Get started

Returnsanalysis

Historical Returns [BigBeluga]🔵 OVERVIEW The Historical Returns indicator visualizes daily and monthly return data to help traders assess seasonal performance and volatility behavior. It provides a clean and informative dashboard showing the current month’s daily return bubbles, monthly return curves, and a snapshot of the current month and year performance. This tool is ideal for spotting recurring return patterns and understanding the broader profitability context of a symbol. 🔵 CONCEPTS Daily Return Bubbles: Each trading day is analyzed for its return percentage, and plotted as a bubble with size proportional to the return magnitude. Monthly Performance Curves: Average or cumulative returns are calculated and plotted to show how the current month is performing relative to historical averages. Current Year Return: Current year performance as a single return value, giving traders context on long-term profitability. Current Month Average Return: Current month average performance as a single return value, giving traders context on short-term profitability. Extreme Return Labels: Optionally highlights daily returns above +4% or below -4% with labeled percentages for spike recognition. 🔵 FEATURES Shows daily return bubbles (1%–7%+), color-coded by direction. Labels monthly returns with the month name and percentage value. Displays a performance dashboard with: Daily return heatmap for the current month. Average return for the current month. Year-to-date return. Toggle between average and cumulative modes for monthly return curves. Clearly marks days with abnormal return spikes using optional labels. Clean fallback warning if not on a daily chart ("⚠️USE DAILY TIMEFRAME"). Custom color themes for bullish and bearish values. 🔵 HOW TO USE Use the monthly return curve to compare how the current month is performing against historical averages. Look for clusters of positive or negative bubbles as signals of strong directional weeks. Watch extreme return labels for volatility spikes or catalyst days. Use year-to-date return to assess how the asset is trending in the broader macro cycle. Combine with other BigBeluga tools to align trades with historically favorable periods. 🔵 CONCLUSION Historical Returns is your visual companion for return analytics — helping you identify profitable months, detect volatility surges, and understand historical seasonality at a glance. With a clean dashboard and insightful overlays, this tool supports better timing and improved statistical edge in both short- and long-term trades.
Pine Script® indicator
by BigBeluga
3
Correlation Heatmap█ OVERVIEW This indicator creates a correlation matrix for a user-specified list of symbols based on their time-aligned weekly or monthly price returns. It calculates the Pearson correlation coefficient for each possible symbol pair, and it displays the results in a symmetric table with heatmap-colored cells. This format provides an intuitive view of the linear relationships between various symbols' price movements over a specific time range. █ CONCEPTS Correlation Correlation typically refers to an observable statistical relationship between two datasets. In a financial time series context, it usually represents the extent to which sampled values from a pair of datasets, such as two series of price returns, vary jointly over time. More specifically, in this context, correlation describes the strength and direction of the relationship between the samples from both series. If two separate time series tend to rise and fall together proportionally, they might be highly correlated. Likewise, if the series often vary in opposite directions, they might have a strong anticorrelation . If the two series do not exhibit a clear relationship, they might be uncorrelated . Traders frequently analyze asset correlations to help optimize portfolios, assess market behaviors, identify potential risks, and support trading decisions. For instance, correlation often plays a key role in diversification . When two instruments exhibit a strong correlation in their returns, it might indicate that buying or selling both carries elevated unsystematic risk . Therefore, traders often aim to create balanced portfolios of relatively uncorrelated or anticorrelated assets to help promote investment diversity and potentially offset some of the risks. When using correlation analysis to support investment decisions, it is crucial to understand the following caveats:  • Correlation does not imply causation . Two assets might vary jointly over an analyzed range, resulting in high correlation or anticorrelation in their returns, but that does not indicate that either instrument directly influences the other. Joint variability between assets might occur because of shared sensitivities to external factors, such as interest rates or global sentiment, or it might be entirely coincidental. In other words, correlation does not provide sufficient information to identify cause-and-effect relationships.  • Correlation does not predict the future relationship between two assets. It only reflects the estimated strength and direction of the relationship between the current analyzed samples. Financial time series are ever-changing. A strong trend between two assets can weaken or reverse in the future. Correlation coefficient A correlation coefficient is a numeric measure of correlation. Several coefficients exist, each quantifying different types of relationships between two datasets. The most common and widely known measure is the Pearson product-moment correlation coefficient , also known as the Pearson correlation coefficient or Pearson's r . Usually, when the term "correlation coefficient" is used without context, it refers to this correlation measure. The Pearson correlation coefficient quantifies the strength and direction of the linear relationship between two variables. In other words, it indicates how consistently variables' values move together or in opposite directions in a proportional, linear manner. Its formula is as follows: 𝑟(𝑥, 𝑦) = cov(𝑥, 𝑦) / (𝜎𝑥 * 𝜎𝑦) Where:  • 𝑥 is the first variable, and 𝑦 is the second variable.  • cov(𝑥, 𝑦) is the covariance between 𝑥 and 𝑦.  • 𝜎𝑥 is the standard deviation of 𝑥.  • 𝜎𝑦 is the standard deviation of 𝑦. In essence, the correlation coefficient measures the covariance between two variables, normalized by the product of their standard deviations. The coefficient's value ranges from -1 to 1, allowing a more straightforward interpretation of the relationship between two datasets than what covariance alone provides:  • A value of 1 indicates a perfect positive correlation over the analyzed sample. As one variable's value changes, the other variable's value changes proportionally in the same direction .  • A value of -1 indicates a perfect negative correlation (anticorrelation). As one variable's value increases, the other variable's value decreases proportionally.  • A value of 0 indicates no linear relationship between the variables over the analyzed sample. Aligning returns across instruments In a financial time series, each data point (i.e., bar) in a sample represents information collected in periodic intervals. For instance, on a "1D" chart, bars form at specific times as successive days elapse. However, the times of the data points for a symbol's standard dataset depend on its active sessions , and sessions vary across instrument types. For example, the daily session for NYSE stocks is 09:30 - 16:00 UTC-4/-5 on weekdays, Forex instruments have 24-hour sessions that span from 17:00 UTC-4/-5 on one weekday to 17:00 on the next, and new daily sessions for cryptocurrencies start at 00:00 UTC every day because crypto markets are consistently open. Therefore, comparing the standard datasets for different asset types to identify correlations presents a challenge. If two symbols' datasets have bars that form at unaligned times, their correlation coefficient does not accurately describe their relationship. When calculating correlations between the returns for two assets, both datasets must maintain consistent time alignment in their values and cover identical ranges for meaningful results. To address the issue of time alignment across instruments, this indicator requests confirmed weekly or monthly data from spread tickers constructed from the chart's ticker and another specified ticker. The datasets for spreads are derived from lower-timeframe data to ensure the values from all symbols come from aligned points in time, allowing a fair comparison between different instrument types. Additionally, each spread ticker ID includes necessary modifiers, such as extended hours and adjustments. In this indicator, we use the following process to retrieve time-aligned returns for correlation calculations:  1. Request the current and previous prices from a spread representing the sum of the chart symbol and another symbol ( "chartSymbol + anotherSymbol" ).  2. Request the prices from another spread representing the difference between the two symbols ( "chartSymbol - anotherSymbol" ).  3. Calculate half of the difference between the values from both spreads ( 0.5 * (requestedSum - requestedDifference) ). The results represent the symbol's prices at times aligned with the sample points on the current chart.  4. Calculate the arithmetic return of the retrieved prices: (currentPrice - previousPrice) / previousPrice  5. Repeat steps 1-4 for each symbol requiring analysis. It's crucial to note that because this process retrieves prices for a symbol at times consistent with periodic points on the current chart, the values can represent prices from before or after the closing time of the symbol's usual session. Additionally, note that the maximum number of weeks or months in the correlation calculations depends on the chart's range and the largest time range common to all the requested symbols. To maximize the amount of data available for the calculations, we recommend setting the chart to use a daily or higher timeframe and specifying a chart symbol that covers a sufficient time range for your needs. █ FEATURES This indicator analyzes the correlations between several pairs of user-specified symbols to provide a structured, intuitive view of the relationships in their returns. Below are the indicator's key features: Requesting a list of securities The "Symbol list" text box in the indicator's "Settings/Inputs" tab accepts a comma-separated list of symbols or ticker identifiers with optional spaces (e.g., "XOM, MSFT, BITSTAMP:BTCUSD"). The indicator dynamically requests returns for each symbol in the list, then calculates the correlation between each pair of return series for its heatmap display. Each item in the list must represent a valid symbol or ticker ID. If the list includes an invalid symbol, the script raises a runtime error. To specify a broker/exchange for a symbol, include its name as a prefix with a colon in the "EXCHANGE:SYMBOL" format. If a symbol in the list does not specify an exchange prefix, the indicator selects the most commonly used exchange when requesting the data. Note that the number of symbols allowed in the list depends on the user's plan. Users with non-professional plans can compare up to 20 symbols with this indicator, and users with professional plans can compare up to 32 symbols. Timeframe and data length selection The "Returns timeframe" input specifies whether the indicator uses weekly or monthly returns in its calculations. By default, its value is "1M", meaning the indicator analyzes monthly returns. Note that this script requires a chart timeframe lower than or equal to "1M". If the chart uses a higher timeframe, it causes a runtime error. To customize the length of the data used in the correlation calculations, use the "Max periods" input. When enabled, the indicator limits the calculation window to the number of periods specified in the input field. Otherwise, it uses the chart's time range as the limit. The top-left corner of the table shows the number of confirmed weeks or months used in the calculations. It's important to note that the number of confirmed periods in the correlation calculations is limited to the largest time range common to all the requested datasets, because a meaningful correlation matrix requires analyzing each symbol's returns under the same market conditions. Therefore, the correlation matrix can show different results for the same symbol pair if another listed symbol restricts the aligned data to a shorter time range. Heatmap display This indicator displays the correlations for each symbol pair in a heatmap-styled table representing a symmetric correlation matrix. Each row and column corresponds to a specific symbol, and the cells at their intersections correspond to symbol pairs . For example, the cell at the "AAPL" row and "MSFT" column shows the weekly or monthly correlation between those two symbols' returns. Likewise, the cell at the "MSFT" row and "AAPL" column shows the same value. Note that the main diagonal cells in the display, where the row and column refer to the same symbol, all show a value of 1 because any series of non-na data is always perfectly correlated with itself. The background of each correlation cell uses a gradient color based on the correlation value. By default, the gradient uses blue hues for positive correlation, orange hues for negative correlation, and white for no correlation. The intensity of each blue or orange hue corresponds to the strength of the measured correlation or anticorrelation. Users can customize the gradient's base colors using the inputs in the "Color gradient" section of the "Settings/Inputs" tab. █ FOR Pine Script® CODERS  • This script uses the `getArrayFromString()` function from our ValueAtTime library to process the input list of symbols. The function splits the "string" value by its commas, then constructs an array of non-empty strings without leading or trailing whitespaces. Additionally, it uses the str.upper() function to convert each symbol's characters to uppercase.  • The script's `getAlignedReturns()` function requests time-aligned prices with two request.security() calls that use spread tickers based on the chart's symbol and another symbol. Then, it calculates the arithmetic return using the `changePercent()` function from the ta library. The `collectReturns()` function uses `getAlignedReturns()` within a loop and stores the data from each call within a matrix . The script calls the `arrayCorrelation()` function on pairs of rows from the returned matrix to calculate the correlation values.  • For consistency, the `getAlignedReturns()` function includes extended hours and dividend adjustment modifiers in its data requests. Additionally, it includes other settings inherited from the chart's context, such as "settlement-as-close" preferences. • A Pine script can execute up to 40 or 64 unique `request.*()` function calls, depending on the user's plan. The maximum number of symbols this script compares is half the plan's limit, because `getAlignedReturns()` uses two request.security() calls.  • This script can use the request.security() function within a loop because all scripts in Pine v6 enable dynamic requests by default. Refer to the Dynamic requests section of the Other timeframes and data page to learn more about this feature, and see our v6 migration guide to learn what's new in Pine v6.  • The script's table uses two distinct color.from_gradient() calls in a switch structure to determine the cell colors for positive and negative correlation values. One call calculates the color for values from -1 to 0 based on the first and second input colors, and the other calculates the colors for values from 0 to 1 based on the second and third input colors. Look first. Then leap.
Pine Script® indicator
by TradingView
11