Strength Indicator

The goal of this project is to implement a Strength Indicator for a basket of stocks. The SI will be calculated using a Machine Learning model that has evaluated previous prices for that stock.

The ideal situation is to have a single generic model that can be used for all stocks. However, it is likely that different stocks will require different features and tuning to achieve optimal performance.

The Strength Indicator will be displayed as a score from -100 to 100, with positive values indicating bullish strength and negative values indicating bearish strength. The data will be updated each minute of the trading day. The data will be available via an API endpoint for use in other applications. The data will also be displayed via a dashboard on this website.

The Strength Indicator model will be trained on 6 months of historical single minute price data. Each tick of data will include the following attributes:

  • Stock
  • Open Price
  • High Price
  • Low Price
  • Close Price
  • Body Size (Close - Open)
  • Head Size (High - max(Close, Open))
  • Tail Size (min(Close, Open) - Low)
  • Volume
  • 50 Minute Moving Average
  • 100 Minute Moving Average
  • 50/100 Moving Average Delta
  • Minute Of the Day
  • Minute Of the Week
  • Week of the Year
  • Minutes till Options Expiration
  • Maximum High over last 1000 minutes
  • Maximum Low over last 1000 minutes
  • Average Volume over last 1000 minutes


stock open high low close body head tail volume 50ma 100ma madelta minday minweek weekofyear minoptexp maxhigh maxlow
AAPL 10 14 10 13 5 2 1 100 20 15 5 200 1503 23 400 12 10
AAPL 10 14 10 13 5 2 1 100 20 15 5 200 1503 23 400 12 10
AAPL 10 14 10 13 5 2 1 100 20 15 5 200 1503 23 400 12 10
AAPL 10 14 10 13 5 2 1 100 20 15 5 200 1503 23 400 12 10
AAPL 10 14 10 13 5 2 1 100 20 15 5 200 1503 23 400 12 10
AAPL 10 14 10 13 5 2 1 100 20 15 5 200 1503 23 400 12 10
AAPL 10 14 10 13 5 2 1 100 20 15 5 200 1503 23 400 12 10


Data Size
100 Bytes Per Tick
390 Ticks Per Day
5 Trading Days Per Week
24 Weeks Model Window
46,800 Ticks in Model Window
5 MB Estimate

The SI value will be a prediction of what is going to happen in the next 2 hours for that stock. The manual calculation of the SI is as follows:

Closing Price in 2 hours - Current Open Price = Price Delta

Look back over last 1000 minutes and calculate the Maximum Bullish Delta and Maximum Bearish Delta.

If Price Delta is positive:

Strength Indicator = (Price Delta) / (Maximum Bullish Delta) * 100

else

Strength Indicator = (Price Delta) / (Maximum Bearish Delta) * 100

An unhandled error has occurred. Reload 🗙