MACD starts with the difference between a faster and a slower moving average: price averages recalculated as new observations arrive. Averaging that difference produces the signal line. Columns called the histogram then show the distance between MACD and its signal line.
Jana turns on MACD in TradingView and sees two lines and a set of columns below the bitcoin price. First, she needs to separate what they represent. The main MACD line compares two price averages. The signal line averages MACD itself, and the histogram subtracts the signal line from the main line. The columns are therefore not the volume of bitcoin bought. They show a different subtraction from the main line, which is why the two can point in opposite directions. [1]
Fidelity describes a common setup: a fast EMA of length 12, a slow one of 26 and a signal EMA of 9. An EMA gives more weight to recent observations, so the faster version usually responds earlier to a change. The lengths refer to the chosen time intervals. On an hourly chart, they do not mean days. Jana is choosing a view of past price movement, not just the shape of a line; default settings are no assurance that those lengths will produce the best result. [2]
TradingView distinguishes two comparisons. A positive MACD means the faster price average is above the slower one. A positive histogram means MACD is above its signal line. Jana can therefore see the main line above zero while the columns are below it. The calculation is consistent: it compares different pairs of values. A single snapshot also shows a state, not a crossover. To identify a crossing, she must know the lines’ previous order too. [1]
If the chart measures price in USD/BTC, differences between its price averages have that unit too. MACD is not a percentage, a probability or profit in an account. Unlike RSI, it has no fixed bounded scale. Raw magnitudes cannot simply be compared across assets with very different prices or across currencies. Fidelity also notes that MACD is not normally an overbought or oversold gauge. A high line by itself does not mean bitcoin is too expensive. [1] [2]
Jana opens another platform and the same MACD label gives a different result. That need not be an error. TradingView lets users choose EMA or SMA separately for the price averages and the signal. SMA gives equal weight to observations in its window, while EMA emphasizes newer ones. The price source and timeframe setting, meaning the time interval, matter too. A comparison needs the same market, price pair, lengths, average types and available history, as well as agreement on whether the latest interval has closed. [1]
Fidelity warns that MACD can repeatedly cross its signal line in a market moving within a range. Someone trading every crossover may keep buying and selling as the price moves back and forth. Divergence, a disagreement between the movement of price and the indicator, also needs careful interpretation. Price might make a higher peak while the indicator makes a lower one. That describes what has happened; it does not by itself tell you when a reversal will come, or whether it will come at all. [2]
Jana receives a crossover alert while a candle is still open. A candle summarizes price over a chosen interval, and its latest price is still changing. TradingView explicitly notes that MACD also changes during this time. The lines may return to their previous order before the interval ends. The historical chart then shows the confirmed final result, not every crossover seen along the way. When assessing an alert, it matters whether it could trigger during the candle or only after the candle closed. [3]
TradingView’s example MACD Strategy enters long, a position that benefits from a rise, when the histogram turns from negative to positive. It switches to short, a position that benefits from a fall, on the opposite change. This is an added rule, not the indicator itself; shorting is also different from selling bitcoin you own. A test must include fees and slippage, the difference between expected and actual execution price. Jana must also guard against overfitting: tuning to past data in a way that fails on other data. Even a successful test promises no future result. [4] [5]
A positive line and negative columns
In an invented snapshot of already calculated averages, the fast EMA is 102000 and the slow EMA is 101500 USD/BTC. Their difference is MACD = 102000 − 101500 = 500 USD/BTC. The signal line is 600 USD/BTC, so histogram = 500 − 600 = −100 USD/BTC. The fast EMA is above the slow one, but the main MACD line is below its signal. Both statements can be true at once. The snapshot does not establish a crossover: without previous values, we cannot know whether the order has just changed. The averages are supplied inputs here, not a full calculation from prices or a real trading offer.
For the clearest picture, read this entry together with Exponential moving average, Moving average, Relative Strength Index, Divergence, Backtesting, Technical indicator. The reverse links also lead from Exponential moving average, Stochastic oscillator, Ichimoku Cloud, Fibonacci Retracement.
01Does a positive MACD mean I should buy?+
A positive MACD says that the faster average is above the slower one. It does not say whether price will rise from here or how much money it is appropriate to put at risk. The histogram can still be negative. A trade requires a separate rule and its evaluation, not just the color or sign of a line.
02Why does another platform show a different MACD?+
Start with the market and price pair. Then compare the price input, timeframe, lengths and EMA or SMA choices for both the price averages and signal. Also check available history and whether the latest candle has closed. The same abbreviation does not mean both platforms are calculating from the same observations.