- we will add a counter for price if the price crosses the target/stoploss 3 times in 15 min interval it is a confirm target/stoploss hit as the price retained at the level for 45 mins
- LLD - when the cron will run on the price hit it will update the counter for +1 and an email function will check the count == 3 send the email
entry alerter -
- when volume is more than 3x than daily average and change percentage more than 3%
- the high of this day will be pivot point
- keep tracking the price against pivot point
- issues - we have to think something about retracement pattern by code
TODO: Done entry filter -
- volume > 1.5
- cross match
- openCloseDiff > 2
- highCloseDiff < 2.5
TODO:
- add all the filtered stocks to a analysis table then we will take the percentage difference between current price and entry price
- we will analyse the total profit whether it will work or not
- add new column - profit, in filter_history and argus will calculate the profit every day on it
- add new column trade_completed bool it will also mark the exit of the stock when price cross down the 20 ema and we will not calculate the profit on these stocks
TODO:
- write code to make the "last_traded_price" to be updated using go routines
- write code to upsert data to historic prices table every working day at 5 pm
- write a code to calculate 20 ema of the stock
TODO: 3 stage filtering
-
Alert
- has to be green candle
- query stock which are active
- if price is in between 2% of cross_match_pivot and 5% of cross_match_pivot and active == true
- active means it is near about equal to 52 week high
- before sending the alert mark the flag entry as true
- if not in entry band mark it as false for the UI top picks page
- top picks will only show entry stocks
- volume > 1.2
-
Insert
- if cross_match == true of todays data insert the record with active = true and cross_match_pivot = ltp
- and also check that this record is not present in the table and active != true
- meaning before inserting query all the records which are active and skip insertion
- volume > 1.5
-
Reset
- if the active == true and red candle then mark the active = false
-
filter cron to start everyday at 2
-
it doesnt have to be accurate match cause we might miss good ones
Changelog 24 Aug 2024: New Ideas
- More Dynamic Stoploss
- top picks should have 2 columns
- active which should be not more than 3 days like 3 din tk to entryable rahega
- uske baad vo chala jaega old top picks me
- health check mails
DB Design
BUG TODO:
- reset entry based on date also
- health check report needs to be generated and then sent properly only once per day
- also test the reset of the stock
- this fucntion will operate on filter_history table
- check that what will happen if the stocks becomes active again will the new entry be create or same entry gets updated
- new target implementation
- add cross matched stocks list with progress bar to see how far it is from entry price
- 2 sections will be there namely - potential entry stocks and Ready to entry stocks
- add a new table 'backtester' which will add stocks and use my 5target-10stoploss-15stoploss strategy to mark entry and exit
- entry bool
- entry_date date
- exit when stoploss hit
- exit bool
- exit_date date
- what will happen if exited stock becomes active again
- i will add a new fucntion named backtester which will run in the end of the stock market session daily
- I am being lenient on the accurate stoploss lets see how much i can get by this
- this new fucntion backtester will operate on new table backtester
- we are decoupling back test results and top picks
- test it by running inbetween market by running twice
- test it at 4 pm by running twice
