Data Mining Millions Of Books For Financial Insight

The Google Ngram project is a mesmerizing database of books in various languages. It provides the ability to search through all these works in just a flash, making it an easy and irresistible source for insight into the patterns of change within financial markets.

This is similar to the data provided by Google trends, another rich vein of data I’ve mined before. But unlike the pattern of search words in Google, Ngrams is a bit dated since there is considerable lag between research, writing and publication of a book. And the most recent data in the Ngrams database is for 2008.

Below are the results that I got by playing with this massive database. For all of the charts, the settings used were:

  • Time period: 1800-2008 (maximum)
  • Keywords and phrases: See respective chart
  • Corpus: English
  • Smoothing: 3

The first pattern I searched for was a sign of the financial crisis that we had gone through recently and how it compared to previous ones:

It is difficult to tell because of the scaling but the term “credit crisis” actually peaked in the 1930’s. I’m sure if the database were to include books written from 2008-2010 that would be very different. And references to “stock market crash” peaked after the 1987 Black Monday crash.

Another concept widely debated both this year and last year was whether we were about to see an inflationary or deflationary effect. I wish we had more recent data because my hunch is that deflation would spike on the chart:


Long time readers will probably remember my repeated arguments for the deflationary position. With the passage of time, this has been confirmed as the correct call.

Next I compared the prevalence of “trust fund” with “mutual fund”. Most are not aware that the term ‘mutual fund’ was a reinvention of the first and it was a marketing gimmick done after the Great Depression to entice investors to return to equity investing.

The re-naming of this financial product was an attempt to mollify a traumatized public. Even decades later, people had vivid recollections of the damage wrought by ‘trust funds’ (the name of the most common investment vehicles at the time) after the crash of 1929 and the ensuing bear market:

After a few decades of getting used to mutual funds and even rushing into them during the great bull market that started in the early 1980’s, investors discovered index funds which provide very low expense along with diversification. And then came ETFs which added the convenience of instantaneous trading and gained both the following of retail and institutional traders:

The move away from mutual funds or index funds towards ETFs (exchange traded funds) goes hand in hand with the shortening of attention spans and a preference for short-term trading gains rather than long term gains:

The first hedge funds was first created by Alfred Winslow Jones in 1949 with $100,000 of capital (40% of which was his own). A few years later in 1952 it came to resemble the modern hedge fund by switching to a limited partnership, charging a 20% performance fee and using both leverage and long/short hedging strategies.

It wasn’t until 1966 when an article by Fortune highlighted the performance of Jones’ fund that this new investment vehicle gained interest from Wall Street and savvy investors. Behold, the rise of the hedge fund!

Check out Google Ngram for yourself and let me know if you find something interesting.

Advertisements
This entry was posted in Uncategorized. Bookmark the permalink.

One Response to Data Mining Millions Of Books For Financial Insight

  1. Pingback: Rydex Leveraged ETF Ratio: A Market Top In Sight | tradersnarrative

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s