Consolidated Volume Highlights Extent of NYSE Market Share Loss
By
Rennie on Thursday, August 20th, 2009 at 2:21 am
It’s no secret that the NYSE has been losing market share to NASDAQ and newcomer BATS, but many may still be surprised at the extent of the loss. In January 2007, the NYSE traded roughly 55% of all NYSE-listed shares. A year later that percentage was down to 40%. This past January it was down to 25%, and just recently we’ve seen a number of sessions in which the NYSE traded less than 20% of total share volume. See this chart for an illustration.
This has important implications for a variety of volume-related indicators, such as up/down volume and TRIN. As the shift away from the floor continues, these indicators will become more and more misleading as they analyze an increasingly small slice of the market. A longer-term spread like the 200-day moving average of NYSE up-down volume can still be a useful tool, especially when paired with its NASDAQ counterpart. But when you hear about a 90% down volume day, that statistic is based only on volume traded on the NYSE. While a consolidated version of this indicator (if it existed) would be similar, it’s impossible to say what the true percentage would be with any certainty.
TRIN suffers from a similar problem. The Arms Index is calculated by dividing the advance/decline ratio by the up/down volume ratio. Because the up/down volume component only takes into account the relatively small amount of volume that actually traded on the floor, TRIN no longer providing an accurate picture of overall market activity. There have been clues that the indicator was broken for some time. Under normal circumstances, TRIN should oscillate above and below the 1.0 level, but in recent years the 20-day average held above 1.0 in both bull and bear markets (see long-term chart).
A favorite sentiment indicator, the Nasdaq/NYSE Volume Ratio, is also affected by the drop-off in NYSE market share. I’ve pointed to this indicator a number of times after the 20-day moving average began to hold in extreme territory (above 1.50) in mid-2007, an area that has historically related to longer-term tops in the market. Unfortunately, it’s become increasingly difficult to rely on this indicator (despite its recent accuracy) given the dwindling amount of volume actually traded on the NYSE. I downloaded consolidated volume figures for the NYSE and NASDAQ back to 2007 and created a consolidated Nasdaq/NYSE Volume Ratio. The result is significantly different from the chart we’ve been posting. See this chart for an illustration (consolidated version is in red, standard version is in green). Note that unlike the standard Nasdaq/NYSE Volume Ratio, the consolidated version has actually been trending the other direction, moving steadily lower from March ’07 through March ’09. It rebounded along with the market from March to July but has since resumed its downtrend. In a nutshell, the consolidated version is showing none of the ‘excessive speculative participation’ seen on the standard version.
The reason is fairly clear from looking at this chart of NYSE volume. The standard version, which is the figure reported in the Wall Street Journal and elsewhere, has been dragging along at multi-year lows for some time now. But that’s misleading. Actual volume in NYSE-listed shares across all exchanges has actually been trending higher over the last 2+ years. Even at its low point last month, consolidated volume (seen in red) was well above average levels in 2007, a major difference from the standard version.
On a short-term basis, one of the more noteworthy developments Wednesday was the fact that S&P futures gapped down at the open and rallied to close above Tuesday’s high. The table below lists the last thirty instances in which the front-month S&P futures contract opened below the previous day’s low and settled above the previous day’s high…
S&P Futures Open <Prev Day’s Low, Close >Prev Day’s High
08/19/09… ???
05/26/09… Lower S&P close one session later
04/01/09… No lower close in next 1-2 days
09/11/08… Lower S&P close two sessions later
07/22/08… Lower S&P close two sessions later
05/06/08… Lower S&P close one session later
05/30/07… Lower S&P close one session later
11/02/05… No lower close in next 1-2 days
10/19/05… Lower S&P close one session later
08/29/05… Lower S&P close one session later
08/18/04… Lower S&P close one session later
03/05/04… Lower S&P close one session later
07/24/02… Lower S&P close one session later
12/19/01… Lower S&P close one session later
10/25/01… Lower S&P close one session later
05/04/01… Lower S&P close one session later
01/10/01… No lower close in next 1-2 days
08/03/00… No lower close in next 1-2 days
06/13/00… No lower close in next 1-2 days
05/05/00… Lower S&P close one session later
04/12/99… Lower S&P close one session later
02/12/98… Lower S&P close one session later
12/05/97… Lower S&P close one session later
02/24/97… Lower S&P close two sessions later
01/21/97… Lower S&P close two sessions later
01/10/97… Lower S&P close one session later
05/08/96… Lower S&P close one session later
01/26/96… No lower close in next 1-2 days
11/28/95… Lower S&P close two sessions later
01/25/95… No lower close in next 1-2 days
08/09/94… Lower S&P close two sessions later
Note that in 23 out of 30 cases, or 77% of the time, the S&Ps posted a lower close within the next 1-2 days as buying power often hit a short-term exhaustion point. That’s not quite enough of an edge over the 59% at-any-time odds to qualify for inclusion on the board, but it’s close.
Consolidated Volume Highlights Extent of NYSE Market Share Loss
By Rennie on Thursday, August 20th, 2009 at 2:21 amIt’s no secret that the NYSE has been losing market share to NASDAQ and newcomer BATS, but many may still be surprised at the extent of the loss. In January 2007, the NYSE traded roughly 55% of all NYSE-listed shares. A year later that percentage was down to 40%. This past January it was down to 25%, and just recently we’ve seen a number of sessions in which the NYSE traded less than 20% of total share volume. See this chart for an illustration.
This has important implications for a variety of volume-related indicators, such as up/down volume and TRIN. As the shift away from the floor continues, these indicators will become more and more misleading as they analyze an increasingly small slice of the market. A longer-term spread like the 200-day moving average of NYSE up-down volume can still be a useful tool, especially when paired with its NASDAQ counterpart. But when you hear about a 90% down volume day, that statistic is based only on volume traded on the NYSE. While a consolidated version of this indicator (if it existed) would be similar, it’s impossible to say what the true percentage would be with any certainty.
TRIN suffers from a similar problem. The Arms Index is calculated by dividing the advance/decline ratio by the up/down volume ratio. Because the up/down volume component only takes into account the relatively small amount of volume that actually traded on the floor, TRIN no longer providing an accurate picture of overall market activity. There have been clues that the indicator was broken for some time. Under normal circumstances, TRIN should oscillate above and below the 1.0 level, but in recent years the 20-day average held above 1.0 in both bull and bear markets (see long-term chart).
A favorite sentiment indicator, the Nasdaq/NYSE Volume Ratio, is also affected by the drop-off in NYSE market share. I’ve pointed to this indicator a number of times after the 20-day moving average began to hold in extreme territory (above 1.50) in mid-2007, an area that has historically related to longer-term tops in the market. Unfortunately, it’s become increasingly difficult to rely on this indicator (despite its recent accuracy) given the dwindling amount of volume actually traded on the NYSE. I downloaded consolidated volume figures for the NYSE and NASDAQ back to 2007 and created a consolidated Nasdaq/NYSE Volume Ratio. The result is significantly different from the chart we’ve been posting. See this chart for an illustration (consolidated version is in red, standard version is in green). Note that unlike the standard Nasdaq/NYSE Volume Ratio, the consolidated version has actually been trending the other direction, moving steadily lower from March ’07 through March ’09. It rebounded along with the market from March to July but has since resumed its downtrend. In a nutshell, the consolidated version is showing none of the ‘excessive speculative participation’ seen on the standard version.
The reason is fairly clear from looking at this chart of NYSE volume. The standard version, which is the figure reported in the Wall Street Journal and elsewhere, has been dragging along at multi-year lows for some time now. But that’s misleading. Actual volume in NYSE-listed shares across all exchanges has actually been trending higher over the last 2+ years. Even at its low point last month, consolidated volume (seen in red) was well above average levels in 2007, a major difference from the standard version.
On a short-term basis, one of the more noteworthy developments Wednesday was the fact that S&P futures gapped down at the open and rallied to close above Tuesday’s high. The table below lists the last thirty instances in which the front-month S&P futures contract opened below the previous day’s low and settled above the previous day’s high…
S&P Futures Open <Prev Day’s Low, Close >Prev Day’s High
08/19/09… ???
05/26/09… Lower S&P close one session later
04/01/09… No lower close in next 1-2 days
09/11/08… Lower S&P close two sessions later
07/22/08… Lower S&P close two sessions later
05/06/08… Lower S&P close one session later
05/30/07… Lower S&P close one session later
11/02/05… No lower close in next 1-2 days
10/19/05… Lower S&P close one session later
08/29/05… Lower S&P close one session later
08/18/04… Lower S&P close one session later
03/05/04… Lower S&P close one session later
07/24/02… Lower S&P close one session later
12/19/01… Lower S&P close one session later
10/25/01… Lower S&P close one session later
05/04/01… Lower S&P close one session later
01/10/01… No lower close in next 1-2 days
08/03/00… No lower close in next 1-2 days
06/13/00… No lower close in next 1-2 days
05/05/00… Lower S&P close one session later
04/12/99… Lower S&P close one session later
02/12/98… Lower S&P close one session later
12/05/97… Lower S&P close one session later
02/24/97… Lower S&P close two sessions later
01/21/97… Lower S&P close two sessions later
01/10/97… Lower S&P close one session later
05/08/96… Lower S&P close one session later
01/26/96… No lower close in next 1-2 days
11/28/95… Lower S&P close two sessions later
01/25/95… No lower close in next 1-2 days
08/09/94… Lower S&P close two sessions later
Note that in 23 out of 30 cases, or 77% of the time, the S&Ps posted a lower close within the next 1-2 days as buying power often hit a short-term exhaustion point. That’s not quite enough of an edge over the 59% at-any-time odds to qualify for inclusion on the board, but it’s close.