Question & Answer Follow-up
By
Rennie on Tuesday, November 24th, 2009 at 9:24 pm
I didn’t get to a number of questions asked during last Friday’s chat at The Kirk Report, so I thought I’d take this opportunity to respond to as many as possible.
I’d love to hear your thoughts on the continued validity of the NYSE TICK with so much volume gone from the NYSE onto the electronic exchanges…
This question was asked more than once, so I thought I’d start here. I don’t think the fact that volume in NYSE stocks is now spread among multiple exchanges has much, if any effect on the NYSE TICK. Remember what the TICK is – the difference between the number of NYSE stocks ticking higher or lower at a particular moment in time. A reading of +500, for instance, indicates 500 more NYSE stocks ticking higher at that moment. If a large buy order hits an NYSE stock, the price is likely to tick higher on all exchanges that trade the stock regardless of which one received the order, which in turn will have a positive effect on the TICK. In other words, the TICK indicator is solely dependent on price, not volume. And while volume may be spread among multiple exchanges, the price of a stock will be virtually identical on all exchanges.
CTA’s got on board quickly in 2003, wonder why so late to the game this year?
This is in reference to the Market Vane survey of commodity trading advisors. I think there was a lot of disbelief earlier this year regarding the rally’s lasting power. While it’s true this ‘smart money’ group didn’t catch the initial turnaround earlier this year, I think that can be forgiven considering they correctly moved to a bearish posture in mid-2008, thereby avoiding all of the late-2008 carnage. In my opinion that’s much more important than catching the bottom.
Why is recent SPY volume tricky?
This is in reference to a remark about end-of-day volume in general and SPY volume in particular being potentially misleading. For exchange volume like the NYSE, the reason is obvious – they’re only trading about 20% of the volume in NYSE-listed issues, so you’re not seeing the whole picture when looking at floor volume. This also has a major effect on figures such as up/down volume percentages and TRIN, which are based on this same floor volume figure. ETF volume is also tricky in that there are an increasingly large number of ETF’s that are essentially going after the same crowd. QQQQ volume used to be a pretty effective indicator when there weren’t so many other ETF’s competing for the same volume. In the last couple of years, previously reliable short-term signals based on QQQQ volume stopped working, and I’m seeing a similar pattern emerge with SPY volume.
What is your usual hold time frame to be in for a trade, hour, days or weeks or longer?
I try to trade on multiple time frames, from intraday to long-term. That said, I’ve found my most reliable signals are on a longer-term time frame, from weeks to months. Short-term trading is seductive and endlessly fascinating, but not easy! From my experience, I think short-term ‘continuation’ signals (those going with the underlying trend) are more reliable and easier to trade than countertrend signals.
Do you think the aversion stage of the sentiment cycle is in front of us or in the rear view mirror?
Just looking at a chart, I think you’d have to say it’s in front of us, but I expect it will play out more in the form of a trading range than a deep pullback. The current rally is looking a lot like 2003.
Didn’t S&P 500 close the 11/16 gap yesterday? If so, does that change your conclusion in any way?
This is in reference to the ‘two upside gaps in one month’ setup I’ve discussed previously and that went into effect on November 16th. No, the fact that the recent gap was filled doesn’t alter the outlook. It’s the fact that the first gap on 11/9 went unfilled and was subsequently followed by another gap on 11/16 that’s bullish.
When analyzing spreadsheet data, I find that the graphics engine of Excel is often inadequate. Do you have an alternative that you recommend?
Well, as subscribers know our charts aren’t the greatest :) but we’re working on that! I currently use Excel along with Metastock.
With some breadth measures (like new 20-day highs minus new 20-day lows or Worden’s T2108), some folks view new lows in breadth unaccompanied by new lows in the index price as a positive divergence since the bad breadth can not impact price proportionately. Others see the same as weakness bubbling below the surface with bearish implications. How do you view these situations?
Generally I view breadth as the lead indicator, so I would view new lows in breadth as having longer-term bearish implications. But I’m also cognizant of the fact that divergences between breadth and the S&P can continue for an extended period of time. So yes, there are definitely periods when price can shrug off weak breadth stats – a good example is the 1999 period. Notice from the long-term chart that breadth topped out well ahead of the market, back in early ’98, while the S&P continued higher for another two years. Of course that’s an extreme example. I would add that on that long-term chart, when cumulative breadth has broken out to a new 10-year high (green arrows), it was usually a longer-term bullish development looking out 1-2 years. We came very close to registering a fresh 10-year high recently but pulled back. This is something to watch in my opinion.
Can you tell us the steps you take to prepare for the next day?
From the time I return from lunch until well after the close I’m usually busy preparing for the next day. This primarily involves plugging in current data as if it were closing figures and checking on what signals could be triggered at that day’s close. I also make it a point of just observing that day’s action from multiple perspectives to see if anything strikes me as unusual, and of course reading various financial blogs and news sources to see what others are discussing. If anything looks interesting I’ll investigate further.
Rennie long time subscriber here and I love this opportunity to real time chat with you. One question on how you actually do your analysis. You take the data from something like pinnacle and then you run them through spreadsheets or some kind of statistical software like SAS, etc?
Thanks! Yes, I use a series of Excel spreadsheets to run my analysis. For me the most important issue is to have all of the data organized and ready to work with. In this way, when something occurs I can immediately go to the spreadsheets and find previous occurrences with minimal effort. This allows me to quickly test an idea and see at a glance whether or not it’s worth pursuing.
Have you tried using interest rates or the dollar in your analysis?
Absolutely. Looking at 7-year highs as sell signals and 15-month lows as buy signals in the 90-day T-Bill rate used to be an excellent market timing tool (from Dick Stoken’s Strategic Investment Timing book), but it last gave a buy in early 2001 that’s still in effect. It almost gave a sell in the 2006-2007 period if rates had just held up a little longer. That was one of the more effective means I’ve seen of applying interest rate movement to the stock market. I’ve never found a consistently reliable means of applying U.S. dollar data, but if someone else has I’d love to hear about it!
What level is an extreme TICK reading?
Well, if you pull up an intraday chart of the NYSE TICK you’ll see the +1000/-1000 levels generally define ‘extreme’ territory, but there are other ways to gauge what’s extreme. Many of these methods involve moving averages and a defined lookback period to adjust to recent market action, so that when the TICK exceeds relatively high or low levels for the period, it’s considered significant even if we’re not seeing traditionally extreme readings.
Great session. What time frame(s) have you found helpful using the $TICK data?
I find it most useful on a 1-minute or 5-minute time frame. With enough of that data, you can create a number of interesting daily indicators as well. For instance, the Cumulative TICK chart that I find useful is a daily chart, but it takes 1-minute data to create. In general I think this is a very promising area of new and useful indicators (those derived from intraday data but not necessarily charted intraday).
That answers just about all of the questions received during Friday’s chat. If I didn’t get to your question, please feel free to send me an email!
Question & Answer Follow-up
By Rennie on Tuesday, November 24th, 2009 at 9:24 pmI didn’t get to a number of questions asked during last Friday’s chat at The Kirk Report, so I thought I’d take this opportunity to respond to as many as possible.
I’d love to hear your thoughts on the continued validity of the NYSE TICK with so much volume gone from the NYSE onto the electronic exchanges…
This question was asked more than once, so I thought I’d start here. I don’t think the fact that volume in NYSE stocks is now spread among multiple exchanges has much, if any effect on the NYSE TICK. Remember what the TICK is – the difference between the number of NYSE stocks ticking higher or lower at a particular moment in time. A reading of +500, for instance, indicates 500 more NYSE stocks ticking higher at that moment. If a large buy order hits an NYSE stock, the price is likely to tick higher on all exchanges that trade the stock regardless of which one received the order, which in turn will have a positive effect on the TICK. In other words, the TICK indicator is solely dependent on price, not volume. And while volume may be spread among multiple exchanges, the price of a stock will be virtually identical on all exchanges.
CTA’s got on board quickly in 2003, wonder why so late to the game this year?
This is in reference to the Market Vane survey of commodity trading advisors. I think there was a lot of disbelief earlier this year regarding the rally’s lasting power. While it’s true this ‘smart money’ group didn’t catch the initial turnaround earlier this year, I think that can be forgiven considering they correctly moved to a bearish posture in mid-2008, thereby avoiding all of the late-2008 carnage. In my opinion that’s much more important than catching the bottom.
Why is recent SPY volume tricky?
This is in reference to a remark about end-of-day volume in general and SPY volume in particular being potentially misleading. For exchange volume like the NYSE, the reason is obvious – they’re only trading about 20% of the volume in NYSE-listed issues, so you’re not seeing the whole picture when looking at floor volume. This also has a major effect on figures such as up/down volume percentages and TRIN, which are based on this same floor volume figure. ETF volume is also tricky in that there are an increasingly large number of ETF’s that are essentially going after the same crowd. QQQQ volume used to be a pretty effective indicator when there weren’t so many other ETF’s competing for the same volume. In the last couple of years, previously reliable short-term signals based on QQQQ volume stopped working, and I’m seeing a similar pattern emerge with SPY volume.
What is your usual hold time frame to be in for a trade, hour, days or weeks or longer?
I try to trade on multiple time frames, from intraday to long-term. That said, I’ve found my most reliable signals are on a longer-term time frame, from weeks to months. Short-term trading is seductive and endlessly fascinating, but not easy! From my experience, I think short-term ‘continuation’ signals (those going with the underlying trend) are more reliable and easier to trade than countertrend signals.
Do you think the aversion stage of the sentiment cycle is in front of us or in the rear view mirror?
Just looking at a chart, I think you’d have to say it’s in front of us, but I expect it will play out more in the form of a trading range than a deep pullback. The current rally is looking a lot like 2003.
Didn’t S&P 500 close the 11/16 gap yesterday? If so, does that change your conclusion in any way?
This is in reference to the ‘two upside gaps in one month’ setup I’ve discussed previously and that went into effect on November 16th. No, the fact that the recent gap was filled doesn’t alter the outlook. It’s the fact that the first gap on 11/9 went unfilled and was subsequently followed by another gap on 11/16 that’s bullish.
When analyzing spreadsheet data, I find that the graphics engine of Excel is often inadequate. Do you have an alternative that you recommend?
Well, as subscribers know our charts aren’t the greatest :) but we’re working on that! I currently use Excel along with Metastock.
With some breadth measures (like new 20-day highs minus new 20-day lows or Worden’s T2108), some folks view new lows in breadth unaccompanied by new lows in the index price as a positive divergence since the bad breadth can not impact price proportionately. Others see the same as weakness bubbling below the surface with bearish implications. How do you view these situations?
Generally I view breadth as the lead indicator, so I would view new lows in breadth as having longer-term bearish implications. But I’m also cognizant of the fact that divergences between breadth and the S&P can continue for an extended period of time. So yes, there are definitely periods when price can shrug off weak breadth stats – a good example is the 1999 period. Notice from the long-term chart that breadth topped out well ahead of the market, back in early ’98, while the S&P continued higher for another two years. Of course that’s an extreme example. I would add that on that long-term chart, when cumulative breadth has broken out to a new 10-year high (green arrows), it was usually a longer-term bullish development looking out 1-2 years. We came very close to registering a fresh 10-year high recently but pulled back. This is something to watch in my opinion.
Can you tell us the steps you take to prepare for the next day?
From the time I return from lunch until well after the close I’m usually busy preparing for the next day. This primarily involves plugging in current data as if it were closing figures and checking on what signals could be triggered at that day’s close. I also make it a point of just observing that day’s action from multiple perspectives to see if anything strikes me as unusual, and of course reading various financial blogs and news sources to see what others are discussing. If anything looks interesting I’ll investigate further.
Rennie long time subscriber here and I love this opportunity to real time chat with you. One question on how you actually do your analysis. You take the data from something like pinnacle and then you run them through spreadsheets or some kind of statistical software like SAS, etc?
Thanks! Yes, I use a series of Excel spreadsheets to run my analysis. For me the most important issue is to have all of the data organized and ready to work with. In this way, when something occurs I can immediately go to the spreadsheets and find previous occurrences with minimal effort. This allows me to quickly test an idea and see at a glance whether or not it’s worth pursuing.
Have you tried using interest rates or the dollar in your analysis?
Absolutely. Looking at 7-year highs as sell signals and 15-month lows as buy signals in the 90-day T-Bill rate used to be an excellent market timing tool (from Dick Stoken’s Strategic Investment Timing book), but it last gave a buy in early 2001 that’s still in effect. It almost gave a sell in the 2006-2007 period if rates had just held up a little longer. That was one of the more effective means I’ve seen of applying interest rate movement to the stock market. I’ve never found a consistently reliable means of applying U.S. dollar data, but if someone else has I’d love to hear about it!
What level is an extreme TICK reading?
Well, if you pull up an intraday chart of the NYSE TICK you’ll see the +1000/-1000 levels generally define ‘extreme’ territory, but there are other ways to gauge what’s extreme. Many of these methods involve moving averages and a defined lookback period to adjust to recent market action, so that when the TICK exceeds relatively high or low levels for the period, it’s considered significant even if we’re not seeing traditionally extreme readings.
Great session. What time frame(s) have you found helpful using the $TICK data?
I find it most useful on a 1-minute or 5-minute time frame. With enough of that data, you can create a number of interesting daily indicators as well. For instance, the Cumulative TICK chart that I find useful is a daily chart, but it takes 1-minute data to create. In general I think this is a very promising area of new and useful indicators (those derived from intraday data but not necessarily charted intraday).
That answers just about all of the questions received during Friday’s chat. If I didn’t get to your question, please feel free to send me an email!