Python for Finance and Algorithmic Trading with QuantConnect

This lecture navigates the Python crash course exercises from the Python for Finance repo, covering tasks on financial data, string and dictionary access, and functions like price and average price.

Explore NumPy, the Python data science library for creating one- and two-dimensional arrays, slicing and indexing, broadcasting, and universal functions for linear algebra, statistics, trigonometry, and random numbers.

Explore NumPy indexing and selection, including single element access, slicing, broadcasting with scalar values, and conditional selection, with two-dimensional array indexing and copy versus view concepts.

Explore the pandas series object, a one-dimensional data structure with a labeled index for label-based data access. Create series from lists or dictionaries and learn core properties and indexing.

Learn to create pandas data frames from Python objects and numpy arrays, set indexes and columns, and read CSV files with correct file paths in a Jupyter notebook.

Learn to retrieve information from pandas data frames by selecting single or multiple columns, create and update columns, and drop columns with axis handling.

Explores managing pandas data frame rows by setting and resetting the index, selecting single or multiple rows with iloc and loc, dropping or appending rows, and making index changes permanent.

Learn to perform pandas conditional filtering on a data frame by column conditions, using boolean series, multi-condition logic with ampersand and pipe, and the is in method.

Explore the pandas apply method by building and applying custom functions to single columns, like extracting the last four digits of credit card numbers, and extend to multi-column inputs.

Explore pandas essential methods for data frames, describing data, sorting by one or multiple columns, locating max/min indices with idxmax/idxmin, and using value_counts, unique, replace, map, duplicates, and sample.

Learn how to combine data frames in pandas using concatenation, including joining along columns or rows, handling misaligned indices, and preparing for later merge and join ideas.

Master pandas merge to perform inner joins on dataframes, using the how parameter and a common key like name to combine registration and login data.

Learn how left and right merges in pandas combine data frames using the on name key, producing nulls for unmatched rows and why table order matters.

Master pandas outer merge to include all rows from both data frames, handle nulls, and merge on columns or indexes with on, left on, right on, and custom suffixes.

Learn to read HTML tables into pandas data frames with read_html, select and clean the desired table from a webpage or file, and export the data frame back to HTML.

Read excel workbooks as a dictionary of sheets with pandas read_excel, mapping each sheet name to a data frame, then write with to_excel, selecting a sheet and disabling the index.

Learn how pandas connects to SQL engines with a driver and SQLAlchemy, using create_engine and read_sql or read_sql_query to load data frames, including an in-memory SQLite example.

Practice pandas manipulation on the S&P 500 data using the finance exercise notebook and its solutions, including merging datasets, filtering by market cap and dividend yield, and applying price formatting.

Learn pandas workflows for finance: import, explore, drop columns, set index, sort by market cap, filter by dividend yield, merge with constituents, and format prices.

Explore matplotlib fundamentals, including the functional pyplot and object-oriented APIs, to create and customize plots with figures and subplots, and learn to plot equations or data points.

Learn the basics of plotting with the simple plot calls, including labels, axis limits, and saving figures, then transition to the robust object‑oriented figure API.

Explore how the figure object serves as a blank canvas for plotting, with a default size of 432 by 288 pixels, then add axes with add_axes and plot on them.

Learn to create a blank figure, add axes, and plot on them using matplotlib, with multiple axes and zoomed-in inserts. Customize with x/y limits, labels, and titles.

Master figure parameters by adjusting the figure size and dpi for desired resolution, then save with the bounding box inches to include the axes, ensuring clear, correctly sized exports.

Learn to style matplotlib plots by controlling colors (names, hex, rgb), adjusting line width and styles, and customizing markers (size and edge/face colors) for clearer visuals.

Explore plotting with numpy and matplotlib by graphing E = mc^2 across masses and comparing yield curves from 2007 to 2020, including single and dual-axis plots.

Master numpy-based data creation and plotting mass versus energy with E=mc^2, then practice matplotlib techniques for legends, subplots, twin axes, and log-scale yields.

Master pandas time methods for date-time indexed data and visualize finance data with plots and map plot live, covering series insights, moving averages, time shifts, percentage changes, and api access.

Explore advanced time series visualization with pandas and matplotlib by configuring major and minor locators and formatters, customizing tick labels, rotation, and alignment for clear monthly and yearly views.

Compute rolling statistics in pandas with a moving window to reveal rolling means and volatility in finance data, using seven- and fourteen-day windows, Bollinger Bands, and QuantConnect.

Learn to manipulate pandas data with time shifting and row-based calculations, including forward and backward shifts, diffs, percent changes, and cumulative sums/products for financial data.

Master programmatic data sources for finance with pandas data reader and yfinance, exploring FRED macro data and Yahoo Finance stocks for backtesting with QuantConnect Lean.

Explore pandas-based finance analytics with SPY data from 2000–2021, covering loading data, inspecting head and dtypes, plotting adjusted close with 200-day rolling, volume histograms, and yearly maxima.

Explore pandas-based finance data workflows: fetch spy data, plot adjusted close and volume, compute 200-day moving averages, and analyze peak-to-recovery and percent-change dynamics.

Explore fundamental financial concepts with Python, including return and risk measures, fair value, efficient market hypothesis, and portfolio optimization with Markowitz, CAPM beta and alpha.

Assess whether the efficient market hypothesis holds, discuss random walk behavior, and consider how algorithmic trading can beat market returns amid imperfect efficiency.

Analyze how to measure portfolio risk using daily percent returns, variance, and standard deviation, compare volatility across time periods, and introduce the Sharpe ratio.

Explore the theory and intuition behind the Sortino ratio, showing how it uses downside-only volatility with a zero threshold, unlike the Sharpe ratio that penalizes all volatility.

Implement the Sortino ratio in Python using downside risk and a zero threshold on daily returns from Apple and Microsoft, then compare with the Sharpe ratio and annualize.

Develop the probabilistic Sharpe ratio in Python by computing the daily Sharpe ratio, incorporating skew and fisher kurtosis, using psi pi to compute the cdf, and annualizing the result.

Learn to use SciPy minimize to optimize portfolio weights by minimizing negative Sharpe ratio, with bounds and a sum-to-one constraint, comparing results to Monte Carlo searches.

Explore CAPM with Python by analyzing long-term data, compute cumulative and percent returns, and visualize market relationships with scatter plots to assess beta and alpha against the S&P 500 benchmark.

Compute beta and alpha from capm using Python by performing linear regression on daily returns against the S&P 500, and examine Apple, Amazon, GE, VIX, and leveraged ETFs.

Explore a bank stock capstone: analyze five years of data, compute returns and Sharpe ratios, plot volumes and moving averages, and experiment with Bollinger Bands.

Plot daily and total dollar volume across banks to compare trading activity and identify the peak day for JPMorgan around March 19, 2021.

Explore algorithmic trading basics on the QuantConnect lean engine, using Python tools and financial theory to build and test strategies with buying and selling rules, shorting, leverage, and hands-on exercises.

Learn how common and preferred stock differ in voting and dividends, and how issuances, buybacks, splits, and reverse splits affect ownership and dilution, with adjusted closing prices.

Explore backtesting core concepts by evaluating a historical trading algorithm against benchmarks like the S&P 500 or Nasdaq 100, and interpret metrics such as probabilistic Sharpe ratio and drawdown.

Backtest buying and holding multiple securities on Quant Connect, with a 50/50 Apple and Microsoft split, orders, and rolling statistics. Download results to compare performance.

Learn how to liquidate a position by selling securities at a target price in QuantConnect, exploring hard-coded vs dynamic targets, trailing stop loss, and the mechanics of exiting holdings.

Explore time-based exits in QuantConnect: liquidate after a holding period or price target by tracking invested time and using time differences.

Explore stop market orders on QuantConnect to trigger a market order when a stop price is reached, contrasting with limit orders, and learn through Boeing backtesting how volatility affects fills.

learn to fetch live price and available shares, compute cash on hand, and execute market orders in a quantconnect algorithm, with profit-driven sell logic and on order events.

Test conditional purchasing by comparing real estate ETFs on daily gains, switching 100 percent holdings when one outperforms the other by more than two percent, using QuantConnect backtests.

explore how leverage uses borrowed capital to amplify returns by buying more shares, while understanding risks like margin calls and forced liquidation if minimum capital isn't met.

Explore how to set leverage and brokerage models in QuantConnect, compare cash vs margin accounts, and model reality factors like fees and slippage for algorithmic trading.

Learn the theory of shorting by borrowing shares via a broker, paying a borrowing fee, selling high, and buying back low, with margin calls and the risk of unlimited losses.

Learn how to implement physical shorting in QuantConnect by borrowing shares, entering a short position, and using liquidation to buy back, with margin call risks.

Learn to conduct research and plotting to design trading algorithms and visualize backtest results with Quant Connect charting. Use notebooks for universe selection and charting to screen securities.

Explore candlestick plots by setting the series type to candle to visualize the open, high, low, close, and volume from trade bars, with daily or minutely resolutions and color conventions.

Learn to combine plots in QuantConnect by using the index argument to place multiple series on the same chart or on separate charts, including line and scatter plots.

Modify plot properties in Quant Connect by adjusting the label name, color, and scatter symbols through the series definition, while recognizing its focus on backtesting rather than full visualization.

Explore fundamental data in QuantConnect research notebooks, retrieving Morningstar fundamentals via the data library with security keys, using get fundamental to fetch metrics, and resampling to daily or quarterly resolutions.

Explore technical indicators in QuantConnect research notebooks, learn to call indicators as classes, plot bollinger bands, and understand symbol requirements and documentation navigation.

Create a coarse universe filter for Quant Connect that uses the sorted function and list comprehension to pick the top three most expensive stocks from 8000 equities at daily resolution.