多个股票投资组合回归并重新平衡

Question

So I have a portfolio composition (some shares included in the portfolio), and at specific dates I need to rebalance -that means the composition of the portfolio changes. Dataframe includes columns ['Names'; 'Date_rebalancing'; 'Weights']

Then I have another big dataframe with daily return of all possible stocks. I need to extract the portfolio return daily.

So when the rebalancing date happens portfolio_return_next_day is sumproduct(weights; returns_only_for_those_companies). Daily return is computed.

stock_return=prices.apply(lambda x: x/x.shift(1)-1)

I expect a list of daily portfolio returns

Answer 1

On your big dataframe, you need to filter so that it matches your portfolio. So use .isin()

df_select = df_big.loc[df_big['name'].isin.[df_port['name'],:]

df_select is now a dataframe with just your stock daily values. To get daily returns, there is a built in function pct_return() . (# defaults to one period)

Let's say your column of closing prices is called 'Close'. Then create a new columns from the daily returns.

df_select['return'] = df['Close'].pct_return()  

This will give you the dataframe with your stocks and percent return daily values.

Answer 2

df_unique_dates are the dates where the rebalancing should take place. I already done the returns for each stock in stock_return but at each rebalancing date I have to select only those stocks that appear in portfolio list_stocks

    for i in range(len(df_unique_dates)):
    list_stocks=dataset.Name.loc[df_unique_dates[i]]
    for j in range(len(stock_return)):
        if stock_return.index[j]>df_unique_dates[i] and stock_return.index[j]<=df_unique_dates[i+1]:
             stock_return['portfolio']=0.1*stock_return[list_stocks].sum(axis=1)
stock_return