Compute Time Series

The compute_time_series() function calculates the growth of a given pandas DataFrame of returns, creating a time series. This can be useful for investors and analysts to understand the growth of their investments over time, as well as for comparing the performance of different financial instruments.

To use the compute_time_series() function, you need to provide a pandas DataFrame containing return data. Optionally, you can also provide a start_value, which is the initial value of the time series (default is 100).

After passing the DataFrame to the compute_time_series() function, it will return a pandas DataFrame containing the growth time series for each column in the input DataFrame. This allows users to visualize and analyze the performance of different financial instruments over time.

Function

def compute_time_series(dataframe: pd.DataFrame, start_value: float = 100) -> pd.DataFrame:
    """
    Compute the growth time series given a DataFrame of returns.

    Parameters
    ----------
    dataframe : pandas.DataFrame
        DataFrame containing return data.
    start_value : float, optional
        Initial value of the time series. Default is 100.

    Returns
    -------
    pandas.DataFrame
        A pandas DataFrame containing the growth time series for each column in the input DataFrame.
    """
    # Validate input
    if not isinstance(dataframe, pd.DataFrame):
        raise TypeError("Input 'dataframe' must be a pandas DataFrame.")

    # Calculate growth time series
    return (np.exp(np.log1p(dataframe).cumsum())) * start_value

Example

import pandas as pd
import numpy as np
import yfinance as yf
import ds4finance as dsf

# Download historical data for SPY
spy = yf.download("SPY", start="2015-01-01", end="2020-12-31", progress=False)

# Calculate daily returns
returns = spy["Adj Close"].pct_change().dropna()

# Calculate growth time series
growth_time_series = dsf.compute_time_series(pd.DataFrame(returns))

# Rename column
growth_time_series.columns = ['SPY Growth']

# Display growth time series
growth_time_series.head()

	SPY Growth
Date
2015-01-05	98.194050
2015-01-06	97.269169
2015-01-07	98.481258
2015-01-08	100.228793
2015-01-09	99.425636

Notes:

• The function calculates the growth time series based on the input DataFrame of returns.

• Make sure the input is a pandas DataFrame containing return data.

• The start_value parameter can be customized to set the initial value of the time series.