A DEEP LOOK AT THE SP500

Snapshot

I want to start by looking at a classic “snapshot” that most sophisticated investors or longer-term traders would try to view before making a decision.

What is included are the Cumulative Returns, Drawdowns, and Daily Returns. Most important is that we are using a large dataset with daily data starting on January 3, 1928, to our current date, which is nearly 100 years of data. The data is from a specialist 3rd party provider. The Python library I am using is quantstats written by Ran Aroussi. This is a very powerful library, I don’t think it gets the attention it deserves.

Key Performance Metrics

This is a very comprehensive report. Ran has done a terrific job with quantstats but it would have been so nice to be able to control the reports in a format different from HTML. Anyway, I spent the afternoon working with the source code and adapting the metrics into tables that I have more control over.

The metrics included below may be a bit of overkill, but I would like to urge you to take the time and spend a few minutes going through all the metrics in the report.

Why I am stressing this is because most experts understand all the characteristics of the subject they are an expert in. Learning from these metrics provides one with an understanding of the SP500’s statistical personality.

Monthly Returns with a Heatmap

The next table is a very helpful visual deep dive into the monthly returns making up the SP500 return history for nearly 100 years.

What is nice about looking at the different numbers with the colour coding highlighting the size of the moves is that you can see that the numbers are not constants. Looking at statistical averages does not tell the whole story of the journey. The journey is sometimes uneventful, and other times it is wild and dangerous. Understanding what has happened over nearly 100 years gives you an incredible map of what is possible. It is also important to know that just because something has never happened in the past doesn’t mean it cannot happen in the future. So one needs to be on the lookout for black swans.

Equity Performance Charts

I am going to throw in quite a few variations of the main equity chart to really get a better understanding of performance. I will probably come back and explain in more detail some of the charts, but I am keen to progress this post.

I am pretty sure the most confusion will be directed at the classic Cumulative Returns chart and the Cumulative Returns (Log Scaled).

This all sounds very technical, but you really don’t need to concern yourself with the fancy math. The first Cumulative Returns chart is the intuitive one we are all familiar with. It measures the performance of your investment over time, from the beginning until the end with all the compounded returns. If you look at the scale of the y-axis (vertical), it goes up in 5k% increments which are equally spaced.

If you look at the Cumulative Returns with Log Scaling it modifies the y-axis so that every % increase is comparable. So a 100% increase in the beginning of the history will have the same size on the y-axis if it happens later in the history.

In summary, the log scaling technique allows one to analyse % moves in a more consistent way through time. Some people are absolutely fanatical about this when reading charts. I don’t have firm views.