Note on BKX Outperformance
By
Rennie on Monday, November 30th, 2009 at 3:20 pm
With the BKX up 2% heading into the final 30 minutes of trade (and the S&P unchanged), I thought this study posted last month would be of interest. While it doesn’t look like it will be triggered (S&P needs to close up 0.5%+), conceptually it’s interesting that in recent times, BKX outperformance has generally been a short-term negative sign. The table below lists the last thirty sessions in which the BKX closed up 2% or more, the SPX closed up 0.5% or more and the BKX tripled the S&P gain. Note that in 27 occurrences, or 90% of the time, the S&P posted a subsequently lower close (below the setup day’s close) within the next three days…
BKX +2.0%, SPX +0.5%, BKX Triples SPX Gain
10/22/09… Lower S&P close one session later
09/22/09… Lower S&P close one session later
08/13/09… Lower S&P close one session later
06/18/09… Lower S&P close two sessions later
06/11/09… Lower S&P close two sessions later
06/04/09… Lower S&P close one session later
05/14/09… Lower S&P close one session later
05/08/09… Lower S&P close one session later
05/06/09… Lower S&P close one session later
05/04/09… Lower S&P close one session later
04/23/09… No lower close within three days
04/21/09… Lower S&P close one session later
04/15/09… Lower S&P close three sessions later
04/09/09… Lower S&P close two sessions later
04/03/09… Lower S&P close one session later
03/31/09… No lower close within three days
03/25/09… Lower S&P close three sessions later
03/18/09… Lower S&P close one session later
02/24/09… Lower S&P close one session later
02/11/09… Lower S&P close two sessions later
02/06/09… Lower S&P close two sessions later
01/28/09… Lower S&P close one session later
01/23/09… No lower close within three days
01/21/09… Lower S&P close one session later
11/28/08… Lower S&P close one session later
11/25/08… Lower S&P close three sessions later
09/19/08… Lower S&P close one session later
09/18/08… Lower S&P close three sessions later
09/16/08… Lower S&P close one session later
09/08/08… Lower S&P close one session later
Again, this is a recent phenomenon only. Prior to November ‘07, you have to look back over four years just to find the next occurrence (on 07/14/03). Then you have to look back another two years to 06/21/01 to find the next. It clearly wasn’t the kind of pattern you see every day. And back then, this pattern didn’t work particularly well. Looking back at the previous 30 signals prior to August ‘08, we find much more average performance. Of the 30 signals between 2000 and August ‘08, accuracy fell to 73% in terms of calling for a lower S&P within three days. That’s only in-line with random odds, indicating that it wasn’t until the recent market turmoil that unusually strong up days for the BKX have represented reliable sell signals.
Note on BKX Outperformance
By Rennie on Monday, November 30th, 2009 at 3:20 pmWith the BKX up 2% heading into the final 30 minutes of trade (and the S&P unchanged), I thought this study posted last month would be of interest. While it doesn’t look like it will be triggered (S&P needs to close up 0.5%+), conceptually it’s interesting that in recent times, BKX outperformance has generally been a short-term negative sign. The table below lists the last thirty sessions in which the BKX closed up 2% or more, the SPX closed up 0.5% or more and the BKX tripled the S&P gain. Note that in 27 occurrences, or 90% of the time, the S&P posted a subsequently lower close (below the setup day’s close) within the next three days…
BKX +2.0%, SPX +0.5%, BKX Triples SPX Gain
10/22/09… Lower S&P close one session later
09/22/09… Lower S&P close one session later
08/13/09… Lower S&P close one session later
06/18/09… Lower S&P close two sessions later
06/11/09… Lower S&P close two sessions later
06/04/09… Lower S&P close one session later
05/14/09… Lower S&P close one session later
05/08/09… Lower S&P close one session later
05/06/09… Lower S&P close one session later
05/04/09… Lower S&P close one session later
04/23/09… No lower close within three days
04/21/09… Lower S&P close one session later
04/15/09… Lower S&P close three sessions later
04/09/09… Lower S&P close two sessions later
04/03/09… Lower S&P close one session later
03/31/09… No lower close within three days
03/25/09… Lower S&P close three sessions later
03/18/09… Lower S&P close one session later
02/24/09… Lower S&P close one session later
02/11/09… Lower S&P close two sessions later
02/06/09… Lower S&P close two sessions later
01/28/09… Lower S&P close one session later
01/23/09… No lower close within three days
01/21/09… Lower S&P close one session later
11/28/08… Lower S&P close one session later
11/25/08… Lower S&P close three sessions later
09/19/08… Lower S&P close one session later
09/18/08… Lower S&P close three sessions later
09/16/08… Lower S&P close one session later
09/08/08… Lower S&P close one session later
Again, this is a recent phenomenon only. Prior to November ‘07, you have to look back over four years just to find the next occurrence (on 07/14/03). Then you have to look back another two years to 06/21/01 to find the next. It clearly wasn’t the kind of pattern you see every day. And back then, this pattern didn’t work particularly well. Looking back at the previous 30 signals prior to August ‘08, we find much more average performance. Of the 30 signals between 2000 and August ‘08, accuracy fell to 73% in terms of calling for a lower S&P within three days. That’s only in-line with random odds, indicating that it wasn’t until the recent market turmoil that unusually strong up days for the BKX have represented reliable sell signals.