You are on a call with a friend and you are both looking at the same stock. Your 50-day moving average reads ₹412; theirs reads ₹409. Your chart shows a large gap in March; theirs shows a smooth advance through the same week. Your RSI says 63, theirs says 58. One of you must have a broken app.
Neither of you does. A chart is not a photograph of the market — it is constructed, and there are half a dozen decisions in the construction that different platforms make differently. Knowing which ones they are takes about ten minutes and saves a great deal of pointless argument.
Six reasons two charts disagree
| The difference | How it shows up | Does it matter? |
|---|---|---|
| Which exchange the data came from | A stock listed on both the NSE and the BSE has two sets of prices and two sets of volume. The prices are usually within a whisker of each other; the volume often is not, because most of the turnover sits on one exchange. | Prices, rarely. Volume, enormously — a chart showing you one exchange’s volume for a stock that trades mostly on the other is understating participation by a large multiple. |
| Whether the series is adjusted for corporate actions | A 1:1 bonus issue halves the share price overnight. An adjusted chart restates the history so the line is continuous; an unadjusted one shows a 50% cliff. | Critically. Every indicator computed across an unadjusted split or bonus is nonsense for as long as its lookback window straddles the event. |
| How much history the app loaded | An exponential moving average and an RSI never fully forget their starting point. Compute a 50-EMA from 200 candles and from 5,000 candles and the current values differ slightly. | A rupee or two on a moving average. Irrelevant for reading a trend; relevant if your stop is placed at a precise indicator value. |
| Which smoothing the indicator uses | RSI computed with the original Wilder smoothing and RSI computed with a simple average of gains and losses are different numbers with the same name. Platforms do not always say which they use. | Enough to move a reading across the 70 line. If your rule contains a threshold, the smoothing method is part of the rule whether you specified it or not. |
| Where the intraday candles are anchored | An hourly chart anchored to the clock gives a stub first bar covering 9.15 to 10.00 and then full bars from 10.00; one anchored to the session start runs 9.15 to 10.15 and 10.15 to 11.15. Every bar after the first covers a different stretch of the day, so the candles look similar and their highs, lows and closes are not the same. | For intraday patterns and any rule read off a specific candle, yes. It is also why two people can disagree about whether an opening-range breakout happened. |
| Whether you are looking at real prices at all | Heikin-Ashi, Renko and similar chart types display synthetic prices — values calculated from the real ones rather than transacted. A Heikin-Ashi open is an average, and no trade ever happened there. | Yes. These charts are legitimate for reading trend, but a stop placed off a synthetic price sits at a level that does not exist in the order book. |
The one that silently ruins backtests
Corporate actions deserve their own paragraph because they cause the most damage and are the easiest to miss. When a company splits its shares five for one, or issues a 1:1 bonus, the share price falls proportionally overnight and nobody has lost anything — the same ownership is simply divided into more pieces.
A properly adjusted price series restates all the earlier prices so the history is comparable. An unadjusted one leaves a cliff. Run a moving-average system across that cliff and it generates a spectacular sell signal on a day when nothing happened; run a backtest across a decade of Indian midcaps without adjustment and you will accumulate dozens of these phantom events.
A five-minute audit of your own chart
- 1Find a stock that had a bonus or split and look at that week
Corporate action histories are published on the NSE and BSE websites. Open the stock in your app around that date. If there is a cliff, your series is unadjusted and every long-lookback indicator on it is unreliable across that point.
- 2Compare one day’s volume with the exchange
Pick yesterday, note the volume your app shows, and check the same figure on the exchange site. A large mismatch usually means you are looking at a single exchange, or at a delayed or sampled feed.
- 3Check where your intraday candles start and end
Open an hourly chart and read the timestamps. Bars running 9.15 to 10.15 are anchored to the session; a stub bar to 10.00 followed by full clock hours is anchored to the clock. Neither is wrong, and any rule that reads a specific bar — an opening range, the first hour’s high — means something different under each.
- 4Write your indicator settings down
Period, price input, smoothing method, and the exact chart type. These are the parameters of your system, not preferences. A rule that says "RSI below 30" without them is not a rule anyone — including you, next year — can reproduce.
- 5Note the data source in the same place
Which app, which exchange, adjusted or not. Data integrity problems are almost always invisible: they do not throw an error, they quietly produce a slightly different answer, and the first sign of trouble is a backtest you cannot reproduce live.
- A two-rupee difference in a moving average between platforms.
- Slightly different colours, candle widths or default lookbacks.
- Small mismatches in the last traded price between two exchanges.
- Your friend’s RSI reading three points away from yours, when neither of you trades a threshold.
- An unadjusted series across a split, bonus or demerger.
- Volume from an exchange where the stock barely trades.
- A stop or alert placed off a Heikin-Ashi or Renko price.
- A written rule with a numeric threshold and no smoothing method specified.
- A backtest whose data source you cannot describe in one sentence.
Your chart shows a 50% single-day fall in a stock last year, but the news archive shows nothing and the company announced a 1:1 bonus that week. What has happened?
Aap aur dost dono ek hi stock dekh rahe ho — aapka 50-day average ₹412, unka ₹409, aur March ka gap unke chart pe hai hi nahi. Koi jhooth nahi bol raha: do alag app, do alag data source, do alag setting. Aankh se trend dekhna hai toh is farak se kuch nahi bigadta. Par stop exact price pe lagana hai, ya backtest karna hai — tab yehi chhota farak poora natija badal deta hai.
- Charts are constructed, and platforms make the construction choices differently — exchange source, adjustment, history length, smoothing, candle anchoring and chart type.
- Most of it changes a number without changing the story — but an unadjusted corporate action and a synthetic chart type change the picture itself.
- Unadjusted corporate actions are the most damaging and the easiest to miss, and Indian bonuses and splits are frequent.
- A stop or alert placed off a synthetic price sits at a level that never traded.
- Write the data source, adjustment setting and indicator parameters into your system — an unspecified chart is an unreproducible result.
Mark it done to track your progress through the curriculum.
Common questions
Short, direct answers to what people ask about this topic.
- why do two apps show different charts for the same stock
- Because a chart is constructed rather than photographed, and platforms make the construction choices differently. The six that account for almost all disagreements are which exchange the data came from, whether the series is adjusted for corporate actions, how much history the app loaded, which smoothing an indicator uses, where intraday candles are anchored, and whether the chart type shows real or calculated prices.
- why does my chart show a different price after a bonus issue
- An adjusted price series restates the older prices so that a split, bonus or similar corporate action does not leave an artificial cliff in the history. After a 1:1 bonus the share price roughly halves overnight while nobody has lost anything, so an unadjusted chart shows a crash that never happened. Any indicator whose lookback window crosses an unadjusted event is computing across two incompatible price scales.
- a chart type that displays calculated rather than traded prices is called
- A synthetic price chart — Heikin-Ashi and Renko are the common examples. A Heikin-Ashi open is an average derived from the previous candle, so no trade ever occurred at that price. They are perfectly legitimate for reading trend, but a stop or an alert placed off one sits at a level that does not exist in the order book.
- how do I check if my chart is adjusted for a bonus or split
- Find a stock that had a bonus or split — corporate action histories are published free on the NSE and BSE websites — and open that week on your own chart. A continuous line means the series is adjusted; a sudden cliff on a day with no news means it is not, and every long-lookback indicator is unreliable across that point.
- does a 1:1 bonus issue halve the share price
- Broadly yes. In a 1:1 bonus every holder receives one additional share for each one held, so the share count doubles and the price adjusts to roughly half, leaving the value of the holding unchanged. The exchanges apply the adjustment on the ex-bonus date, and a properly adjusted price series restates the earlier history so the chart stays continuous through it.