Owen,

At its most basic level leverage is when you have to borrow funds to complete an order, so any time you have to borrow money your leverage goes up. Consequently, set_universe has no bearing on the amount of leverage your algo takes on, it is usually down to you order execution. Here a are a couple tips to help avoid over leveraging your account(Note: leverage is not bad).

• Use order_target_percent() this allows greater control over what your portfolio allocation is to a given security, and is much better and more robust than calculating the number of shares to order manually.

• Check for open orders using the get_open_orders() method. If your order hasn't been filled yet and you make another order for that security you can risk over leveraging your account, so define a set of behaviors for what to do with open orders such as exiting that bar of trade execution.

• Record your leverage and if you see any spikes do some debugging to see what that specific issue is, it may be obscure. For example last night I had leverage spikes that were a result of a double-negative portfolio weight.

• Finally, be aware of your portfolio's holdings, sometimes if a security gets delisted it remains in your portfolio (something we are working on fixing), among other things this is probably a good idea anyway.

Thanks! I use order_target_percent and get_open_orders but my leverage, for some reason, is always in the range of 23-800. Here is the code

if bool(get_open_orders(stock)):  
            return  
        if current_price <= buy_long_band and cash >= 5000 and notional < context.max_notional:  
            order_target_percent(stock, .1)  
            log.debug("Buying long" + str(stock))  
        elif current_price >= close_band:  
            order_target(stock, 0)  
            log.debug("Selling long" + str(stock))  
        elif current_price <  long_stop_band:  
            order_target(stock, 0)  
            log.debug("Lost too much money, stopping position in" + str(stock))  
        if bool(get_open_orders(stock)):  
            return  
        if context.order_placed == True and current_price <= average_band or current_price > short_stop_band:  
            order_target(stock, 1000)  
            log.debug("Covering short position in" + str(stock))  
            context.order_placed = None  
        elif context.order_placed == False and current_price > short_band and notional > context.min_notional:  
            order_target(stock, 1000)  
            log.debug("Short selling 10% of " + str(stock))  
            context.order_placed = True  

I tried debugging but I simply could not find the issue.

@James Christopher

Actually never mind, I fixed the problem.

Hi Owen,

Do you mind explaining the fix? I'm stuck with something similar.

Regards,
AVM