How do I backtest using a list of specific stocks for my algorithm ?

Hi wonderful Quantopians!

I have a pretty straightforward model. It uses a ranking filter of fundamentals and AVG Dollar Volume strategy when purchasing US stocks. The model uses a bucket of quality assets available for purchase once every week.

I want to test my model using only a list of a certain amount of specific stocks(For example 2-10 different stocks). I would need to remove the filters I have in my model just in case. However, I cannot seem to figure out the simple code needed to make my model purchase from only a list instead or such a large universe.

I keep coming across errors in the IDE, would love some guidance on this one.

def initialize(context):  
    # Universe Selection  
        market = (  
        MarketCap().percentile_between(MIN_Market_Cap_Percentile, 100)  
        & Sector().notnull()  
        & builtin.USEquityPricing.close.latest.notnull()  
        & (builtin.USEquityPricing.volume.latest > 0)  
         )

        monthly_top_volume = (  
        AverageDollarVolume(window_length=LiquidityHistory)  
        .top(MarketSize, mask=market)  
        .downsample(Market_Data_Frequency)  
         )  

  universe = monthly_top_volume & market

The above code is my universe, when I seem to transform it to a list it does not seem to work. I'm clearly doing something wrong. You can see the rest of my code in the backtest!

Cheers everyone and thanks ahead.

To limit a pipeline to specific securities one can use either the StaticAssets or StaticSids filters (https://www.quantopian.com/docs/api-reference/pipeline-api-reference#quantopian.pipeline.filters.StaticAssets).

Remember to import the filters first and then simply set your universe to the specific securities you wish.

from quantopian.pipeline.filters import StaticAssets, StaticSids  
# Here we limit out universe to four assets  
universe = StaticAssets(symbols('AAPL', 'IBM', 'C', 'FB'))

Limiting the universe in this way may help debugging but be cautious to not read too much into the results. As an example, the above algo has a min and max position size constraint. With only four assets, the entire portfolio may not be allocated and leave a very small gross exposure (ie leverage). Also, the constraint on sector exposure may be impossible to satisfy (or be greatly skewed) if, for example, the chosen static assets are from just one or two sectors.

See the attached backtest. Simply added the two lines above to the original code.