This repository provides Python code for adding anchored VWAPs to OHLC data, generating charts, and saving them as .png files. It can help you apply the trading strategies described in Brian Shannon's book, Maximum Trading Gains with Anchored VWAP. The precondition is that you're familiar with running Python scripts and have some knowledge of Pandas.
The principal function vwaps_plot_build_save is located in a file of the same name in the root folder. Additionally, the repository includes several other functions that use this function. These functions build and save daily charts and charts with intervals of 1 to 60 minutes. There are also functions in the import_ohlc folder for importing OHLC data from various providers Yahoo Finance.
You can import the functions into the run_main.py file and run them from there. Before doing so, it's a good idea to complete the following tasks.
- Install all required packages listed in the
requirements.txtfile, especiallykaleido. - Try to understand in detail how the function
vwaps_plot_build_saveworks. - Check out the
draw_all_daily_chartsfunction. It can be helpful to run this function before the beginning of each trading day. - Take a look at the function
draw_daily_chart_ticker. It will come in handy when you need to quickly draw a daily chart for some ticker. An example of its usage is in therun_main.pyfile. - Advanced, optional, risky If you want, try to use the function
draw_qqq_intradayin therun_main.pyfile for intraday trading.
The primary function responsible for building and saving Anchored VWAPs charts is vwaps_plot_build_save. Its declaration is as follows:
def vwaps_plot_build_save(
input_df: pd.DataFrame,
anchor_dates: List[str],
chart_title: str = "",
chart_annotation_func: Callable = get_chart_annotation,
add_last_min_max: bool = False,
file_name: str = DEFAULT_RESULTS_FILE,
print_df: bool = True,
hide_extended_hours: bool = False,
):To plot the anchored VWAPs chart, you'll need a DataFrame that contains the following columns: Open, High, Low, Close, and Volume. Also, you'll have to provide a list of anchor dates.
By default, the first point on the chart's X-axis will be the earliest date in the anchor dates list. To adjust this behavior, add x to the date you want the chart to start. For example, x2024-08-03 00:00:00 instead of 2024-08-03 00:00:00.
Anchored VWAPs that start from the first day of the year or the first minute of the trading day are popular and valuable. However, as the year or trading day progresses, the charts can become overburdened and less helpful. For example, the intraday chart below with a 1-minute interval, plotted towards the end of the trading day, is overcrowded. It doesn't allow us to see the details of the recent market activity.
Use the method described above to create a new chart that isn’t cluttered with old data. This approach shows recent market activity clearly.
You won’t need to increase the candle interval, such as switching from one minute to five minutes or from one day to a week.
I have found that VWAPs anchored to the dates of the last minimum and maximum are very important for swing trading. They are more significant than the Maximum Trading Gains with Anchored VWAP book says. Therefore, the vwaps_plot_build_save function acquired an additional parameter add_last_min_max. It saves me a lot of time and effort because I no longer have to follow and update these dates manually.
For intraday charts, VWAPs anchored to the dates of the last minimum and maximum are of little help. They are usually redundant. When building such charts, it is better not to add them.
Creating chart titles is straightforward. You can refer to the example below in the section on drawing intraday charts. Also, explore the code of the get_chart_annotation_1d function to see what information is included in the default annotation.
You may want to put different data in the titles and annotations of your charts. To make the annotation more informative, consider modifying the get_chart_annotation_1d function. It is even better to create one or more custom annotation functions.
To draw intraday charts, you can use the following function as inspiration and a starting example.
def draw_qqq_intraday():
ticker = "QQQ"
interval = "1m"
hist = get_ohlc_from_yf(ticker=ticker, period="2d", interval=interval)
anchor_dates = [
"2024-10-07 13:30:00",
"2024-10-07 18:00:00",
"2024-10-07 19:42:00",
]
chart_title = {"ticker": ticker, "interval": interval}
chart_title_str = str(chart_title)
vwaps_plot_build_save(
input_df=hist,
anchor_dates=anchor_dates,
chart_title=chart_title_str,
chart_annotation_func=get_chart_annotation,
add_last_min_max=False,
file_name=f"intraday_{ticker}.png",
print_df=True,
hide_extended_hours=True,
)At the start of the trading day, you only have the initial date and time to build the first and most important Anchored VWAP. As the day progresses, other key anchor points gradually become evident.
Throughout the day, you might want to tidy up the chart by removing outdated data. Rather than increasing the candle interval, you can set one of your Anchored VWAPs as the minimum threshold for the X-axis.
For example, this is what happened after I replaced 2024-10-07 18:00:00 with x2024-10-07 18:00:00 in the anchor_dates list.
It's a good idea to set the print_df parameter to True. It allows you to monitor the quality of the intraday data you receive from your provider using the DataFrame's tail displayed on the screen.
The draw_all_daily_charts function makes it easy to build updated charts for all the tickers you're tracking. Previously, these charts primarily depended on custom anchor dates provided by the user. While the option to add custom anchor dates remains, you can now generate helpful charts without needing to specify them.
Key dates include January 1 of the current year and the dates of recent significant price highs and lows. The system automatically adds VWAPs anchored to these material dates to the chart.
Below is an example of a daily OHLC chart that doesn't include any custom anchor dates but features a custom annotation.
Before running the draw_all_daily_charts function, enter the tickers you're interested in into the tickers_follow_daily.xlsx file. This file contains two worksheets. On the first worksheet, specify the tickers and any relevant notes, as shown in the image below. Adding notes is optional.
On the second worksheet, enter the tickers in the columns and list their custom anchor dates below them. If you place an "X" before a date, the system will use that date as the minimum threshold for the chart's X-axis.
For every ticker you are interested in, you can specify one or more custom anchor dates or none at all. The charts remain highly useful even without custom dates. If you don't want to assign custom anchor dates, there is no need to enter that ticker on the second worksheet.
You can modify the value of first_day_of_year in the constants.py file. It’s generally better not to set it to the first day of the current year in January. Instead, consider waiting until February or even March before updating this value.
See also the function draw_daily_chart_ticker. It will come in handy when you need to quickly draw a daily chart for some ticker. Fill in the ticker and anchor dates, then call it as shown below.
ticker = "NVDA"
anchor_dates = [
"2024-01-01 00:00:00",
] + ["2024-08-03 00:00:00", "2024-07-11 00:00:00"]
draw_daily_chart_ticker(ticker=ticker, anchor_dates=anchor_dates)Check the function code to see how the .png file name for saving the chart is generated.
Brian Shannon, author of Maximum Trading Gains with Anchored VWAP, emphasizes the importance of 5-day moving averages, particularly on 15-minute and 30-minute candlestick charts. You can use the draw_5_days_avg function to plot such charts.
Its declaration:
def draw_5_days_avg(ticker: str, interval: str = "15m"):
"""
Create and save plot 5_d_avg_{ticker}.png
containing OHLC candles and 5 days simple moving average (SMA).
Usage: avoid buying the dip until the price consolidates above 5 days SMA.
For details, see Appendix B
of the book "Maximum Trading Gains With Anchored VWAP".
"""Here is an example of a chart produced by this function.
from draw_ratio import draw_ratio
...
draw_ratio(ticker_1="IWM", ticker_2="QQQ", cutoff_date="2020-01-01")Here the result.
You can follow me on Twitter and connect with me on LinkedIn.
See also my Trading strategy template that uses Python backtesting library.