Hi Andrew,

Welcome to Quantopian! It's great to start like a simple strategy like this to help you grow more comfortable with the platform. With the code you posted, it looks like you're on the right track implementing the strategy you described. A couple of comments/suggestions:

• It looks like your logic for entering position is backwards, your condition checks if the current price is greater than or equal to 8% of yesterday's close price. To follow the logic you described, you will need to get history(bar_count=3, frequency='1d', field='price') and then you will want to look at price_history[s][-3] and price_history[s][-2] in your for loop.

• For your exit strategy, since you have two exit conditions, you will want to store your entry price in a variable in context. For example, you can set context.enter_prices = {} in the initialize(context) function, and then when you enter a position in the for loop, you can store your enter price by doing something like context.enter_price[s] = current_price. Then, every day, you can compare the current price to the entry price for each stock in data and exit when one of your conditions is met.

Does this help?

Thanks Jamie!

That was very helpful.

I was able to correct the oversight on the buy condition, and I've included my code below. The only issue is that I would like to place a market buy order to execute at market open the next day, and I would like the sell order to use minute data instead of daily close prices. I cleaned up and commented my current algorithm, if you would be so kind as to take a look!

Here's what I'm trying to model:

• TVIX closes at $10.00 on Monday.
  
• TVIX drops to $9.20 on Tuesday - an 8% drop.
  
• A market buy order is placed to execute Wednesday morning at market open.
  
• On the next Monday, the stock goes up 8% at any point during the day and a sell order is sent to the market.

Thanks again in advance!

def initialize(context):  
    context.etf = symbol('TVIX')        # Select security.  
    context.entry_price = 0             # Initialize variable.  
def handle_data(context, data):  
    # Create objects to contain open and close prices.  
    close_history = history(bar_count=3, frequency='1d', field='close_price')  
    open_history = history(bar_count=3, frequency='1d', field='open_price')  
    for s in data:  
        # Checks to see if yesterday's close is 8% less than the close of  
        # the previous day.  
        twodaysago = close_history[s][-3]  
        yesterday = close_history[s][-2]  
        # Place a buy order if the condition is met.  
        if yesterday <= (twodaysago * .92):  
            context.entry_price = open_history[s][-1]  
            print context.entry_price  
            order_target(s, 1)  
    # Create a variable to see the current price.  
    current_price = data[context.etf].price  
    if current_price >= context.entry_price * 1.08:  
        # Take profit of 8%.  
        order_target(context.etf, 0)  
    if current_price <= context.entry_price * .92:  
        # Stop loss of 4%.  
        order_target(context.etf, 0)  

In order to implement the logic you explained to me, you will need to run your backtest in minute mode (a better simulation anyway) and you will want to use the schedule_function() method to call a function every day that you define to check for buy orders. You will then have to handle your checks for selling separately in handle_data(), which will get called every minute in minute mode.

In the end, you will have 3 functions:

• initialize(context)
  
• handle_data(context, data) -- where sell orders will be handled
  
• check_buy(context, data) -- called by schedule function, the name is arbitrary!

There are some good examples of schedule function in the help docs (follow the link above) and if that's not enough, there are more throughout the forums.

Thanks Jamie!

I was hesitant to use minute mode unnecessarily, because I imagine the server load for backtesting is much greater, but I will work on recreating this strategy with the schedule function!

Thanks for pointing me in the right direction!

• Andrew