},[354,1491,1492],{},"Both labels are history, not forecasts."," They describe a stretch of the past once it can be measured, which is why the counts in the table above can be compiled at all, and why none of them tells you where today's market sits inside its own cycle.",[21,1495,1497],{"id":1496},"bear-markets-and-recessions","Bear markets and recessions",[11,1499,1500],{},"The two words travel together in headlines, but the record is looser than that. Hartford's summary counts 27 bear markets since 1928 against only 15 recessions in the same period, using recession dates from the National Bureau of Economic Research.",[11,1502,1503,1504,1509],{},"According to ",[280,1505,1508],{"href":1506,"target":283,"rel":1507},"https:\u002F\u002Fwww.fidelity.com\u002Flearning-center\u002Fsmart-money\u002Fbear-vs-bull-market",[285],"Fidelity's comparison",", about a quarter of bear markets came with no recession at all. The Crash of 1987 is the famous one: the S&P 500 fell about a third in a few months and the economy kept growing. The 2022 bear is the recent one.",[11,1511,1512,1513],{},"Depth varies too. In the Edward Jones data, only three of the 12 bear markets since 1946 fell 40% or more. ",[354,1514,1515],{},"The deep, multi-year collapses that define the word for many people are the exception in the record, not the rule.",[11,1517,1518],{},"None of this means a bear market is mild. A 25% fall in an index can still mean far larger falls in single stocks, and the recovery to the old high can take years after the bottom.",[21,1520,1522],{"id":1521},"what-an-index-label-means-for-one-stocks-chart","What an index label means for one stock's chart",[11,1524,1525],{},"Bull and bear markets are labels for an index. A single stock can sit in its own downtrend while the S&P 500 climbs to records, or hold up through a bear market while the index falls a third. The label tells you about the average of 500 companies, not about the one on your screen.",[11,1527,1528],{},"That is where reading the individual chart comes in. Primo Trade's Analyze Chart takes a photo or screenshot of a stock chart and gives an AI read of the trend, volatility, volume and sentiment it sees, along with the chart pattern it recognises and what that pattern usually signals. From a ticker, the app also pulls the stock's prices, news and financial metrics into one AI analysis.",[11,1530,268],{},[11,1532,1533,1534,1536,1537,1541],{},"The AI's read is a description of what one chart shows, so it has the same limit as the bull and bear labels: it describes the past and present of the chart, not what comes next. If you want the vocabulary behind the shapes it names, the guide to ",[280,1535,443],{"href":442}," covers the common ones. If you use a moving-average crossover as your own trend filter, the post on the ",[280,1538,1540],{"href":1539},"https:\u002F\u002Fprimotrade.app\u002Fblog\u002Fgolden-cross-what-seven-decades-of-signals-show","golden cross signal"," looks at what seven decades of those signals show.",[21,1543,1545],{"id":1544},"what-the-numbers-add-up-to","What the numbers add up to",[11,1547,1548],{},"Bull markets have lasted longer than bear markets in every published count, from about 2.7 years against 9.6 months on the Ned Davis rule to about five and a half years against 16 months since 1946. The typical bear fell about a third, a figure the sources roughly agree on, while their lengths differ because each source splits long declines differently and starts in a different year.",[11,1550,1551],{},"The 20% line is a convention measured on closing prices, which is why April 2025 was a correction and 1990 was not a bear market at all. Both labels are applied after the threshold is crossed, so they mark history rather than the turn itself. And a bear market is a statement about an index, not about the economy and not about any one stock.",{"title":661,"searchDepth":662,"depth":662,"links":1553},[1554,1555,1558,1561,1564,1565,1566,1567,1568],{"id":1240,"depth":662,"text":1241},{"id":1267,"depth":662,"text":1268,"children":1556},[1557],{"id":1286,"depth":1196,"text":1287},{"id":1293,"depth":662,"text":1294,"children":1559},[1560],{"id":1397,"depth":1196,"text":1398},{"id":1422,"depth":662,"text":1423,"children":1562},[1563],{"id":1441,"depth":1196,"text":1442},{"id":1451,"depth":662,"text":1452},{"id":1473,"depth":662,"text":1474},{"id":1496,"depth":662,"text":1497},{"id":1521,"depth":662,"text":1522},{"id":1544,"depth":662,"text":1545},"https:\u002F\u002Fmedia.primotrade.app\u002Fblog\u002Fbull-vs-bear-market-how-long-each-one-really-lasts.jpg","Bull vs bear market history in numbers: how long each lasted, why sources count them differently, and why the label only arrives after the turn.",{},"\u002Fblog\u002F2026\u002F11\u002Fbull-vs-bear-market-how-long-each-one-really-lasts","2026-11-02",{"title":1222,"description":1570},{"loc":1572,"images":1576},[1577],{"loc":1486},"blog\u002F2026\u002F11\u002Fbull-vs-bear-market-how-long-each-one-really-lasts","vDAHsL_Owa6jjMYY-ujdRd8anb1HMIh6NvOIZWG7R5M",{"id":1581,"title":1582,"author":6,"body":1583,"coverImg":1928,"description":1929,"extension":681,"head":6,"meta":1930,"navigation":683,"path":1931,"publishedAt":1932,"schemaOrg":6,"seo":1933,"sitemap":1934,"stem":1937,"updatedAt":6,"__hash__":1938},"blog\u002Fblog\u002F2026\u002F10\u002Fdoji-candlestick-5-types-and-when-each-one-misleads.md","Doji Candlestick: 5 Types and When Each One Misleads",{"type":8,"value":1584,"toc":1915},[1585,1588,1591,1595,1598,1607,1610,1614,1625,1629,1632,1635,1722,1729,1733,1736,1739,1742,1745,1748,1752,1755,1764,1767,1776,1780,1783,1786,1794,1797,1801,1804,1813,1816,1819,1823,1826,1829,1832,1835,1839,1842,1845,1848,1865,1868,1872,1875,1879,1882,1885,1887,1902,1906,1909,1912],[11,1586,1587],{},"A stock climbs for six sessions, then prints a candle with almost no body and a wick on each side. Someone in a trading chat calls it the top. A doji candlestick is a candle whose open and close sit at or very near the same price, so the session ended roughly where it began. It records a standoff between buyers and sellers, and on its own it does not say which side comes out ahead.",[11,1589,1590],{},"There are five common types, and each comes with a story traders like to tell about it. Testing across millions of candles found that most of those stories hold about half the time. This guide walks through each type: what the shape shows, what the tests found, and what has to happen next before a doji tells you anything.",[21,1592,1594],{"id":1593},"what-a-doji-candlestick-actually-records","What a doji candlestick actually records",[11,1596,1597],{},"A candle has four prices: open, high, low and close. The body spans open to close, and the wicks, or shadows, reach out to the high and the low. In a doji the body shrinks to a thin line, because the close landed back at the open after price moved around during the session.",[11,1599,1600,1601,1606],{},"StockCharts' ",[280,1602,1605],{"href":1603,"target":283,"rel":1604},"https:\u002F\u002Fchartschool.stockcharts.com\u002Ftable-of-contents\u002Fchart-analysis\u002Fcandlestick-charts\u002Fintroduction-to-candlesticks",[285],"ChartSchool guide"," describes the result as looking like a cross, an inverted cross or a plus sign, and puts the key point plainly: alone, doji are neutral patterns, and any bullish or bearish bias comes from the price action before them and the confirmation after them.",[11,1608,1609],{},"\"Near the same price\" depends on the stock. The same guide gives a scale: a 20 dollar stock might form a doji with a 0.125 point gap between open and close, while a 200 dollar stock might form one with a 1.25 point gap. What matters is that the body looks like a thin line next to the candles around it.",[289,1611],{"src":1612,"alt":1613,"width":293,"height":294},"https:\u002F\u002Fmedia.primotrade.app\u002Fblog\u002Fdoji-candlestick-5-types-and-when-each-one-misleads\u002Fdoji-types.jpg","Diagram of five doji candlesticks side by side, labelled standard, long-legged, dragonfly, gravestone and four-price",[11,1615,1616,1617,1620,1621,1624],{},"Context changes the weight of a doji too. The same guide cites Steve Nison, who noted that a doji among other small-bodied candles is not important, while a doji that appears among candles with long bodies deserves attention. ",[354,1618,1619],{},"A doji is only as loud as the candles around it."," If candle anatomy is new to you, the ",[280,1622,1623],{"href":514},"candlestick chart basics"," cover the parts first.",[21,1626,1628],{"id":1627},"how-the-doji-types-were-tested","How the doji types were tested",[11,1630,1631],{},"The numbers in this guide come from Thomas Bulkowski, who tested 103 candlestick patterns for his Encyclopedia of Candlestick Charts, using almost 5 million candle lines across bull and bear markets. He counts a breakout as a close above the top of the candle or below its bottom, then measures whether price reversed the prior trend or continued it.",[11,1633,1634],{},"Two numbers matter for each type. The reversal or continuation rate says how often price went the way the textbook story expects. The overall performance rank, from 1 to 103, says how far price tended to travel after the breakout, with 1 being the strongest mover.",[29,1636,1637,1653],{},[32,1638,1639],{},[35,1640,1641,1644,1647,1650],{},[38,1642,1643],{},"Doji type",[38,1645,1646],{},"The story traders tell",[38,1648,1649],{},"Tested behaviour",[38,1651,1652],{},"Performance rank",[48,1654,1655,1669,1683,1696,1709],{},[35,1656,1657,1660,1663,1666],{},[53,1658,1659],{},"Any doji in an uptrend (northern)",[53,1661,1662],{},"Bearish reversal",[53,1664,1665],{},"Bullish continuation 51% of the time",[53,1667,1668],{},"83 of 103",[35,1670,1671,1674,1677,1680],{},[53,1672,1673],{},"Any doji in a downtrend (southern)",[53,1675,1676],{},"Bullish reversal",[53,1678,1679],{},"Bullish reversal 52% of the time",[53,1681,1682],{},"78 of 103",[35,1684,1685,1688,1691,1693],{},[53,1686,1687],{},"Long-legged doji",[53,1689,1690],{},"Big move coming",[53,1692,1665],{},[53,1694,1695],{},"37 of 103",[35,1697,1698,1701,1703,1706],{},[53,1699,1700],{},"Dragonfly doji",[53,1702,1676],{},[53,1704,1705],{},"Reversal 50% of the time",[53,1707,1708],{},"98 of 103",[35,1710,1711,1714,1716,1719],{},[53,1712,1713],{},"Gravestone doji",[53,1715,1662],{},[53,1717,1718],{},"Bearish reversal 51% of the time",[53,1720,1721],{},"77 of 103",[11,1723,1724,1725,1728],{},"These are measured frequencies from a past sample, not odds for the next candle you see. ",[354,1726,1727],{},"A rate of 50 to 52% is a coin flip, and Bulkowski calls it random himself."," The four-price doji is not in the table: none of the tested figures here cover it, for reasons the fifth section makes clear.",[21,1730,1732],{"id":1731},"_1-the-standard-doji-the-plain-cross","1. The standard doji: the plain cross",[11,1734,1735],{},"The standard doji has a thin body and short wicks of roughly equal length. It is the most common shape, and its traditional reading depends entirely on where it shows up.",[11,1737,1738],{},"After a rise, or after a long white candle, StockCharts reads a doji as a sign that buying pressure may be weakening. After a fall, or after a long black candle, it reads the same shape as selling pressure easing. The guide adds the condition that matters: on their own, doji are not enough to mark a reversal.",[11,1740,1741],{},"Bulkowski's tests back that caution. He calls any doji that forms in an uptrend a northern doji and any doji in a downtrend a southern doji. The northern doji is supposed to signal a bearish reversal, yet in his sample it acted as a bullish continuation 51% of the time. The southern doji is supposed to signal a bullish reversal, and it did, 52% of the time.",[11,1743,1744],{},"Both are also very common. The southern doji ranks 8th for frequency out of 103 patterns, and the northern doji ranks 6th, which Bulkowski compares to finding sand at a beach.",[11,1746,1747],{},"That is where the standard doji misleads. Because it appears so often, spotting one after a rise feels like evidence, when the tested record says the next move was close to random. Treat it as a reason to watch the next candle closely, not as a call on direction.",[21,1749,1751],{"id":1750},"_2-the-long-legged-doji-a-loud-session-that-went-nowhere","2. The long-legged doji: a loud session that went nowhere",[11,1753,1754],{},"A long-legged doji has long upper and lower shadows, often close to equal, around a body that is barely there. Price traded well above and well below the open during the session and still closed almost exactly where it started. StockCharts sums it up as a whole lot of yelling and screaming with little change at the end.",[11,1756,1757,1758,1763],{},"The usual story is that this much indecision comes before a big move. Bulkowski's verdict on the ",[280,1759,1762],{"href":1760,"target":283,"rel":1761},"https:\u002F\u002Fthepatternsite.com\u002FLongLegDoji.html",[285],"long-legged doji"," is blunt: it means nothing, and his tests show the breakout direction is random. It acted as a bullish continuation 51% of the time.",[11,1765,1766],{},"Two details are worth keeping. Its overall performance rank is 37 out of 103, the best of the doji types in his tests, so price did tend to travel after the breakout. And candles with shadows taller than the median saw price move 50% farther than those with shorter shadows.",[11,1768,1769,1770,1775],{},"So the long-legged doji says something about size and nothing about direction. That fits how Investopedia frames ",[280,1771,1774],{"href":1772,"target":283,"rel":1773},"https:\u002F\u002Fwww.investopedia.com\u002Fterms\u002Fd\u002Fdoji.asp",[285],"doji in general",": common in periods of consolidation, and useful for spotting a potential breakout, with some analysts reading reversal and others reading a pause before the trend continues. The misleading part is reading a direction into a candle whose whole message is that nobody knew the direction.",[21,1777,1779],{"id":1778},"_3-the-dragonfly-doji-the-t-that-is-supposed-to-be-bullish","3. The dragonfly doji: the T that is supposed to be bullish",[11,1781,1782],{},"The dragonfly doji forms when the open, high and close are equal and the low leaves a long lower shadow. It looks like a capital T. Sellers drove price down during the session, then buyers pushed it all the way back to the open by the close.",[11,1784,1785],{},"That recovery is why many traders read the dragonfly as bullish. StockCharts is more careful. After a long downtrend, a long black candle or at support, it could signal a potential bottom. After a long uptrend, a long white candle or at resistance, the same long lower shadow could foreshadow a top. Confirmation is required in both cases.",[11,1787,1034,1788,1793],{},[280,1789,1792],{"href":1790,"target":283,"rel":1791},"https:\u002F\u002Fwww.thepatternsite.com\u002FDragonfly.html",[285],"dragonfly doji tests"," show it reversed the prior trend 50% of the time. Its overall performance rank is 98 out of 103, near the bottom, and Bulkowski's advice is not to expect price to trend for long after one. Because the close is pinned at the top of the candle, upward breakouts were the more common outcome, which is a different thing from a reversal.",[11,1795,1796],{},"His own worked example shows the trap. A dragonfly formed at the end of an uptrend, where the long lower shadow invited a bullish read. Price broke out downward instead and gave back about half of the prior rise before climbing again.",[21,1798,1800],{"id":1799},"_4-the-gravestone-doji-the-rally-that-failed","4. The gravestone doji: the rally that failed",[11,1802,1803],{},"The gravestone doji is the dragonfly upside down. The open, low and close are equal, and the high leaves a long upper shadow. Buyers pushed price up during the session, and sellers dragged it back to the open, which was also the low, by the close.",[11,1805,1806,1807,1812],{},"It is the candle traders most often treat as a warning at the top of a run. In Bulkowski's sample the ",[280,1808,1811],{"href":1809,"target":283,"rel":1810},"https:\u002F\u002Fthepatternsite.com\u002FGravestone.html",[285],"gravestone doji"," did act as a bearish reversal, but only 51% of the time, and it ranked 77 out of 103 for performance, so price did not usually trend far after the breakout. He also found it far more often in bull markets than in bear markets, at a ratio of about 15 to 1.",[11,1814,1815],{},"His testing points to where it is least useful. He advises ignoring gravestone doji that form in congestion areas, the sideways stretches where price has no trend to reverse. Taller gravestones did better: those taller than the median saw price move about 50% farther after the breakout.",[11,1817,1818],{},"StockCharts adds the other half of the picture. The long upper shadow shows a failed rally, but the high itself shows that buyers were active. After a downtrend or at support, that buying pressure can matter more than the failure.",[21,1820,1822],{"id":1821},"_5-the-four-price-doji-the-dash","5. The four-price doji: the dash",[11,1824,1825],{},"The four-price doji is the rarest shape: open, high, low and close are all the same price. There is no body and no wick, only a short horizontal dash.",[11,1827,1828],{},"Four equal prices mean every trade in the session happened at one price. That takes either very thin trading or a bar short enough to hold a single price, so on the daily chart of a busy stock it is a rare sight.",[11,1830,1831],{},"It is also the one doji that offers nothing to read. There is no shadow to show a rejected push and no range to show a fight. It tells you that price did not move, and often that very few shares changed hands. None of the tested figures in this guide cover it, and there is little a test could measure.",[11,1833,1834],{},"The misleading part here is subtler: a run of four-price doji on a thin stock can make a chart look calm when the real story is a lack of trading. Check volume before reading anything into a dash.",[21,1836,1838],{"id":1837},"why-every-doji-needs-the-next-candle","Why every doji needs the next candle",[11,1840,1841],{},"This is the most important point in the guide. A doji is a pause, and a pause only becomes a turn when the next candles confirm it.",[11,1843,1844],{},"StockCharts spells out what confirmation looks like. After a rise and a doji, bearish confirmation could come as a gap down, a long black candle or a decline below the open of the long white candle before the doji. After a fall and a doji, bullish confirmation could come as a gap up, a long white candle or an advance above the open of the long black candle before it. Greg Morris adds a simpler test in Candlestick Charting Explained: a reversal pattern needs a prior trend to reverse.",[11,1846,1847],{},"Before reading a doji as anything more than a pause, run through a few checks:",[548,1849,1850,1853,1856,1859,1862],{},[551,1851,1852],{},"Is there a clear trend before it, or is price moving sideways?",[551,1854,1855],{},"Is the body thin compared with the candles around it, the way Nison describes?",[551,1857,1858],{},"Does it sit at a level that matters, such as support, resistance or a recent high or low?",[551,1860,1861],{},"How long are the shadows compared with recent candles?",[551,1863,1864],{},"Did the next candle close beyond the doji's high or low, and in which direction?",[11,1866,1867],{},"Bulkowski's data adds one more pattern across several types: long-legged, dragonfly, gravestone and southern doji that formed within a third of the yearly low performed best in his sample.",[21,1869,1871],{"id":1870},"a-gravestone-that-did-not-reverse","A gravestone that did not reverse",[11,1873,1874],{},"Here is how the coin flip plays out on a chart. A stock rises for a week and prints a gravestone doji right at last month's high. The bearish read is tempting. The next session opens higher and closes above the gravestone's high. By Bulkowski's definition, that is an upward breakout: the gravestone did not reverse anything. That outcome is the other 49% in his sample, and it is why a doji earns a second look rather than a decision.",[21,1876,1878],{"id":1877},"where-primo-trade-fits","Where Primo Trade fits",[11,1880,1881],{},"A doji's meaning sits in its surroundings: the trend before it, the level it formed at, the candle after it. That is the kind of context Primo Trade's Analyze Chart reads. You take a photo or screenshot of a chart, and the AI reads the trend, volatility, volume and sentiment, names the chart pattern it sees and explains what that pattern usually signals.",[11,1883,1884],{},"The result carries a signal, from Strong Buy to Strong Sell, which is the AI's directional read and not an instruction, alongside a Primo Score, a Trading Strategy and Key Insights. It works as a second opinion on a chart you are already looking at, especially when a doji has you second-guessing what you see.",[11,1886,268],{},[11,1888,1889,1890,1893,1894,1898,1899,1901],{},"The honest limit applies here too. An AI read of a chart runs into the same problem as a human one when a doji sits alone without confirmation, and a screenshot without a ticker or an axis gives it less to work with. The post on ",[280,1891,1892],{"href":541},"how AI reads a chart"," covers where that reading goes wrong. For the shapes on one page, the ",[280,1895,1897],{"href":1896},"https:\u002F\u002Fprimotrade.app\u002Fcandlestick-patterns-cheat-sheet","candlestick patterns cheat sheet"," lists every common candle, and the ",[280,1900,530],{"href":529}," covers the larger formations a doji often sits inside.",[21,1903,1905],{"id":1904},"the-short-version","The short version",[11,1907,1908],{},"A doji candlestick records a session that ended where it began. The standard, long-legged, dragonfly, gravestone and four-price types each describe how price got there, and each has a traditional story attached.",[11,1910,1911],{},"Bulkowski's tests on almost 5 million candle lines found those stories held about half the time: 50% for the dragonfly, 51% for the gravestone and the long-legged doji, 51 and 52% for doji in uptrends and downtrends. The long-legged doji showed the most movement after a breakout, without a reliable direction.",[11,1913,1914],{},"What turns a doji from noise into information is everything around it: a real prior trend, a thin body among long ones, a level that matters, and above all a next candle that closes beyond its range. Without those, a doji is a pause, and reading it as more is where it misleads.",{"title":661,"searchDepth":662,"depth":662,"links":1916},[1917,1918,1919,1920,1921,1922,1923,1924,1925,1926,1927],{"id":1593,"depth":662,"text":1594},{"id":1627,"depth":662,"text":1628},{"id":1731,"depth":662,"text":1732},{"id":1750,"depth":662,"text":1751},{"id":1778,"depth":662,"text":1779},{"id":1799,"depth":662,"text":1800},{"id":1821,"depth":662,"text":1822},{"id":1837,"depth":662,"text":1838},{"id":1870,"depth":662,"text":1871},{"id":1877,"depth":662,"text":1878},{"id":1904,"depth":662,"text":1905},"https:\u002F\u002Fmedia.primotrade.app\u002Fblog\u002Fdoji-candlestick-5-types-and-when-each-one-misleads.jpg","A doji candlestick records a standoff, not a direction. The 5 types, what testing found for each, and what has to happen next before one means anything.",{},"\u002Fblog\u002F2026\u002F10\u002Fdoji-candlestick-5-types-and-when-each-one-misleads","2026-10-15",{"title":1582,"description":1929},{"loc":1931,"images":1935},[1936],{"loc":1612},"blog\u002F2026\u002F10\u002Fdoji-candlestick-5-types-and-when-each-one-misleads","okjmE8CR_NKcWOPzzHIbXxULGc92wM7fmhN1iqE-f7Q",1794742768361]