Major Changes

UPDATE: Some of this is incorrect

There are a few changes which will be made to The Bubble Index website; changes which will require massive computation. So, over the next few weeks there will be fewer daily updates.

Changes include:

  • Adding search functionality in the navigation bar
  • The standardization of every Bubble Index
  • Creating a standard set of windows (52 days, 153 days, 256 days, 512 days, 1260 days, 1764 days, 2520 days, and 5040 days) for all stocks, indices, commodities, and currencies.
The last two changes will allow every Bubble Index for a given window to be comparable with each other. In other words, say I want to compare the relative herding in two securities — The Bubble Index: GOOG (512 days) with The Bubble Index: DJIA (512 days). This will be possible after the changes are complete. The standardization will set 100.00 as the level of The Bubble Index on October 23, 1929 for all windows.
So, if The Bubble Index: GOOG (512 days) is at 18.65 today, then I would know that herding behavior is only 18.65% of that seen in The Bubble Index: DJIA (512 days) on October 23, 1929, several days before the Wall Street Crash of 1929.

As of today, the issue with The Bubble Index for a single security is that the index levels are relative to their own values. This hinders the usefulness of time series with shorter time horizons, like Tesla (TSLA) with only three years of history. In addition, on time series with long time horizons, like the DJIA, the massive difference in absolute price affects the level of the index (not the shape and structure, just the relative size of the “spikes”). This is bothersome and I should have corrected this a long time ago. Example of the normalization change in the DJIA (RED is the non-normalized curve):