Category:

Data Analysis

Create a Personal Portfolio/Wealth Simulation in Python (Part 2)

by Stuart Jamieson 18 July 2021

Welcome to Part 2 of the series of posts dealing with how to build your own python based personal portfolio /wealth simulation model. At the end of the first post (which can be found here), we got to the point where we had modelled some inflows, some outflows, we had applied an annual salary raise to our future income flows, along with applying various tax rates to both our active income (salary) and investment income.

We had also factored in a stochastic element to our model for generating investment returns, using the historic mean monthly return and volatility of the S&P Total Return index as a proxy for the “market”. Finally, we had ended by adding a couple of lines of code that would record whether our wealth/asset value were reduced to (or below) zero at any point throughout the simulated time period; we will be using this later to help calculate our “risk of ruin”, i.e. attach probabilities to the likelihood of ending up with assets worth less than a threshold value (one that we deem unacceptable to fall below – which doesn’t have to be zero of course).

The full code snippet from the end of Part 1 is shown below for convenience (I have altered some of the outflow and inflow values back to their original levels just fyi). IMPORTANT: I have also had to include the use of the yfinance package to allow the over-riding of the Pandas DataReader package as at the time of writing, Pandas DataReader is currently not stable and essentially not working. The code changes are commented below to help identify them more easily.

18 July 2021 0 comment

Create a Personal Portfolio/Wealth Simulation in Python

by Stuart Jamieson 13 June 2021

written by Stuart Jamieson

This post will introduce the first part (of multiple) where we build up a personal finance model to help simulate future time periods based on certain chosen input variables. We will input variables such as our current investable asset base, our annual salary, expected monthly inflows and outflows and a range of other relevant values. Firstly, after our necessary imports, we look to start on modelling inflows over time.

import pandas as pd
import numpy as np
import random
import pandas_datareader.data as web
import datetime
import matplotlib as mpl
import matplotlib.pyplot as plt
%matplotlib inline
plt.rcParams['figure.figsize'] = (12,6)

# Set up model inflows
inflows = {'active_annual_income':50_000}
variables = {'start_date' : "01/01/2020",
             'years': 10}
income_gains_storage = []
months = variables['years'] * 12

for month in range(months):
    income = inflows['active_annual_income'] / 12
    income_gains_storage.append(income)
plt.plot(pd.Series(income_gains_storage).cumsum())
plt.show()

This looks like a pretty simplistic results, the cumulative income we have earned over 10 years, at 50k annual salary is unsurprisingly 10 x 50,000 = 500,000. One thing we might want to introduce is some logic to account for the fact the nasty tax man gets his hands on at least some (way too much) of it. For the moment, for simplicity’s sake, I will assume a flat tax rate of 25%. Of course most countries have a sliding scale of income bands which are subject to increasing rates of tax the more one earns. We can come back to this later and see if we can implement something more along those lines – but for now lets stick with our flat 25%.

13 June 2021 9 comments

Data Analysis Portfolio Optimisation

Black-Litterman Portfolio Allocation Model in Python

by Stuart Jamieson 27 November 2020

written by Stuart Jamieson

A while ago I posted an article titled “INVESTMENT PORTFOLIO OPTIMISATION WITH PYTHON – REVISITED” which dealt with the process of calculating the optimal asset weightings for a portfolio according to the classic Markowitz “mean-variance” approach. With this method we aim to maximise our level of return for any given level of risk, in doing so we develop the concept of an “efficient frontier” and usually seek to identify the point/portfolio on that frontier which represents the best trade-off between risk and return (i.e. the portfolio with the highest Sharpe Ratio).

As nicely as the model allows us to identify our supposed “optimal asset weightings”, there are several, rather severe problems we face when using mean-variance optimisation.

The model assumes asset returns are normally distributed.
It can generate unintuitive, highly-concentrated portfolios.
The inputs to the model include each individual asset’s predicted/expected return and volatility, but we of course can never know those values for certain. Common practice is to calculate the assets’ historic returns and standard deviations and use them as proxies – this makes the massive presumption that all the assets will continue to behave and perform just as they have done in the past. Not only does it assume the returns and volatility will remain the same, it also assumes that correlations between all the assets in question will remain stable through time. We know these assumptions are just not realistic.
To make matters worse, not only do we use inputs that are reliant on our own “best guess” or forecasts, the model happens to be extremely sensitive to variations in these input values (especially to the return inputs, less so the volatility inputs). If the input values are changed, even by relatively small amounts, the optimal portfolio weightings created by the model can swing and vary wildly. Ideally we would like our model to be as “robust” as possible in this regard and generate stable/slowly changing suggested asset weightings when faced with changing input values.

In this post I am going to take a look at the Black-Litterman model, which is an adaptation of the classic mean-variance framework which enables investors to combine their unique views regarding the performance of various assets with the market equilibrium in a manner that results in intuitive, diversified portfolios.

27 November 2020 9 comments

Data Analysis Data Wrangling

Build a Financial Data Database with Python

by Stuart Jamieson 24 October 2020

written by Stuart Jamieson

Hi all, and welcome back to the site – I appreciate it has been an unexpectedly long time since I last posted…in fact my last post was around this time last year. Hopefully I can get back on the “treadmill” and churn out some articles at a somewhat faster rate than 1 a year over the next couple of months! Well that’s my aim anyway.

Ok so this post will be based on how to build and structure what is often referred to as a “data pipeline”; essentially it is the part of the overall project workflow concerned with the gathering and storing of data, along with any wrangling/munging/pre-processing/transforming of that data for later use.

Ideally this endeavour should aim to produce a code module which is robust, efficient, reusable, scalable and maintainable. It should also aim to produce a well structured, easily accessible store of “high-quality”, ready to use financial data sets and series.

I thought I’d start by sketching out a rough mind-map style representation of the concepts and ideas I want to cover and the basic points I want to make; this is shown below.

24 October 2020 20 comments

Data Analysis Trading Strategy Backtest

Equities Market Intraday Momentum Strategy in Python – Part 1

by Stuart Jamieson 23 October 2019

written by Stuart Jamieson

For this post, I want to take a look at the concept of intra-day momentum and investigate whether we are able to identify any positive signs of such a phenomenon occurring across (quite a large) universe of NYSE stocks. It has been suggested that, for the wider market in general at least, there is a statistically significant intra-day momentum effect resulting in a positive relationship between the direction of returns seen during the first half an hour of the trading day (taking the previous day’s closing price as the “starting value”) and the last half an hour of the day’s session. That is to say, it may be that a stock/index which displays a positive return early in the trading session, will be more likely to experience a positive return over the last part of the session.

The effect seems to have been first identified/posited by Gao, Han, Li and Zhou in their 2015 research paper (https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2440866). In their research paper, they specifically look at high-frequency data regarding the S&P 500 ETF, and they test over 20 years’ worth of data – so it’s worth pointing out that I am going the “other way” somewhat. Where their study lacked depth (number of instruments studied), my data contains around 3000 individual stocks, however, where they tested over a long time period (20 years) my data spans only 1 year.

I have a feeling already that the mechanisms and forces that move the “overall market” and result in certain price patterns and behaviours, may not necessarily translate exactly over into the individual constituent stocks. We can but try…

23 October 2019 3 comments