0bps

Strategy vault / Machine learning

Timeframe Power Trading Strategy

The Timeframe Power Trading Strategy is a strategy that utilizes the price trend patterns of stocks during different timeframes within a day. It seeks to identify the optimal long or short opportunities across 48…

Pine ScriptMachine learningTrend following
Author
ChaoZhang
Complexity
Intermediate · 159 lines
Parameters
54
Backtest bar
1h · intraday
Tested on
BTC/USDT · futures
Jump to codeView in vaultOriginal on FMZ

IMG

Overview

The Timeframe Power Trading Strategy is a strategy that utilizes the price trend patterns of stocks during different timeframes within a day. It seeks to identify the optimal long or short opportunities across 48 half-hourly timeframes in a day.

Strategy Logic

The core logic of this strategy is that stock prices tend to exhibit certain patterns during different periods in a day. The strategy sets up 48 half-hourly timeframes throughout the day and determines whether to go long, go short or do nothing during each timeframe. When time enters a certain timeframe, if the setting is "long", it will open a long position. If the setting is "short", it will open a short position. At the end of each timeframe, it checks the operation type of the next timeframe. If it's the same as the current one, it will continue holding the position. If different, it will close out positions before the timeframe ends.

For example, if the timeframe 6:30am - 7:00am is set to "long", the strategy will open a long position at 6:30am. If 7:00am - 7:30am is set to "short", it will close long positions before 7am and open short positions at 7am.

The advantage of this strategy lies in its ability to capitalize on intraday price swings of stocks. The risk is that such patterns may change over time and render the strategy ineffective.

Advantage Analysis

The biggest edge of this strategy is that it utilizes the "Prices is Right" attribute of stocks - prices tend to have different mean and variance during different periods of the day. This allows the strategy to adopt range trading tactics during volatile periods and trend trading tactics during stable periods to adapt to varying market conditions.

Another advantage is the flexibility of parameter configuration. Optimal parameter sets could be used for different stocks to offset uncertainties.

Risk Analysis

The main risk comes from instability of assumptions - if the intraday price pattern changes substantially for a stock, the profitability expectations of the strategy will be affected. Such changes could come from fundmental shifts or black swan events affecting the overall market.

Also, high trading frequency poses risks in terms of transaction costs. Without sufficient trading volume, accumulation of fees could erode end returns.

Optimization Guidelines

Consider introducing machine learning models to enable dynamic adjustment of parameters - e.g. LSTM models to forecast next-period prices and fine-tune long/short settings accordingly.

Alternatively, combine stock fundamentals to gauge likelihood of pattern shift, to determine optimal timing for strategy activation.

Conclusion

The Timeframe Power Trading Strategy generates alpha by identifying optimal intraday operations during different periods when analyzing recurring price patterns. With flexible parameter adjustment and risk controls, it is an efficient algo trading strategy. Future optimization paths involve ML adoption or fundmental combos to expand profitability and enhance robustness against uncertainties.

|| Timeframe Power Trading Strategy

Overview

The Timeframe Power Trading Strategy is a strategy that utilizes the price trend patterns of stocks during different timeframes within a day. It seeks to identify the optimal long or short opportunities across 48 half-hourly timeframes in a day.

Strategy Logic

The core logic of this strategy is that stock prices tend to exhibit certain patterns during different periods in a day. The strategy sets up 48 half-hourly timeframes throughout the day and determines whether to go long, go short or do nothing during each timeframe. When time enters a certain timeframe, if the setting is "long", it will open a long position. If the setting is "short", it will open a short position. At the end of each timeframe, it checks the operation type of the next timeframe. If it's the same as the current one, it will continue holding the position. If different, it will close out positions before the timeframe ends.

For example, if the timeframe 6:30am - 7:00am is set to "long", the strategy will open a long position at 6:30am. If 7:00am - 7:30am is set to "short", it will close long positions before 7am and open short positions at 7am.

The advantage of this strategy lies in its ability to capitalize on intraday price swings of stocks. The risk is that such patterns may change over time and render the strategy ineffective.

Advantage Analysis

The biggest edge of this strategy is that it utilizes the "Prices is Right" attribute of stocks - prices tend to have different mean and variance during different periods of the day. This allows the strategy to adopt range trading tactics during volatile periods and trend trading tactics during stable periods to adapt to varying market conditions.

Another advantage is the flexibility of parameter configuration. Optimal parameter sets could be used for different stocks to offset uncertainties.

Risk Analysis

The main risk comes from instability of assumptions - if the intraday price pattern changes substantially for a stock, the profitability expectations of the strategy will be affected. Such changes could come from fundmental shifts or black swan events affecting the overall market.

Also, high trading frequency poses risks in terms of transaction costs. Without sufficient trading volume, accumulation of fees could erode end returns.

Optimization Guidelines

Consider introducing machine learning models to enable dynamic adjustment of parameters - e.g. LSTM models to forecast next-period prices and fine-tune long/short settings accordingly.

Alternatively, combine stock fundamentals to gauge likelihood of pattern shift, to determine optimal timing for strategy activation.

Conclusion

The Timeframe Power Trading Strategy generates alpha by identifying optimal intraday operations during different periods when analyzing recurring price patterns. With flexible parameter adjustment and risk controls, it is an efficient algo trading strategy. Future optimization paths involve ML adoption or fundmental combos to expand profitability and enhance robustness against uncertainties.

Parameters

NameDefaultWhat it controls
v_input_16From Month
v_input_214From day
v_input_32021From Year
v_input_412To Month
v_input_531To day
v_input_62100To Year
v_input_700000-0030: None | Short | Long
v_input_800030-0100: Long | Short | None
v_input_900100-0130: Long | Short | None
v_input_1000130-0200: Long | Short | None
v_input_1100200-0230: Long | Short | None
v_input_1200230-0300: None | Short | Long
v_input_1300300-0330: None | Short | Long
v_input_1400330-0400: None | Short | Long
v_input_1500400-0430: None | Short | Long
v_input_1600430-0500: None | Short | Long
v_input_1700500-0530: None | Short | Long
v_input_1800530-0600: None | Short | Long
v_input_1900600-0630: None | Short | Long
v_input_2000630-0700: None | Short | Long
v_input_2100700-0730: None | Short | Long
v_input_2200730-0800: None | Short | Long
v_input_2300800-0830: None | Short | Long
v_input_2400830-0900: None | Short | Long
v_input_2500900-0930: None | Short | Long
v_input_2600930-1000: None | Short | Long
v_input_2701000-1030: None | Short | Long
v_input_2801030-1100: None | Short | Long
v_input_2901100-1130: None | Short | Long
v_input_3001130-1200: None | Short | Long
v_input_3101200-1230: None | Short | Long
v_input_3201230-1300: None | Short | Long
v_input_3301300-1330: None | Short | Long
v_input_3401330-1400: None | Short | Long
v_input_3501400-1430: None | Short | Long
v_input_3601430-1500: None | Short | Long
v_input_3701500-1530: None | Short | Long
v_input_3801530-1600: None | Short | Long
v_input_3901600-1630: None | Short | Long
v_input_4001630-1700: None | Short | Long
v_input_4101700-1730: None | Short | Long
v_input_4201730-1800: None | Short | Long
v_input_4301800-1830: None | Short | Long
v_input_4401830-1900: None | Short | Long
v_input_4501900-0930: None | Short | Long
v_input_4601930-2000: None | Short | Long
v_input_4702000-2030: None | Short | Long
v_input_4802030-2100: None | Short | Long
v_input_4902100-2130: None | Short | Long
v_input_5002130-2200: None | Short | Long
v_input_5102200-2230: None | Short | Long
v_input_5202230-2300: None | Short | Long
v_input_5302300-2330: None | Short | Long
v_input_5402330-0000: None | Short | Long

Source Pine Script

/*backtest
start: 2023-10-23 00:00:00
end: 2023-11-22 00:00:00
period: 1h
basePeriod: 15m
exchanges: [{"eid":"Futures_Binance","currency":"BTC_USDT"}]
*/

// This source code is subject to the terms of the Mozilla Public License 2.0 at https://mozilla.org/MPL/2.0/

//@version=4
strategy("Timeframe Time of Day Buying and Selling Strategy", overlay=true)

frommonth = input(defval = 6, minval = 01, maxval = 12, title = "From Month")
fromday = input(defval = 14, minval = 01, maxval = 31, title = "From day")
fromyear = input(defval = 2021, minval = 1900, maxval = 2100, title = "From Year")

tomonth = input(defval = 12, minval = 01, maxval = 12, title = "To Month")
today = input(defval = 31, minval = 01, maxval = 31, title = "To day")
toyear = input(defval = 2100, minval = 1900, maxval = 2100, title = "To Year")

timeframes = array.new_string(48, '')
timeframes_options = array.new_string(49, 'None')

array.set(timeframes,0,'2330-0000')
array.set(timeframes_options,0, input(defval='None', options=['Long','Short','None'], title='0000-0030'))
array.set(timeframes,1,'0000-0030')
array.set(timeframes_options,1, input(defval='Long', options=['Long','Short','None'], title='0030-0100'))
array.set(timeframes,2,'0030-0100')
array.set(timeframes_options,2, input(defval='Long', options=['Long','Short','None'], title='0100-0130'))
array.set(timeframes,3,'0100-0130')
array.set(timeframes_options,3, input(defval='Long', options=['Long','Short','None'], title='0130-0200'))
array.set(timeframes,4,'0130-0200')
array.set(timeframes_options,4, input(defval='Long', options=['Long','Short','None'], title='0200-0230'))
array.set(timeframes,5,'0200-0230')
array.set(timeframes_options,5, input(defval='None', options=['Long','Short','None'], title='0230-0300'))
array.set(timeframes,6,'0230-0300')
array.set(timeframes_options,6, input(defval='None', options=['Long','Short','None'], title='0300-0330'))
array.set(timeframes,7,'0300-0330')
array.set(timeframes_options,7, input(defval='None', options=['Long','Short','None'], title='0330-0400'))
array.set(timeframes,8,'0330-0400')
array.set(timeframes_options,8, input(defval='None', options=['Long','Short','None'], title='0400-0430'))
array.set(timeframes,9,'0400-0430')
array.set(timeframes_options,9, input(defval='None', options=['Long','Short','None'], title='0430-0500'))
array.set(timeframes,10,'0430-0500')
array.set(timeframes_options,10, input(defval='None', options=['Long','Short','None'], title='0500-0530'))
array.set(timeframes,11,'0500-0530')
array.set(timeframes_options,11, input(defval='None', options=['Long','Short','None'], title='0530-0600'))
array.set(timeframes,12,'0530-0600')
array.set(timeframes_options,12, input(defval='None', options=['Long','Short','None'], title='0600-0630'))
array.set(timeframes,13,'0600-0630')
array.set(timeframes_options,13, input(defval='None', options=['Long','Short','None'], title='0630-0700'))
array.set(timeframes,14,'0630-0700')
array.set(timeframes_options,14, input(defval='None', options=['Long','Short','None'], title='0700-0730'))
array.set(timeframes,15,'0700-0730')
array.set(timeframes_options,15, input(defval='None', options=['Long','Short','None'], title='0730-0800'))
array.set(timeframes,16,'0730-0800')
array.set(timeframes_options,16, input(defval='None', options=['Long','Short','None'], title='0800-0830'))
array.set(timeframes,17,'0800-0830')
array.set(timeframes_options,17, input(defval='None', options=['Long','Short','None'], title='0830-0900'))
array.set(timeframes,18,'0830-0900')
array.set(timeframes_options,18, input(defval='None', options=['Long','Short','None'], title='0900-0930'))
array.set(timeframes,19,'0900-0930')
array.set(timeframes_options,19, input(defval='None', options=['Long','Short','None'], title='0930-1000'))
array.set(timeframes,20,'0930-1000')
array.set(timeframes_options,20, input(defval='None', options=['Long','Short','None'], title='1000-1030'))
array.set(timeframes,21,'1000-1030')
array.set(timeframes_options,21, input(defval='None', options=['Long','Short','None'], title='1030-1100'))
array.set(timeframes,22,'1030-1100')
array.set(timeframes_options,22, input(defval='None', options=['Long','Short','None'], title='1100-1130'))
array.set(timeframes,23,'1100-1130')
array.set(timeframes_options,23, input(defval='None', options=['Long','Short','None'], title='1130-1200'))
array.set(timeframes,24,'1130-1200')
array.set(timeframes_options,24, input(defval='None', options=['Long','Short','None'], title='1200-1230'))
array.set(timeframes,25,'1200-1230')
array.set(timeframes_options,25, input(defval='None', options=['Long','Short','None'], title='1230-1300'))
array.set(timeframes,26,'1230-1300')
array.set(timeframes_options,26, input(defval='None', options=['Long','Short','None'], title='1300-1330'))
array.set(timeframes,27,'1300-1330')
array.set(timeframes_options,27, input(defval='None', options=['Long','Short','None'], title='1330-1400'))
array.set(timeframes,28,'1330-1400')
array.set(timeframes_options,28, input(defval='None', options=['Long','Short','None'], title='1400-1430'))
array.set(timeframes,29,'1400-1430')
array.set(timeframes_options,29, input(defval='None', options=['Long','Short','None'], title='1430-1500'))
array.set(timeframes,30,'1430-1500')
array.set(timeframes_options,30, input(defval='None', options=['Long','Short','None'], title='1500-1530'))
array.set(timeframes,31,'1500-1530')
array.set(timeframes_options,31, input(defval='None', options=['Long','Short','None'], title='1530-1600'))
array.set(timeframes,32,'1530-1600')
array.set(timeframes_options,32, input(defval='None', options=['Long','Short','None'], title='1600-1630'))
array.set(timeframes,33,'1600-1630')
array.set(timeframes_options,33, input(defval='None', options=['Long','Short','None'], title='1630-1700'))
array.set(timeframes,34,'1630-1700')
array.set(timeframes_options,34, input(defval='None', options=['Long','Short','None'], title='1700-1730'))
array.set(timeframes,35,'1700-1730')
array.set(timeframes_options,35, input(defval='None', options=['Long','Short','None'], title='1730-1800'))
array.set(timeframes,36,'1730-1800')
array.set(timeframes_options,36, input(defval='None', options=['Long','Short','None'], title='1800-1830'))
array.set(timeframes,37,'1800-1830')
array.set(timeframes_options,37, input(defval='None', options=['Long','Short','None'], title='1830-1900'))
array.set(timeframes,38,'1830-1900')
array.set(timeframes_options,38, input(defval='None', options=['Long','Short','None'], title='1900-0930'))
array.set(timeframes,39,'1900-0930')
array.set(timeframes_options,39, input(defval='None', options=['Long','Short','None'], title='1930-2000'))
array.set(timeframes,40,'1930-2000')
array.set(timeframes_options,40, input(defval='None', options=['Long','Short','None'], title='2000-2030'))
array.set(timeframes,41,'2000-2030')
array.set(timeframes_options,41, input(defval='None', options=['Long','Short','None'], title='2030-2100'))
array.set(timeframes,42,'2030-2100')
array.set(timeframes_options,42, input(defval='None', options=['Long','Short','None'], title='2100-2130'))
array.set(timeframes,43,'2100-2130')
array.set(timeframes_options,43, input(defval='None', options=['Long','Short','None'], title='2130-2200'))
array.set(timeframes,44,'2130-2200')
array.set(timeframes_options,44, input(defval='None', options=['Long','Short','None'], title='2200-2230'))
array.set(timeframes,45,'2200-2230')
array.set(timeframes_options,45, input(defval='None', options=['Long','Short','None'], title='2230-2300'))
array.set(timeframes,46,'2230-2300')
array.set(timeframes_options,46, input(defval='None', options=['Long','Short','None'], title='2300-2330'))
array.set(timeframes,47,'2300-2330')
array.set(timeframes_options,47, input(defval='None', options=['Long','Short','None'], title='2330-0000'))


string_hour = hour<10?'0'+tostring(hour):tostring(hour)
string_minute = minute<10?'0'+tostring(minute):tostring(minute)
current_time = string_hour+string_minute


f_strLeft(_str, _n) =>
    string[] _chars = str.split(_str, "")
    int _len = array.size(_chars)
    int _end = min(_len, max(0, _n))
    string[] _substr = array.new_string(0)
    if _end <= _len
        _substr := array.slice(_chars, 0, _end)
    string _return = array.join(_substr, "")

f_strRight(_str, _n) =>
    string[] _chars = str.split(_str, "")
    int _len = array.size(_chars)
    int _beg = max(0, _len - _n)
    string[] _substr = array.new_string(0)
    if _beg < _len
        _substr := array.slice(_chars, _beg, _len)
    string _return = array.join(_substr, "")


for i = 0 to array.size(timeframes) - 1
    start_time = f_strLeft(array.get(timeframes, i), 4)
    end_time = f_strRight(array.get(timeframes, i), 4)
    
    if current_time == end_time and array.get(timeframes_options, i)!='None' and array.get(timeframes_options, i) != array.get(timeframes_options, i==47?0:i+1) and timestamp(toyear, tomonth, today, 00, 00)
        strategy.close_all()

    if current_time == start_time and array.get(timeframes_options, i)!='None' and array.get(timeframes_options, i) != array.get(timeframes_options, i==0?47:i-1)
        if array.get(timeframes_options, i) == 'Long'
            strategy.entry("Long", strategy.long, when=(time > timestamp(fromyear, frommonth, fromday, 00, 00) and time < timestamp(toyear, tomonth, today, 00, 00)))
        else if array.get(timeframes_options, i) == 'Short'
            strategy.entry("Short", strategy.short, when=(time > timestamp(fromyear, frommonth, fromday, 00, 00) and time < timestamp(toyear, tomonth, today, 00, 00)))

Published by ChaoZhang on the FMZ strategy square · last modified 2023-11-23. Collected in the Quant Trading Vault. Unvalidated research: re-test with your own data and costs before trusting it.

Build this with AgenKit

A spec is not a trading system. Paste this into your coding agent and AgenKit runs a specialist team — architecture, test-first build, risk gates, review — with paper trading on by default.

in Claude Code, Cursor, Codex & 4 more
/agenkit build a production-grade trading bot with a backtest engine, risk controls and paper trading by default from strategies/Timeframe-Power-Trading-Strategy.md
  1. npx agenkit install engineering-kit in your project
  2. Paste the command into your agent and approve each phase
  3. Get tests, backtest, risk controls and a paper-trading switch
Get AgenKitHow /agenkit works

Understand it first

Before you run this, read the lesson behind the idea and how to backtest it honestly. Specs like this one are the survivors of an unknown number of trials; deflate their results accordingly.

Related specs

esc

Try , , or .

↑↓ move↵ open