PDA

View Full Version : 11-3-11 Close Comments



Mike
11-03-2011, 09:06 PM
I have been quiet for a couple of days. I do some CANSLIM teaching each week which takes time away but the real reason is that the Market School I attended in Las Vegas has began a flurry of activity building an automated market model. What I found at Market School which I have already commented on was a market timing model that fits my trading style. It is designed to get me in when it is appropriate, keep me in when it is appropriate and take me out when it is appropriate. This means when the market is rallying that I stay in for the length of the rally. This model has navigated the choppy waters of the current market so well that my hat is off to Bill O'Neil and his merry band of portfolio managers who seemed to have caputured the essence of O'Neil in a rule-based model.

The model went to +4 market exposure count two days ago from +5 and back up to +5 today which equates to a drop back to 90% invested two days ago and a green light for 100% in after today's market action. Today's signal was an additional FTD (up more than 1.25% on higher volume than yesterday) bringing the count to +5.

Essentially the market has been in an undeniable rally choppy as it is. The leadership has firmed up and many stocks have rallied in a way that you can buy them at propper buy points, not get shaken out, and then hold them to harvest points. Essentailly a CANSLIM rally that has been working.

The flip side of the coin is the action at the 200-day moving average. This has stopped the last five bear markets in their tracks with a subsequent move to the downside. We have been visiting this level for a while. In history this resolves itself fairly quickly. We either take the high road and move up above the 200-day and stay there or we continue down for the next leg down. My personal opinion is that we will take the low road but my opinion means little. Price and Volume will lead the way and keep us sane. In IBD tomorrow they keep the market Under Pressure.

Riskslayer
11-04-2011, 06:49 AM
Hi Mike,

I have found your last several comments extremely useful and interesting. Please keep them coming!

Your positions from your Oct 29 note (SWI, SBH, UA, ISRG, GLD, DLTR, GNC) are doing great!

I am trying to follow along the model counts. I understand the traditional CANSLIM criterion around FTD and distribution days (DD's), but had some Q's about the new concepts:

1. Do we continue to count DD's? So, we keep an Exposure Count (EC) and DD count (DDC). It seems from your Oct 24 note that this is the case .. and it relates to B10... Under B10, if the DDC is above 4, then we increase the EC after a DD falls off (25d window)? But, under traditional CANSLIM, the rallied ended after 4 DD's? It seems like the EC and DDC are maintained even when their is no rally condition in place (as defined under traditional CANSLIM)? e.g. your Oct 24 note shows Oct 10 B3 event and EC =1 despite Buy Switch OFF.

2. Does B3 event only increase EC following a FTD and only the 1st occurrence? Can the 1st FTD (starting a traditional CANSLIM rally) be used to trigger both B1 and B3? e.g. $COMPQX on Oct 18, both an FTD and B3 event? Then Oct 19 on $COMPQX does not increase EC even though no intraday excursions below 20d ema? Is the B3 event reset only following a Buy Switch move to off?

3. Are B4, B5 & B6 events similar and only can increment EC once per Buy Switch ON? similarly reset? Can B6 occur and increment EC on same day as FTD, B1?

4. Or, is it the case that the FTD and related rules, B1, B2 and B7-B10 occur and are counted toward subsequent day EC so long as Buy Switch is ON, whereas B3-B6 are examined each day and add to EC only if they are true for that day?

5. I tend to put significantly more weight on $COMPQ than NYSE, so is it appropriate to only run the analysis against $COMPQ? I would think that we do not want to count a FTD on NYSE, put Buy Switch to ON, and then run the EC and DDC based on $COMPQ? right? Oct 24 Note says under certain conditions Buy Switch can be ON, but EC = 0, so long as the uptrend is still intact - how is "intacted" measured? What takes the Buy Switch to OFF?

I have some Sell signal Q's, but I will hold those for now :)

Thanks for all your help!

Shawn

Mike
11-04-2011, 09:41 AM
Hi Mike,

I have found your last several comments extremely useful and interesting. Please keep them coming!

Your positions from your Oct 29 note (SWI, SBH, UA, ISRG, GLD, DLTR, GNC) are doing great!

I am trying to follow along the model counts. I understand the traditional CANSLIM criterion around FTD and distribution days (DD's), but had some Q's about the new concepts:

1. Do we continue to count DD's? So, we keep an Exposure Count (EC) and DD count (DDC). It seems from your Oct 24 note that this is the case .. and it relates to B10... Under B10, if the DDC is above 4, then we increase the EC after a DD falls off (25d window)? But, under traditional CANSLIM, the rallied ended after 4 DD's? It seems like the EC and DDC are maintained even when their is no rally condition in place (as defined under traditional CANSLIM)? e.g. your Oct 24 note shows Oct 10 B3 event and EC =1 despite Buy Switch OFF.

2. Does B3 event only increase EC following a FTD and only the 1st occurrence? Can the 1st FTD (starting a traditional CANSLIM rally) be used to trigger both B1 and B3? e.g. $COMPQX on Oct 18, both an FTD and B3 event? Then Oct 19 on $COMPQX does not increase EC even though no intraday excursions below 20d ema? Is the B3 event reset only following a Buy Switch move to off?

3. Are B4, B5 & B6 events similar and only can increment EC once per Buy Switch ON? similarly reset? Can B6 occur and increment EC on same day as FTD, B1?

4. Or, is it the case that the FTD and related rules, B1, B2 and B7-B10 occur and are counted toward subsequent day EC so long as Buy Switch is ON, whereas B3-B6 are examined each day and add to EC only if they are true for that day?

5. I tend to put significantly more weight on $COMPQ than NYSE, so is it appropriate to only run the analysis against $COMPQ? I would think that we do not want to count a FTD on NYSE, put Buy Switch to ON, and then run the EC and DDC based on $COMPQ? right? Oct 24 Note says under certain conditions Buy Switch can be ON, but EC = 0, so long as the uptrend is still intact - how is "intacted" measured? What takes the Buy Switch to OFF?

I have some Sell signal Q's, but I will hold those for now :)

Thanks for all your help!

Shawn


Shawn,

We keep seperate Exposure Counts and Distribution Counts. The Distribution Counts relate to sell signals S3 and S4 and buy signal B10. The normal way that we exit the market is with more and more S3 and S4 signals decrementing the Exposure Count along with any other sell signals that occur. On the day the Exposure Count is reduced to zero (portfolio in cash) AND the Distribution Count is full (6 or higher) the Buy Switch is turned off. The Buy Switch can also be turned off with the S2 signal which is an undercut of the entire rally. When the Buy Switch is turned off we are waiting for a new Follow Through Day to move back into the market. A FTD turns the Buy Switch on. We can have buy signals prior to a FTD such as a B3 signal (lows above the 21-day ema) but we can not act on them until the FTD or B1 signal. At the FTD we we act on all of the signals since the Rally Day or day 1 of the advance.

When you read about ending market rallies with high distribution counts you need to keep the year in mind. Right now and since 2005 full distribution count is 6 days. From 1991-2004 it was 5 and prior to 1991 it was 4. The increasing volatility of the market has led to greater tollerance to distribution counts.

The distribution count is reset to zero on the initial FTD or B1 signal. This occured on the NASDAQ on 10/18/11. We have had three days of distribution since this date. We count all distribution days in a 25 day window or since the FTD. We also drop off distribution days out of the count if the market intraday high exceeds the close of a distribution day by 6%. If the market moves up that far that particular distribution day has less meaning to the whole picture. I have attached the 11/02 Market School chart. Since 11/02 we have had one more signal the 23rd day additional FTD or B2 signal that occured yesterday. The exposure count is now +5 fully invested. We do not count B2 signals after the 25th day of the advance. B7 days or Accumulation Day signals are then counted. An Accumulation Day B7 signal is a FTD-tyoe day with extra requirements. Volume must be above average and we must be at a 13-week high and close in the upper 25% of the range.

A B3 signal (lows above the 21-day) comes when ever it comes and sometime before the FTD. You can not count another B3 signal in the advance unless an S5 signal (closing below the 21-day) occurs. After the S5 reset another B3 signal can be counted. This reset by S5 works the same with B3, B4, B5 signals. On the flip side of the market S5, S6, S7, S8 signals are reset by a B3 signal. This means you cannot count additional S5,6,7,8 signals until the B3 reset.

Bill O'Neil and his portfolio managers place the most weight on the NASDAQ market. The procedure they use when a FTD occurs on a different index such as the NYSE Composite is to pay attention to what ever index had the FTD but shift over to the NASDAQ as soon as they can. Procedurely they wait until the exposure count on the NASDAQ matches the exposure count on what ever index had the FTD. At this point they move to the NASDAQ for the remainder of the rally. They will still pay attention to other markets for signs of weakness to decide if more caution is advised. Also none of this obviates the need to pay attention to what the leading stocks are doing.

I think I addressed your questions but if I missed something ask again.

11305