Statistics
Below are all of the Conditional Statistics™ available throughout the Patterns ID app with a brief description of how the values are calculated.
Alpha
The alpha is calculated by taking all of the stock returns that occurred under a specific condition and comparing them against S&P 500 index returns over the same period. The alpha value is the excess stock returns when compared to their sensitivity to the broader market.
Autocorrelation
The autocorrelation is calculated by taking all of the stock returns that occurred under a specific condition and calculating their correlation with prior day’s returns. This value gives an indication of the historical 1-day momentum of returns under the condition in question.
Average
The average is calculated by taking all of the stock returns that occurred under a specific condition and calculating their average.
Beta
The beta is calculated by taking all of the stock returns that occurred under a specific condition and comparing them against S&P 500 index returns over the same period. The beta value is the stock prices’ sensitivity to the broader market and acts as a multiple of the index.
Count
The count is calculated by taking all of the stock returns that occurred under a specific condition and counting the records.
Max
The maximum is calculated by taking all of the stock returns that occurred under a specific condition and extracting the maximum value.
Median
The median is calculated by taking all of the stock returns that occurred under a specific condition and extracting the middle value after ordering the returns from low to high.
Min
The minimum is calculated by taking all of the stock returns that occurred under a specific condition and extracting the minimum value.
Positive
The positive value is calculated by taking all of the stock returns that occurred under a specific condition and counting the records that are greater than 0.
P-Value
The p-value is calculated by taking all of the stock returns that occurred under a specific condition and calculating their average.
Relative Average
The relative average is calculated by first taking all of the stock returns that occurred under a specific condition and calculating the average. Then the unconditional average is subtracted from the conditional average to get the relative average value. The relative average removes any momentum effects from the stock price history to highlight the unique performance under that specific condition.
Standard Deviation (St Dev, SD)
The standard deviation is calculated by taking all of the stock returns that occurred under a specific condition and calculating the standard deviation of those returns. The standard deviation value gives insight into the volatility of the conditional returns.
Indicators
Below are all of the indicators available throughout the Patterns ID app with a brief description of how the values are calculated.
10Y – 2Y
This indicator groups price returns based on the prior day’s 10y treasury yield less the 2y treasury yield. The 10y – 2y is the difference between the two maturities and is the most popular way to calculate the yield curve. The 10y – 2y values are ordered and grouped into quartiles so that values used for the screener are Q1 – Q4, representing the 10y – 2y quartile.
20D Volatility
This indicator groups price returns based on the prior day’s 20 day volatility of closing prices. The value is annualized using the square root of the number of trading periods in a year. The calculated volatility values are ordered and grouped into quartiles so that values used for the screener are Q1 – Q4, representing the 20D volatility quartile.
365D Price Change
This indicator groups price returns based on the prior day’s 365 day price change using closing prices. The calculated 365 day price change values are ordered and grouped into quartiles so that values used for the screener are Q1 – Q4, representing the 365 day price change quartile.
365D Price Percentile
This indicator groups price returns based on the prior day’s 365 day historical price percentile using closing prices. The historical price percentile is calculated by ranking the prior 365 calendar days’ closing prices and determining the ranked percentile of the last close. The historical percentile values are ordered and grouped into quartiles so that values used for the screener are Q1 – Q4, representing the historical-price-percentile quartile.
All
This indicator is a meta-indicator and averages the results of all Conditional Statistics™ available. There are subgroups of ‘All’ that only average across a type of indicator, e.g. ‘All Technical’ averages values across all technical indicators.
Bollinger Bands
This indicator groups price returns based on the prior day’s place in the Bollinger Bands. The Bollinger Bands are calculated using closing prices from the prior 20 periods and uses 2 standard deviations. The last closing price is compared to the Bollinger Bands and uses the following values for the screener: Above Upper, Middle Upper, Middle Lower, and Below Lower.
Chande Oscillator
This indicator groups price returns based on the prior day’s Chande Momentum Oscillator value using a 20 day period. The Chande Oscillator values are grouped based on the following ranges: 0-0.2, 0.2-0.5, 0.5-0.8, 0.8-1.0
CPI YoY
This indicator groups price returns based on the prior day’s CPI YoY change. The CPI value is calculated by taking the latest CPI value and dividing it by the CPI value from a year ago. The CPI change values are ordered and grouped into quartiles so that values used for the screener are Q1 – Q4, representing the CPI change quartile.
Current Ratio
This indicator groups price returns based on the prior day’s current ratio. The current ratio value is calculated by dividing current assets by current liabilities. The current ratio values are ordered and grouped into quartiles so that values used for the screener are Q1 – Q4, representing the current ratio quartile.
Day of Week
This indicator groups price returns based on the day of week. Values used for the screener are 1 – 5.
D/E
This indicator groups price returns based on the prior day’s debt to equity ratio. The debt to equity ratio is calculated using total debt divided by total stockholder equity. The D/E values are ordered and grouped into quartiles so that values used for the screener are Q1 – Q4, representing the D/E quartile.
Election Cycle
This indicator groups price returns based on the year relative to the presidential election cycle. Values used for the screener are 1 – 4 where 1 is an election year.
EMA
This indicator groups price returns based on the prior day’s difference between the 20 day and 50 day exponential moving average. The calculated values are ordered and grouped into quartiles so that values used for the screener are Q1 – Q4, representing the short-term – long-term SMA quartile.
GDP YoY
This indicator groups price returns based on the prior day’s quarterly GDP YoY change. The GDP rate of change value is calculated by taking the latest GDP value and dividing it by the GDP value from a year ago. The GDP change values are ordered and grouped into quartiles so that values used for the screener are Q1 – Q4, representing the GDP change quartile.
MACD
This indicator groups price returns based on the prior day’s MACD value being positive or negative. MACD is calculated using closing prices with a fast MA period of 12, slow MA period of 26, and signal MA period of 9. The final values are calculated by subtracting the signal from the difference between the fast and slow moving averages. The MACD values are grouped by positive and negative and uses the following values for the screener: Positive and Negative.
Money Flow Index
This indicator groups price returns based on the prior day’s Money Flow Index and uses a 20 day lookback period. The Money Flow Index values are grouped based on the following ranges: 0-20, 20-50, 50-80, 80-100.
Month of Year
This indicator groups price returns based on the month of year. Values used for the screener are 1 – 12.
Profit Margin
This indicator groups price returns based on the prior day’s profit margin. The profit margin value is calculated by dividing net income by total revenue. The profit margin values are ordered and grouped into quartiles so that values used for the screener are Q1 – Q4, representing the profit margin quartile.
None
Price returns are not grouped by any indicator and show the unconditional daily return statistics. It’s considered the control group and useful for comparing the differences between its values and values of other indicators.
On Balance Volume
This indicator groups price returns based on the prior day’s on balance volume using a 20 day lookback period. The calculated values are grouped by positive and negative and uses the following values for the screener: Positive and Negative.
P/S
This indicator groups price returns based on the prior day’s price to sales ratio. The P/S ratio is calculated using closing prices and dividing by the revenue per share. The P/S values are ordered and grouped into quartiles so that values used for the screener are Q1 – Q4, representing the P/S quartile.
P/B
This indicator groups price returns based on the prior day’s price to book ratio. The P/B ratio is calculated using closing prices and dividing by the book value. The book value is calculated as (total assets – total liabilities) / shares outstanding. The P/B values are ordered and grouped into quartiles so that values used for the screener are Q1 – Q4, representing the P/B quartile.
P/E
This indicator groups price returns based on the prior day’s price to earnings ratio. The P/E ratio is calculated using closing prices and compared to the latest prior 12 months EPS. The EPS calculation uses total shares outstanding plus convertibles. The P/E values are ordered and grouped into quartiles so that values used for the screener are Q1 – Q4, representing the P/E quartile.
P/EG
This indicator groups price returns based on the prior day’s price to earnings growth ratio. The P/EG ratio is calculated using closing prices and the latest prior 12 months EPS growth. The EPS growth is calculated by comparing the current EPS value with the EPS from 1 year ago. The EPS calculation uses total shares outstanding plus convertibles. The P/EG values are ordered and grouped into quartiles so that values used for the screener are Q1 – Q4, representing the P/EG quartile.
P/FCF
This indicator groups price returns based on the prior day’s price to free cash flow ratio. The P/FCF ratio is calculated using closing prices and dividing by free cash flow. The P/FCF values are ordered and grouped into quartiles so that values used for the screener are Q1 – Q4, representing the P/FCF quartile.
Post Holiday
This indicator groups price returns based on whether the period was immediately after a holiday. The holidays used are the following: New Year’s Day, Martin Luther King Jr. Day, Washington’s Birthday, Easter Sunday, Memorial Day, Juneteenth, Independence Day, Labor Day, Thanksgiving Day, and Christmas Day.
Pre Holiday
This indicator groups price returns based on whether the period was immediately before a holiday. The holidays used are the following: New Year’s Day, Martin Luther King Jr. Day, Washington’s Birthday, Easter Sunday, Memorial Day, Juneteenth, Independence Day, Labor Day, Thanksgiving Day, and Christmas Day.
Price
This indicator groups price returns based on the prior day’s nominal closing price of the security. The price history is ordered and grouped into quartiles so that values used for the screener are Q1 – Q4, representing the nominal price quartile.
ROA
This indicator groups price returns based on the prior day’s return on assets (ROA). The ROA value is calculated using net income and dividing by the total assets. The ROA values are ordered and grouped into quartiles so that values used for the screener are Q1 – Q4, representing the ROA quartile.
ROE
This indicator groups price returns based on the prior day’s return on equity (ROE). The ROE value is calculated using net income and dividing by the shareholders equity. The ROE values are ordered and grouped into quartiles so that values used for the screener are Q1 – Q4, representing the ROE quartile.
RSI
This indicator groups price returns based on the prior day’s RSI value. The RSI is calculated using close, high and low prices over the last 14 periods. The first iteration is calculated using a simple average and the Wilder approach for a smoothed average is used afterwards. The RSI values are grouped based on the following ranges: 0-0.2, 0.2-0.5, 0.5-0.8, 0.8-1.0
SMA
This indicator groups price returns based on the prior day’s difference between the 20 day and 50 day simple moving average. The calculated values are ordered and grouped into quartiles so that values used for the screener are Q1 – Q4, representing the short-term – long-term SMA quartile.
Stochastics
This indicator groups price returns based on the prior day’s stochastic value. Stochastics are calculated using closing prices over the last 14 periods. The stochastic value is notably the %K value. The stochastic values are grouped based on the following ranges: 0-0.2, 0.2-0.5, 0.5-0.8, 0.8-1.0.
UM Sentiment
This indicator groups price returns based on the prior day’s University of Michigan Consumer Sentiment (UMCS) survey value. The UMCS values are ordered and grouped into quartiles so that values used for the screener are Q1 – Q4, representing the UMCS quartile.
Unemployment
This indicator groups price returns based on the prior day’s unemployment rate. The unemployment rate values are ordered and grouped into quartiles so that values used for the screener are Q1 – Q4, representing the unemployment rate quartile.
Volume
This indicator groups price returns based on the prior day’s trading volume. The price history is ordered and grouped into quartiles so that values used for the screener are Q1 – Q4, representing the trading volume quartile.
Williams %R
This indicator groups price returns based on the prior day’s Williams %R value and uses a 20 day lookback period. The Williams %R values are grouped based on the following ranges: -100 to -80, -80 to -50, -50 to -20, and -20 to 0.