67 lines
2.4 KiB
Markdown
67 lines
2.4 KiB
Markdown
# VWAP Trend Bot
|
|
|
|
## Overview
|
|
|
|
This is an implementation of the strategy presented in the paper [Volume Weighted Average Price (VWAP) The Holy Grail for Day Trading Systems](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4631351) (2023).
|
|
|
|
The strategy is very basic, initiating long positions when price closes above VWAP and short positions when price closes below VWAP.
|
|
|
|
Trades are held either until price closes in the opposite direction with respect to VWAP or the end of the trading day.
|
|
|
|
## Backtest Results
|
|
|
|
### ETFs
|
|
|
|
#### TQQQ, 1 Minute, 2018-2024
|
|
|
|
Starting from 2018 with TQQQ which is what the authors of the paper did.
|
|
|
|
|
|
|
|
|
|
|
|
#### TQQQ, 1 Minute, 2010-2024
|
|
|
|
1-minute data for TQQQ from my data provider begins in February 2010.
|
|
|
|
|
|
|
|
|
|
|
|
Unfortunately the results for TQQQ don't look as good as the paper suggests, at least not anymore.
|
|
|
|
#### QQQ, 1 Minute, 2007-2024
|
|
|
|
1-minute data for QQQ from my data provider begins in April 2007.
|
|
|
|
|
|
|
|
|
|
|
|
#### SPY, 1 Minute, 2007-2024
|
|
|
|
1-minute data for SPY from my data provider begins in April 2007.
|
|
|
|
|
|
|
|
|
|
|
|
### Futures
|
|
|
|
**NOTE**: Backtest results for futures are using US equities extended trading hours as the results look more favorable than when including all of globex.
|
|
|
|
Also, I'm using a higher time frame here (15 minute) in order to limit the number of trades due to fees + commissions on futures, but results for lower time frames should be similar.
|
|
|
|
#### NQ, 15 Minute, 2018-2024
|
|
|
|
|
|
|
|
|
|
|
|
#### ES, 15 Minute, 2018-2024
|
|
|
|
|
|
|
|
|
|
|
|
--- |