Volume is the oxygen of the market, or so they say. 

What has been noticeable about this year up until the sell-off was that as the market was rising, volume was falling. 

Until it sold off violently, that is.  Since the break on February 27, volume has picked up, and it has been heavier on down days than on up days.

As you can see in my snazzy graph, volume of the S&P 500 - the blue bars - has generally been higher when the S&P 500 - the red line - has gone down. 

Since the market broke on February 27, there have been 16 trading days, including the 27th and today. The average daily volume of the stocks in the S&P 500 has been 1.45 billion. There have been 9 days when the market has gone up and 7 days when the market has fallen. 

On days when the market has risen, volume has averaged 1.31 billion.  On only 3 days has the market risen on volume higher than the previous day.  The average volume for those 3 days has been 1.31 billion.  (I won't have today's volume numbers on the S&P 500 until tomorrow, but total volume on the Big Board was up slightly compared to yesterday, so I assumed it was up slightly for the S&P 500 today.  I'll finalize the data tomorrow.)  On the 6 declining days, volume has also averaged 1.31 billion.  However, when you look at the change in volume compared to the day before, when the market has risen on heavier volume, the increase in volume has averaged 0.04 billion shares, or not much.  On the six days the market has risen on falling volume, the average decline in volume has been 0.25 billion shares.  In total, for the 9 days when the market has risen, on average the volume change compared to the previous day has been -0.15 billion shares.  Or, in other words, the market has been rising on lower volume.

In contrast, there have been 7 days when the market has fallen.  On those 7 days, the average volume has been 1.62 billion shares, or 23% more volume than the days when the market has risen.  There have been 4 days when the market fell on rising volume.  On average, daily volume was 1.77 billion shares for those 4 days.  On the 3 days when the market was falling on declining volume, the average volume was 1.41 billion.  When the market has fallen on rising volume, the average change in volume compared to the previous day has been 0.43 billion shares.  On the days when the market has fallen on declining volume, the average change has been -0.16 billion.  On all 7 days when the market has fallen, volume has on average increased by 0.18 billion shares compared to the previous day.  Or, in other words, the market has been falling on heavier volume.

Of the 9 up days, there were 3 days when the average daily volume was greater than 1.41 billion,  one of which was the day after the original break, and another was on March 6 when 1.42 billion shares traded. Seven of the 10 and 6 of the 8 heaviest volume days were down days.  The market's best two days during this stretch have been the last two days, which have also been the two lowest volume days since the sell-off began.

Last night, I came across this exert from the book, How Legendary Traders Made Millions, which am I currently reading. It is regarding Gerald Loeb and the market in 1937.

Beginning the second week of March, the market experienced some heavy selling action.  From March 8 through March 22, the market sold off four times on heavy volume in little over two weeks.  On April 2 and 7, it did the same thing - sold off hard on heavy volume.  It was here that Gerald Loeb began to see a parallel to the 1929 market - the market began experiencing sell-offs in heavy volume after a strong run-up.  He noticed that many leading stocks were topping and the market was having trouble advancing.  it was also a time when the market had just staged an incredible run-up and the country was in the beginning of a newly formed recession.  He knew from his experiences that the market needs time to digest strong gains after years of almost uninterrupted advances.  He also listened to his intuition, and he decided to trust his judgment.  He therefore cashed out completely and once again was 100% in cash when the market broke wide open.  A sharp bounce back in August back to near its high for the year didn't last long though and actually turned out to be a "suckers rally." ...

From August through November the market sold off hard, dropping from near 185 to 130, or nearly 30%.  it regained its footing a bit in December and finished the year at 135.  It was still a 25% negative return following two strong years. 

Just thought it was interesting.