Manipulating Financial Data in Python

Module 1: Reading, slicing and plotting stock data

• Overview of the data we'll be working with.
• Introduction to our primary library: Pandas
• Reading CSV data into Pandas
• Filtering to specific dates
• Sorting
• Plotting

script

• Overview of data we'll be working with: AAPL.csv, SPY.csv (note date order)
  • Meaning of various columns
• The Pandas dataframe
• Read CSV into a dataframe (AAPL example)
• Slice according to dates
• [quiz: read SPY.csv and slice against different dates]
• Plot (note date order wrong)
• Sort
• Plot

Module 2: Building a dataframe with lots of stocks

• Our target dataframe structure
• Reading data for multiple stocks into the structure
• Plotting
• Normalizing

script

• What we want to end up with: Rows: Dates, Columns: Symbols
• Step by step how to build it
• SPY.csv will be our reference -- it trades every day the market is open.
• Read SPY.csv, slice to date range, sort
• Read AAPL.csv, merge() into existing dataframe
• Repeat with GLD, IBM, GOOG
• Plot and display legend
• Observe: Scale not good, let's normalize
• Print some of the numbers
• Plot after normalization
• [quiz: normalize at a different date]

Module 3: Numpy fundamentals

• What is Numpy and how it relates to Pandas
• Creating Numpy arrays
• Indexing and slicing Numpy arrays
• Important data processing on Numpy arrays

script

• Numpy relationship to Pandas
• Creating Arrays
  • empty, zeros, ones
• Basic Indexing and Slicing (start at 0 not 1)
• [quiz: print 2nd & 3rd columns]
• Index one array by another
• Reshaping
• Data Processing using Arrays
  • Sum rows, Sum columns
  • Statistics on columns: Mean, Median, stddev
• See: http://wiki.quantsoftware.org/index.php?title=Numpy_Tutorial_1

Module 4: Statistical analysis of time series

• Rolling statistics example
  • Read SPY
  • 20 day rolling average
  • +- 20 day rolling stdev * 2
  • Plot above as Bollinger bands
• Discussion of daily returns, what they are, how to calculate
• [quiz: compute and plot SPY and XOM daily returns]
• Scatter plot (plot SPY vs XOM)
• Compare SPY vs GLD. How can we quantify these differences?
• Fit a line and plot it, print slope and corrcoef for SPY & XOM
• Discussion of correlation not the same as slope

Module 5: Incomplete data

• [for this lesson: need to create 4 assets: SPY no missing, X ends midway, Y begins midway, Z has periodic outages]
• Read SPY
  • 20 day rolling average
  • Plot
• Attempt above with X, what happens? (incomplete data)
• Look at data; NaN!
• What to do?
• Discussion & drawing of the types of incomplete data characterized by 4 examples above.
• What is the proper way to handle?
• [quiz: implement and plot fill forward on X]
• Show Pandas methods for forward fill and backward fill, plot rolling averages

Module 6: Computing statistics on a portfolio

• Statistics we'll look at:
  • Average daily return
  • Stddev of daily return: volatility
  • Total return
  • Sharpe ratio
• Buy and hold: SPY, GLD, GOOG, XOM, $100K each
• Show how to read assets, normalize so each starts at $100K, goes forward.
• [quiz: compute total daily portfolio value and daily rets for portfolio]
• Show how to compute, avg daily rets, stdev, total ret, Sharpe ratio

Module 7: Optimizers: Building a parameterized model

• What does an optimizer do?
• Show syntax of optimizer use
• Create fake data
• Try to fit it using an optimizer
• [quiz: add another type of curve to fit (e.g., sine)]

Module 8: Optimizers: How to optimize a portfolio

• Frame the portfolio optimization problem for the optimizer
• Add target return
• Plug the parts together in code for 4 assets
• [quiz: Add constraints on holdings]