MesoTrader Metrics Methodology

Analytics Overview

MesoTrader analyzes how a stock behaves around market-moving events using a structured event study framework. Each event is evaluated through a configurable time window that includes:

  • A sample period to establish baseline price behavior and volatility
  • An event window, broken into three phases:
    • Pre-Event Phase – the lead-up to the event
    • Event Day – the day of the announcement
    • Post-Event Phase – the reaction period that follows

The number of days in each phase—as well as the length of the sample period—is configurable per event study. This flexibility allows us to adapt to different event types, detect trend shifts, and apply contextual filters that surface repeatable patterns and opportunities.

Each metric is calculated at the event level, then aggregated across multiple events to reveal historical patterns, including:

  • Frequency of positive or negative stock moves after events.
  • Changes in volatility around events.
  • Observed market behavior around specific event types

Additionally, the analytics engine supports:

  • Multiple event types (earnings, FOMC meetings, CPI releases, and more).
  • Contextual filtering based on historical outcomes, macro trends, and market behavior.
  • Performance-based and sector-level aggregation (e.g., identifying top-10 stocks by average returns, volatility, or sector performance).

Earnings Strategy Metrics (Updated V2 Output)

Summary Information

Each analysis begins with a strategy summary embed at the top of the bot’s response. This includes:

  • Company Name and Ticker: Identifies the analyzed stock.
  • Period: Start and end dates of the included earnings announcements. Default is the 8 most recent events and optional filters may be used to set a to and from date to expand the range.
  • Number of Events: How many earnings events are included in the analysis.
  • Estimated Next Earnings: The estimated date and time (Morning, Afternoon, or Timing TBD) of the company’s next earnings call.
  • Return Period: Fixed as the 3-day period after each earnings announcement.
  • Outcomes and Pre-Event Trend: Currently listed as "Unfiltered" (future versions may apply filters here).

Data Quality Warning

If the bot detects data anomalies, a Data Quality Warning is displayed above the strategy summary. This occurs if two or more earnings announcements are 30 calendar days or fewer apart, or if there is a gap of 120 days or more between consecutive events. These flags indicate possible duplicates, backdated entries, or missing events that could affect reliability.

Performance Summary

The bot summarizes cumulative strategy performance across the 3-day post-earnings return window. This includes:

  • Total Cumulative Return: The compounded return across all included earnings events.
  • Latest YoY Return: Rolling return over the most recent 12 months of earnings data.
  • Latest Annual Returns: Calendar-year returns for up to three years based on available event dates. If only partial years are available (e.g. year-to-date), this is noted in the label.

Trade Performance Metrics

This section provides stats based on individual 3-day post-earnings returns:

  • Hit Rate: Percentage of earnings events where the return was greater than 0%.
  • Average Gain: Average return across only the earnings events with positive returns.
  • Average Loss: Average return across only the events with zero or negative returns.
  • Profit Factor: Sum of gains divided by sum of losses (excluding zero returns).
  • Risk Reward Ratio: Average gain divided by average loss (excluding zero returns).

Important Clarifications

  • The 3-day return period always starts at market close on the earnings announcement date and ends at the close three trading days later.
  • Profit Factor and Risk Reward Ratio exclude flat (0%) returns from the denominator to avoid division errors.
  • The "Latest YoY Return" is a rolling figure—not tied to a calendar year—and reflects returns during the most recent 12-month period covered by the included earnings events.
  • Results are purely descriptive. They summarize past behavior and do not constitute predictions or trading advice.

Estimated Upcoming Earnings

This command displays estimated future earnings announcement dates for U.S.-listed common stocks.

The bot groups upcoming earnings by calendar date and displays each entry with the following information:

  • Ticker: The stock symbol in bold.
  • Timing: The expected time of the earnings call—either Morning, Afternoon, or Timing TBD.
  • Company Name: The full name of the reporting company.

At the top of the response, the bot also displays a summary of the search context:

  • Date Range: The range of earnings dates included in the result.
  • Market Cap: The selected market cap filter, or “All” if none was applied.
  • Sector/Industry: The selected sector and industry filters, or “All/All” if none were applied.
  • Ticker: The selected ticker filter, or “All” if none was applied.

If more than 50 total events or 10 distinct dates are found, the response is capped and a message is included to indicate that only partial results are shown.

Important Clarifications

  • The dates shown represent estimated earnings call dates, not guaranteed press release times.
  • The timing labels are based on expected call times, not exact start times or SEC filing timestamps.
  • Only U.S.-based common stocks are included. ADRs, ETFs, and foreign listings are excluded.
  • Company tickers containing a period (e.g., BRK.B) are displayed with a hyphen instead (e.g., BRK-B).
  • If no earnings are scheduled within the search range, the bot returns an empty result set.
  • If more than 50 upcoming earnings events or more than 10 distinct dates match the filters, the result set is trimmed and a warning is shown.

Data Requirements and Filters

The bot applies specific data requirements and filtering criteria before including any earnings announcement in the analysis:

  • Minimum Required Events:
    • At least 2 earnings announcements meeting pricing data requirements (as detailed below) are required to display any metrics.
    • The EPS surprise groups require a minimum of 4 earnings announcements where both actual and expected EPS data are available.
  • Number of Events:
    • Demo/Free access metrics are calculated from a maximum of the 8 most recent quarterly earnings announcements per ticker. All other usage will default to the 8 most recent announcements and will allow to and from filters to be used to expand the events analyzed.
    • These announcements may not be contiguous due to company-specific reporting irregularities or delisting status.
  • Ticker Selection (U.S. Stocks Only):
    • The bot analyzes only U.S.-based common stocks, excluding ADRs.
    • Specifically, tickers must meet these criteria:
      • Type: Common Stock
      • Home Category: Domestic, Domestic Primary, Domestic Secondary, or undefined (only if a GICS Sector classification is provided).
  • Pricing Data Requirements:
    • Each earnings announcement must have available Close price data for:
      • The Pre-Event Window: 25 trading days immediately preceding the event window.
      • The Event Window: the full 6-trading-day period (2 days before the event, the event day itself, and 3 days after).
    • Close price data must be available for the event day itself; otherwise, that earnings announcement is excluded from the analysis.
    • To account for weekends and market holidays, typically, any two-week period within the required windows should contain at least 8 trading days.
  • Market Holidays and Weekend Adjustments:
    • Earnings announcements occurring on market holidays (e.g., Good Friday) or weekends for selected exceptions (BRK-A, BRK-B) are adjusted to the following trading day (typically the next Monday).
  • IPO Date Filter:
    • Only earnings announcements occurring on or after the stated IPO date for each ticker are included in the analysis.
    • If no IPO date is available, this filter is not applied.
  • Recent Data Cutoff:
    • Because metrics include a 3-day period following the earnings announcement, events generally need to have occurred at least 3 trading days before the most recent data refresh to be included.
  • EPS and Price Data Completeness:
    • Earnings announcements missing either the reported or estimated EPS data (or both) are excluded from EPS-dependent metrics and grouped under "Missing EPS Data."
    • Price data is split-adjusted; dividends are not included.
  • EPS Definition:
    • EPS values shown are Adjusted EPS.
  • Earnings Announcement Date Definition:
    • The bot uses the earnings call date when a company conducts an earnings call.
    • If no earnings call occurs, the announcement date is the date of the SEC filing or press release announcing earnings results.
  • 3-Day Post-Earnings Return Definition:
    • The 3-day post-earnings return is calculated as the cumulative daily compounded return from market close on the earnings announcement date through market close three trading days later.

Known Limitations

To the best of our knowledge, the following limitations apply to the metrics and analysis provided by the bot (these are not necessarily all-inclusive):

  • After-Hours and Pre-Market Activity: The bot does not incorporate pre-market or after-hours trading activity, nor does it include volume or options data.
  • Reported Date Ambiguity: The "Reported Date" return can reflect either the trading day before or after the earnings announcement, depending on whether the announcement occurred before market open or after market close.
  • Ticker Symbol Formatting: Ticker symbols containing a period (.) are reformatted to use a hyphen (-) instead. Example: BRK.B is displayed as BRK-B.
  • Earnings Announcement Date Source: The bot relies on third-party data providers to determine the earnings announcement date. Providers may mix earnings call dates, SEC filing dates, and press-release dates, potentially creating inconsistencies in how the announcement date is defined.
  • Small Sample Sizes: The bot analyzes up to the 8 most recent quarterly earnings announcements per ticker. Such small sample sizes limit statistical significance and reliability of conclusions drawn.
  • Price Data Adjustments: Metrics use split-adjusted prices, but dividend adjustments are not included. If comparing these results with other sources or calculations that include dividend adjustments, differences may be observed.
  • Historical Data Only (Not Predictive): All metrics presented by the bot are purely historical observations. They do not constitute investment advice, nor are they indicative of future stock performance. Past behavior may not repeat in future events.

Copyright 2025 MesoTrader, Inc. All rights reserved.

Terms of Service

Privacy Policy