mercurial

###Summary I wanted to write a simple framework that would allow me to easily write new strategies, backtest them, take their combined output, execute them and finally, analyze them.

###Why is it called Mercurial? Mercurial is an adjective that describes a person 'subject to sudden or unpredictable changes of mood or mind' ... just like the stock market. Furthermore, Mercury is the God of trade/financial gain in ancient Roman mythology.

###What does it do? Mercurial is a framework that allows you to:

1. Get access to market data from yahoo
2. Plug-in strategies
3. Consolidate output of multiple strategies based on their weighting
4. Backtest the strategies
5. Execute trades via Interactive Brokers
6. Analyze your portfolio and compare to benchmark (S&P)

There are multiple components to Mercurial:

1. Strategy

• This script consolidates output of multiple strategies and stores the data in a mysql table called orders. For example, currently, there are two strategies: strategy_MA and strategy_coin_flip.

2. Trade

• This script reads strategy output from 'orders' table and executes the orders via Interactive Brokers. As orders are sent to Interactive Brokers, messages are received and stored in 'orders' table to reflect latest status of each order. For example, as an order is sent, it's status changes to 'opened' and as it is executed, it changes to 'filed'.

3. Analysis

• As orders are executed, their executed price is stored in the same table. This transaction data is analyzed in the Analysis.py script and is compared to a benchmark (S&P).

4. Backtester

• The backtester can run the strategies for a certain historical period and analyze the results so you can analyze the performance of your strategies.

How do you set it up?

1. First, you need to install mysql. You can find instructions on how to install mysql here: https://www.tutorialspoint.com/mysql/mysql-installation.htm

Then, you need to create two tables: 'orders' and 'positions'. Here is the schema of the two tables:

mysql> describe orders;
+------------+-------------+------+-----+---------+----------------+
| Field      | Type        | Null | Key | Default | Extra          |
+------------+-------------+------+-----+---------+----------------+
| order_id   | int(11)     | NO   | PRI | NULL    | auto_increment |
| date       | datetime    | YES  |     | NULL    |                |
| sim_id     | varchar(60) | YES  |     | NULL    |                |
| security   | varchar(20) | YES  |     | NULL    |                |
| action     | varchar(20) | YES  |     | NULL    |                |
| size       | int(11)     | YES  |     | NULL    |                |
| ask_price  | varchar(20) | YES  |     | NULL    |                |
| exec_price | varchar(20) | YES  |     | NULL    |                |
| status     | varchar(50) | YES  |     | NULL    |                |
+------------+-------------+------+-----+---------+----------------+

mysql> describe positions;
+-------------+-------------+------+-----+---------+-------+
| Field       | Type        | Null | Key | Default | Extra |
+-------------+-------------+------+-----+---------+-------+
| security    | varchar(50) | YES  |     | NULL    |       |
| avg_cost    | varchar(20) | YES  |     | NULL    |       |
| total_value | varchar(20) | YES  |     | NULL    |       |
| quantity    | int(11)     | YES  |     | NULL    |       |
+-------------+-------------+------+-----+---------+-------+

2. Fill in your table/db details in the config file (config.yaml)

3. Download Interactive Brokers demo tool - Trader Workstation, and fill in your port and client_id in config.yaml You can download it from here: https://gdcdyn.interactivebrokers.com/en/index.php?f=1286

4. Change universe details for each strategy in config.yaml to whatever you like.

5. Run

python strategy.py

6. Check 'orders' table to make sure strategy has output orders into the table.

7. Execute the orders by running:

python trade.py

8. Check 'orders' table to see if status of your orders has changed and confirm on Trader Workstation GUI that orders have been placed.

9. Run

python analysis.py

to analyze your portfolio.

Running the backtester

1. Have your strategies ready and make sure they are configured in config.yaml
2. Run this command:

python backtest.py -start <YYYYMMDD> -end <YYYYMMDD> -mode simulation -sim_id <simulation_id>

A simulation id is an id you want to use to differentiate results of your simulation from any previous or future simulation. -end is an optional parameter. If you don't specify it, the backtester will use today's date as the end date.