For what it's worth -
Discussions on Str-Squ pops up now and then and I thought I'd share an observation.
When Str-Squ was introduced I really tried to wrap my mind round it (finally an indicator ... of sorts). I soon realized that I couldn't make any sense of the histogram though (I used qcharts for this). Sometimes it seemed to work but equally often - not. After some frustration I thought that I would need a way to go back at least a few minutes and debrief. Finally I started to plot the value second by second instead of using the histogram.
<img src=http://www.elitetrader.com/vb/attachment.php?s=&postid=1715654>
Well... no improvement there. Str-squ still seemed to exhibit a rather erratic behavior. Even strong 'signals' as some shown in the image above would sometimes be true and sometimes not. So thinking I was going further down the rabbit hole than I should I dropped Str-Squ for some time.
Then it dawned on me... the indicator works but the feed is the problem. Str-Squ depends on data from different sources, and the feed from those sources have to be in sync with each other. If they get out of sync (one lags more or less than the other) you'll see a false signal. Note that this is independent of the lag of the feed to your computer. Retail feeds like qcharts even throttle ticks at times when market volumes pick up, which is sure to add to the erratic behavior of Str-Squ. There's more to this but I'll not delve into it here (off topic already).
So it's important not only to see the strength of the signal in terms of value, it is equally important to see strength in terms of time. The longer a stretch or squeeze lasts, the greater the probability that the signal is really a signal (well, almost anyway). Like the one at 10:53 above (which incidentally was not a signal if I remember correctly
. As Spydertrader keeps saying, you must know when to look).
cheers,
tobbe
Discussions on Str-Squ pops up now and then and I thought I'd share an observation.
When Str-Squ was introduced I really tried to wrap my mind round it (finally an indicator ... of sorts). I soon realized that I couldn't make any sense of the histogram though (I used qcharts for this). Sometimes it seemed to work but equally often - not. After some frustration I thought that I would need a way to go back at least a few minutes and debrief. Finally I started to plot the value second by second instead of using the histogram.
<img src=http://www.elitetrader.com/vb/attachment.php?s=&postid=1715654>
Well... no improvement there. Str-squ still seemed to exhibit a rather erratic behavior. Even strong 'signals' as some shown in the image above would sometimes be true and sometimes not. So thinking I was going further down the rabbit hole than I should I dropped Str-Squ for some time.
Then it dawned on me... the indicator works but the feed is the problem. Str-Squ depends on data from different sources, and the feed from those sources have to be in sync with each other. If they get out of sync (one lags more or less than the other) you'll see a false signal. Note that this is independent of the lag of the feed to your computer. Retail feeds like qcharts even throttle ticks at times when market volumes pick up, which is sure to add to the erratic behavior of Str-Squ. There's more to this but I'll not delve into it here (off topic already).
So it's important not only to see the strength of the signal in terms of value, it is equally important to see strength in terms of time. The longer a stretch or squeeze lasts, the greater the probability that the signal is really a signal (well, almost anyway). Like the one at 10:53 above (which incidentally was not a signal if I remember correctly
. As Spydertrader keeps saying, you must know when to look).cheers,
tobbe