T2 Beta Test
By
Rennie on Tuesday, September 22nd, 2009 at 9:38 pm
The S&P closed in new highs for the year Tuesday as nearly 250 issues hit new 52-week highs. While seasonally bullish tendencies are in effect until 2pm ET, the fresh six-month high for the S&P suggests we could be in for a choppy finish Wednesday (see this September 9th column for a recent discussion). The S&P continues to hold the ‘overbought breakout’ trendline, which stands right around SPX 1066.50 on Wednesday. TICKscore closed at +12, Cumulative TICK +75,000, the thirteenth consecutive session with net positive NYSE TICK action.
VXO closed in new lows for the year at 21.96, and history suggests we’re likely to see a further drop in the VXO on Wednesday. Volatility indexes generally trade lower on FOMC announcement days. The table below highlights the VXO’s performance from open to close during each of the last thirty regularly scheduled FOMC days…
VXO Performance on FOMC Announcement Day
09/23/09… VXO ??? from open to close
08/12/09… VXO -3.4% from open to close
06/24/09… VXO -1.7% from open to close
04/29/09… VXO -2.6% from open to close
03/18/09… VXO -2.6% from open to close
01/28/09… VXO -1.6% from open to close
12/16/08… VXO -7.8% from open to close
10/29/08… VXO +2.0% from open to close
09/16/08… VXO -9.9% from open to close
08/05/08… VXO -9.0% from open to close
06/25/08… VXO -2.4% from open to close
04/30/08… VXO +2.1% from open to close
03/18/08… VXO -13.0% from open to close
01/30/08… VXO -4.1% from open to close
12/11/07… VXO +12.0% from open to close (*)
10/31/07… VXO -13.3% from open to close
09/18/07… VXO -21.3% from open to close
08/07/07… VXO -6.2% from open to close
06/28/07… VXO -3.0% from open to close
05/09/07… VXO -8.6% from open to close
03/21/07… VXO -12.0% from open to close
01/31/07… VXO -8.4% from open to close
12/12/06… VXO -3.3% from open to close
10/25/06… VXO -1.6% from open to close
09/20/06… VXO -1.4% from open to close
08/08/06… VXO -2.0% from open to close
06/29/06… VXO -17.5% from open to close
05/10/06… VXO -1.9% from open to close
03/28/06… VXO +0.6% from open to close
01/31/06… VOX +1.4% from open to close
12/13/05… VXO -9.3% from open to close
Note that in only one case out of the last thirty did the VXO gain 3% or more from the open, while it lost 3% or more sixteen times. This tendency remains consistent throughout all FOMC announcements since the beginning of 1993. Invariably, the VXO will trend sideways or head convincingly lower off the open.
On Monday, we quietly launched the latest incarnation of our original TrendCatcher strategy. T2 is quite different from the original TrendCatcher in terms of its construction, but the intention is the same – identify a solid trend day as early as possible and ride it into the close. The past two years have provided a challenging environment for this sort of mechanical strategy, and we’ve run through many, many variations of the original concept. In the end, a breakthrough came with the invention of the T2 indicator, a simple but effective means of measuring intraday buy/sell activity that combines aspects of our two primary NYSE TICK indicators – TICKscore and Cumulative TICK.
This is the first trend day strategy we tested that managed to perform well during relatively calm periods as well as during the extreme volatility of last fall. It’s not easy to call for a higher (lower) S&P at the end of the day following an early morning rally (decline) and be right two-thirds of the time (see track record). Over the next few months, we’ll see if it can keep it up. I plan on reporting on the indicator each morning via the T2 page on our site as well as on Twitter (for the time being), and will most likely continue to track it in this manner for the rest of the year. If it continues to perform as expected, we’ll drop the Twitter updates and move it solely to the subscriber site. Email alerts will be sent out when a signal is triggered. We’ll also have an intraday chart of T2 available on the intraday snapshot page as soon as we work out some technical issues.
One of the key components of the strategy is time. The strongest trends usually start early in the session and persist. There’s only one actual input, the T2 indicator. Over the past 26 months, there have been 137 signals, so it only averages about 5 trades a month. Average profit only slightly exceeded the average loss (1.2% vs 1.1%), but there were twice as many winning signals as losing signals (92 vs. 45). That’s pretty impressive considering the signal is triggered in the first hour of trading and remains in effect until the close (with no stop). Adding in some sort of fail-safe stop could improve results, and is something I’ll look into over the coming days. Here is a look at the equity curve – note that the maximum drawdown was limited to 7% over the past 26 months. It trades relatively infrequently, which may frustrate some active traders, but for now this appears to be one of the more promising methods of mechanically identifying trend days that are likely to persist.
T2 Beta Test
By Rennie on Tuesday, September 22nd, 2009 at 9:38 pmThe S&P closed in new highs for the year Tuesday as nearly 250 issues hit new 52-week highs. While seasonally bullish tendencies are in effect until 2pm ET, the fresh six-month high for the S&P suggests we could be in for a choppy finish Wednesday (see this September 9th column for a recent discussion). The S&P continues to hold the ‘overbought breakout’ trendline, which stands right around SPX 1066.50 on Wednesday. TICKscore closed at +12, Cumulative TICK +75,000, the thirteenth consecutive session with net positive NYSE TICK action.
VXO closed in new lows for the year at 21.96, and history suggests we’re likely to see a further drop in the VXO on Wednesday. Volatility indexes generally trade lower on FOMC announcement days. The table below highlights the VXO’s performance from open to close during each of the last thirty regularly scheduled FOMC days…
VXO Performance on FOMC Announcement Day
09/23/09… VXO ??? from open to close
08/12/09… VXO -3.4% from open to close
06/24/09… VXO -1.7% from open to close
04/29/09… VXO -2.6% from open to close
03/18/09… VXO -2.6% from open to close
01/28/09… VXO -1.6% from open to close
12/16/08… VXO -7.8% from open to close
10/29/08… VXO +2.0% from open to close
09/16/08… VXO -9.9% from open to close
08/05/08… VXO -9.0% from open to close
06/25/08… VXO -2.4% from open to close
04/30/08… VXO +2.1% from open to close
03/18/08… VXO -13.0% from open to close
01/30/08… VXO -4.1% from open to close
12/11/07… VXO +12.0% from open to close (*)
10/31/07… VXO -13.3% from open to close
09/18/07… VXO -21.3% from open to close
08/07/07… VXO -6.2% from open to close
06/28/07… VXO -3.0% from open to close
05/09/07… VXO -8.6% from open to close
03/21/07… VXO -12.0% from open to close
01/31/07… VXO -8.4% from open to close
12/12/06… VXO -3.3% from open to close
10/25/06… VXO -1.6% from open to close
09/20/06… VXO -1.4% from open to close
08/08/06… VXO -2.0% from open to close
06/29/06… VXO -17.5% from open to close
05/10/06… VXO -1.9% from open to close
03/28/06… VXO +0.6% from open to close
01/31/06… VOX +1.4% from open to close
12/13/05… VXO -9.3% from open to close
Note that in only one case out of the last thirty did the VXO gain 3% or more from the open, while it lost 3% or more sixteen times. This tendency remains consistent throughout all FOMC announcements since the beginning of 1993. Invariably, the VXO will trend sideways or head convincingly lower off the open.
On Monday, we quietly launched the latest incarnation of our original TrendCatcher strategy. T2 is quite different from the original TrendCatcher in terms of its construction, but the intention is the same – identify a solid trend day as early as possible and ride it into the close. The past two years have provided a challenging environment for this sort of mechanical strategy, and we’ve run through many, many variations of the original concept. In the end, a breakthrough came with the invention of the T2 indicator, a simple but effective means of measuring intraday buy/sell activity that combines aspects of our two primary NYSE TICK indicators – TICKscore and Cumulative TICK.
This is the first trend day strategy we tested that managed to perform well during relatively calm periods as well as during the extreme volatility of last fall. It’s not easy to call for a higher (lower) S&P at the end of the day following an early morning rally (decline) and be right two-thirds of the time (see track record). Over the next few months, we’ll see if it can keep it up. I plan on reporting on the indicator each morning via the T2 page on our site as well as on Twitter (for the time being), and will most likely continue to track it in this manner for the rest of the year. If it continues to perform as expected, we’ll drop the Twitter updates and move it solely to the subscriber site. Email alerts will be sent out when a signal is triggered. We’ll also have an intraday chart of T2 available on the intraday snapshot page as soon as we work out some technical issues.
One of the key components of the strategy is time. The strongest trends usually start early in the session and persist. There’s only one actual input, the T2 indicator. Over the past 26 months, there have been 137 signals, so it only averages about 5 trades a month. Average profit only slightly exceeded the average loss (1.2% vs 1.1%), but there were twice as many winning signals as losing signals (92 vs. 45). That’s pretty impressive considering the signal is triggered in the first hour of trading and remains in effect until the close (with no stop). Adding in some sort of fail-safe stop could improve results, and is something I’ll look into over the coming days. Here is a look at the equity curve – note that the maximum drawdown was limited to 7% over the past 26 months. It trades relatively infrequently, which may frustrate some active traders, but for now this appears to be one of the more promising methods of mechanically identifying trend days that are likely to persist.