# Backtest

Assuming you already have created your first strategy, imported historical candles, and have set correct routes for your strategy, it is time to actually backtest it:

WARNING

Spot support is currently very limited - you can use futures mode with leverage 1 for now.

jesse backtest 'start_date' 'finish_date'

start_date and finish_date must be valid date strings in YY-MM-DD format. A working example would be:

jesse backtest '2016-01-01' '2020-04-06'

And here is the output:

 CANDLES              |
----------------------+--------------------------
 period               |   1557 days (4.27 years)
 starting-ending date | 2016-01-01 => 2020-04-06


exchange    | symbol   | timeframe   | strategy
------------+----------+-------------+--------------------+-------
 Bitfinex   | BTC-USD   | 6h          | TrendFollowingStrategy


Executing simulation...  [####################################]  100%
Executed backtest simulation in:  107.89 seconds


METRICS                          |
---------------------------------+------------------------------------
 Total Closed Trades             |                                192
 Total Net Profit                |                 64735.12 (647.35%)
 Starting => Finishing Balance   |                   10000 => 74659.0
 Total Open Trades               |                                  0
 Open PL                         |                                  0
 Total Paid Fees                 |                           10620.84
 Max Drawdown                    |                            -24.83%
 Sharpe Ratio                    |                                1.2
 Annual Return                   |                             38.43%
 Expectancy                      |                     337.16 (3.37%)
 Avg Win | Avg Loss              |                   1261.49 | 351.89
 Ratio Avg Win / Avg Loss        |                               3.58
 Percent Profitable              |                                43%
 Longs | Shorts                  |                          58% | 42%
 Avg Holding Time                | 3.0 days, 20.0 hours, 15.0 minutes
 Winning Trades Avg Holding Time | 6.0 days, 11.0 hours, 19.0 minutes
 Losing Trades Avg Holding Time  |  1.0 day, 21.0 hours, 14.0 minutes

There are further metrics that are disabled by default. To enable them open your project's config.py file and set the values to True:

'metrics': {
    'sharpe_ratio': True,
    'calmar_ratio': False,
    'sortino_ratio': False,
    'omega_ratio': False,
    'winning_streak': False,
    'losing_streak': False,
    'largest_losing_trade': False,
    'largest_winning_trade': False,
    'total_winning_trades': False,
    'total_losing_trades': False,
}

# Charts

Performing backtest with the chart flag would print out charts for the balance change of your portfolio in the backtest period, and buy/sell points on the asset price change%.

jesse backtest '2016-01-01' '2020-04-06' --chart

When the backtest is finished, Jesse prints the path to the output chart image file:

Chart output saved at:
storage/charts/BT-2020-04-13T15-44-42.png

And here's the image (click to zoom or open image in a new tab to see the full size): chart

# TradingView

This feature is useful for those who use TradingView (opens new window) for their chart needs. Add the --tradingview flag to the backtest command:

jesse backtest '2019-01-01' '2019-10-30' --tradingview

At the end of the process, it will print out the path to a .txt file with a pine script as content. Then open the file, copy its content and paste it inside TradingView's (opens new window) Pine Editor. Now click on "Add to chart" and you will see entries and exits of the strategy.

WARNING

Make sure to see the chart in the same timeframe as the backtest simulation was executed on. Also, only the last ~30 trades are displayed because of a limit on TradingView's side.

# CSV

Jesse can output completed trades into a CSV file. You may open this CSV file with Excel, Google sheets, or whatever tool you like to analyse completed trades, draw further charts, etc.

Usage: Add the --csv flag to the backtest command:

jesse backtest '2019-01-01' '2019-10-30' --csv

Once the backtest is completed, Jesse will print the path to the CSV file.

# JSON

Jesse can output completed trades into a file with JSON format. JSON format is a popular javascript format that is usually used to create APIs on the web.

Usage: Add the --json flag to the backtest command:

jesse backtest '2019-01-01' '2019-10-30' --json

Once the backtest is completed, Jesse will print the path to the JSON file.

# Full Report

Generates QuantStats' (opens new window) HTML output.

Usage: Add the --full-reports flag to the backtest command:

jesse backtest '2019-01-01' '2019-10-30' --full-reports

# Visual analyze of completed trades

There is a very cool tool called "Jesse trades info" made by one of the community members. It helps to parse JSON backtest files and to display them in a convenient way using tables and charts.

You can find screenshots and a video on its Github repository page (opens new window).

← Utilities Live Trading →