Market Volatility in Up vs Down Trends

10Aug08

This is a follow up to World Beta’s excellent post on “volatility clustering”

A lot has been made (example) of the fact that the big days we’ve been experiencing lately are far more common in down than up trends (perhaps implying that this market is still headed south).  World Beta’s (WB) post shows that the likelihood of both big up and big down moves increases in downtrends because something else increases in downtrends too – volatility.  According to WB, between 1929 and 2008, down markets were 72% more volatile that up markets.

Here we put our own spin on WB’s results to show whether their conclusions have consistently held true over the last 50+ years.

The above graph shows a 5-year rolling average of the volatility (standard deviation) of daily returns during uptrends (green) versus downtrends (red) from 1956 to the present on the S&P 500.  I’ve defined up/down trends as whether or not the S&P 500 is above/below its 200-day moving average.

These rolling averages have been further divided into positive (dark) and negative (light) volatility.  By positive/negative volatility I mean the volatility of up days versus down day returns in each period.

WB’s conclusions have been mostly consistent over the last 50+ years…

  1. Volatility in downtrends (red) has run consistently higher than volatility in uptrends (green).
  2. Positive (dark) and negative (light) volatility has moved in lockstep.  When the propensity for big up days increases, so does the propensity for big down days (and vice-versa).

I’ll add one additional observation: I don’t see a strong connection between the average volatility of uptrends versus downtrends at any given moment, meaning, just because the current trend has been more or less volatile might not mean the subsequent opposing trend will be as well.  The market seems to “reset” itself at trend changes.

In conclusion: when the market is in a downtrend, both positive and negative volatility increases leading to more big up AND big down days compared to when in an uptrend.  This observation has been consistent over the last 50+ years.  Additionally, an increase in volatility in the current trend doesn’t necessarily carry over to the following trend.

Happy Trading,
ms

 

To stay up to date with what’s happening at the MarketSci Blog, we recommend subscribing to our RSS Feed or Email Feed.

Geek notes: (1) daily returns have been capped at +/- 5 standard deviations to mitigate the influence of single large days and (2) all averages are geometric.



One Response to “Market Volatility in Up vs Down Trends”


  1. 1 Monday links: volatility clustering « Abnormal Returns

Leave a Reply

Please log in using one of these methods to post your comment:

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