SLAyers' Notes

[dbphoenix]

The "century" number thing has been discussed for decades. But it's an easy hook, most convenient for those who don't use charts, who are after all those who move markets (see yesterday's posts). Consider also our relationship to the July high.

But I found it interesting this morning that today's high and yesterday's high were identical, making the calculation of the DP simple. Once one has that, all he has to do is decide how much risk he's willing to assume and then assume it. Having done that, he'd have four hours to enter at that price before price dropped to 64.

Other than that, given all the meandering, my position is neutral.

Db

 

[YoungBuller]

When price blew through the lower limit, does the next oldest channel become the pertinent one? Would the line of least resistance then become down as price bounces off the upper limit of that next oldest channel from August until about the 2nd week of October? Then it works its way back into the current channel and LOLR becomes up as it bounces off the bottom of it?

 

[dbphoenix]

A year ago, the NQ was slightly overbought. When it corrected this summer, it was oversold (this is of course the case with corrections). Given that it returned to the channel, business may be as usual depending on what the longer-timeframe participants want (see post 235). There's no particular reason to believe that we won't reach the top of the channel, which as of today is just north of 5000.

In the meantime, supply lines will help the trader keep track of what the longer-term people are doing. These cues -- such as if price were to reach 4740+/- -- may provide the trader with attractive trading opportunities.

Db

 

[YoungBuller]

A year ago, the NQ was slightly overbought. When it corrected this summer, it was oversold (this is of course the case with corrections). Given that it returned to the channel, business may be as usual depending on what the longer-timeframe participants want (see post 235). There's no particular reason to believe that we won't reach the top of the channel, which as of today is just north of 5000.

In the meantime, supply lines will help the trader keep track of what the longer-term people are doing. These cues -- such as if price were to reach 4740+/- -- may provide the trader with attractive trading opportunities.

Db

So you're saying the next-oldest channel never came into play? And that after the climatic low, one would be looking to get long to ride out a reversion to the mean?

If price does indeed reach 4740ish, would you still be looking to possibly short despite it being around the channel mean and therefore choppier trading?

 

[dbphoenix]

If longer-term traders saw that price was headed for the median of the old channel, assuming they maintain trend channels, assuming that they maintain charts, then that channel may have come into play. Who knows? But it's not pertinent. What is pertinent is what the trader did, if anything, on August 24th.

As for 4740, what do you plan to do? Why?

Db

 

[YoungBuller]

If longer-term traders saw that price was headed for the median of the old channel, assuming they maintain trend channels, assuming that they maintain charts, then that channel may have come into play. Who knows? But it's not pertinent. What is pertinent is what the trader did, if anything, on August 24th.

As for 4740, what do you plan to do? Why?

Db

Do you place the importance on 8/24 because of the climactic action? Does climactic action like that trump any sort of channel that may be in place? Is there anything wrong with waiting until the 30th or 10/1 to enter on the Hourly given that the entry on the 24th or 26th would've required assuming too much risk for my tolerance?

If I have this right, longer-term traders (aka market movers...right?) don't necessarily keep track of channels or look at charts, but rather we use channels and charts to visually interpret their footprint of mean reversion action. Yay or nay?

At 4740, I'll be looking for a LH and convincing demand line break by more than 10 points on the Hourly to get out of my current long I entered ~4510 and flip to a short based on that supply line you have drawn (whose importance isn't gained by your having drawn it, but rather the footprint of longer-term trader supply coming in repeatedly). If price instead sails through it, I'll be holding my long and trailing for a trip to the top of the channel.

 

[dbphoenix]

1. Price action is in the market. Lines and channels and so forth are in the trader's head.

2. We can track their footprints by the trades they make. Whether or not those trades result in mean reversion will depend on the instrument traded.

3. This evening may be interesting.

 

[YoungBuller]

1. Price action is in the market. Lines and channels and so forth are in the trader's head.

2. We can track their footprints by the trades they make. Whether or not those trades result in mean reversion will depend on the instrument traded.

3. This evening may be interesting.

Indeed it may. Even though price only got to ~4700, would you still be looking for a reversal?

P.S. What tool do you use for those nice red and green dots?

 

[YoungBuller]

A bit of a sidetrack here going back to your original SLA.doc. I have a question about the trend channels you have drawn in the document.

In one, you have this channel:

In another, you have this channel:

And in yet another, you have this channel, which I perceive to be the "correct" one since it begins with the 2009 low:

Which is to say the only real issue I'm having with understanding this fully is deciphering the "correct" trend channel to trade mean reversions. Would you mind if I posted a Weekly chart of the NQ from its inception and we could go through marking the correct channels? I feel as if this would give me a lot more confidence in this method. I can start a new thread if needed.

One final question: Is there ever a time when a mean-reverting market such as the NQ changes and is no longer mean-reverting? I know you said the NQ has been mean-reverting since its inception, but what gives you confidence that it will continue to be in the future? Is it just that the majority of the money in this market is traded by large traders who obviously behave in a mean-reverting manner as a group?

 

[Gring0]

Which is to say the only real issue I'm having with understanding this fully is deciphering the "correct" trend channel to trade mean reversions.

They're all good channels. Instead of working your way inwards, try to work your way outwards from the mean. You might see things differently. The easiest thing is to pick two lows from a bottom of an up trend and a the highest point between those lows for the top of that up channel. At least you'll be consistent. Clarity comes from realizing the mean and moving outwards. This way you can see price even turning a bit earlier or later at the extremes as not being a big thing.

One final question: Is there ever a time when a mean-reverting market such as the NQ changes and is no longer mean-reverting? I know you said the NQ has been mean-reverting since its inception, but what gives you confidence that it will continue to be in the future? Is it just that the majority of the money in this market is traded by large traders who obviously behave in a mean-reverting manner as a group?

You never know, but the exceedingly higher number of stop outs should alert to something being not right. Keep in mind though that the mean reverting market isn't always mean reverting. There has to exist a clear mean first. That same mean is the answer to your first query. AMT looks at the median to determine the extremes. Just step back until things clear up. Staying out happens more often than staying in.

Gringo

Note: The term mean & median is being interchangeably used to show coherence of thought. Both aren't the same thing.

 

[YoungBuller]

They're all good channels. Instead of working your way inwards, try to work your way outwards from the mean. You might see things differently.

Gringo

You mean try to draw the mean before the limits? Are you saying there's not necessarily one "correct" channel but as long as you have something that resembles price revolving around a mean then you should be alright?

 

[dbphoenix]

Indeed it may. Even though price only got to ~4700, would you still be looking for a reversal?

P.S. What tool do you use for those nice red and green dots?

1. According to the SLA, the trade is now short.

2. Paint Shop Pro.

 

[dbphoenix]

A bit of a sidetrack here going back to your original SLA.doc. I have a question about the trend channels you have drawn in the document.

In one, you have this channel:

In another, you have this channel:

And in yet another, you have this channel, which I perceive to be the "correct" one since it begins with the 2009 low:

Which is to say the only real issue I'm having with understanding this fully is deciphering the "correct" trend channel to trade mean reversions. Would you mind if I posted a Weekly chart of the NQ from its inception and we could go through marking the correct channels? I feel as if this would give me a lot more confidence in this method. I can start a new thread if needed.

One final question: Is there ever a time when a mean-reverting market such as the NQ changes and is no longer mean-reverting? I know you said the NQ has been mean-reverting since its inception, but what gives you confidence that it will continue to be in the future? Is it just that the majority of the money in this market is traded by large traders who obviously behave in a mean-reverting manner as a group?

There have been a number of "original" SLA documents, which is why I decided to write the book rather than make post after post about it. If you provide the commentary that went with these charts, I may try to straighten it all out, but I don't know that there would be any point. Straightening it all out is why I wrote "About Back-testing, Hindsight Bias, Curve-Fitting, and Putting The Cart Before The Horse" in Notes.

As for the "correct" trend channel, this is also addressed in the above as well as throughout the book. I also point out that as trendlines don't provide support or resistance, the mean matters more than the exactitude of the trendlines in that the distance traders can and are willing to travel from value is arguably of more importance to the trader than value itself. Gringo addresses at least some of this as well.

As to the NQ and mean-reversion, I'll worry about that if and when the time comes. Most likely I'll be dead so it won't matter. But the NQ is based on stocks, as are other indices, and this whole mean-reversion thing got started with stocks and PEs, so as long as these indices are based on stocks, the indices will continue to revert to the mean.

 

[dbphoenix]

Note: The term mean & median is being interchangeably used to show coherence of thought. Both aren't the same thing.

True. I've addressed this before, but it never hurts to reiterate, boring as it may be.

The "mean" in mean reversion is not a mean. It's a median. The middle. Calculating the mean of every excursion would render the mean useless. Determining the trend channel and the "mean" of it from the first three swing points and leaving it the hell alone provides far more useful information regarding other traders than continuously re-calculating it. Which is also why redrawing the channel lines with every twitch and switch renders them useless as well.

But everybody calls it a mean, so waddayagonna do?

 

[YoungBuller]

They're all good channels. Instead of working your way inwards, try to work your way outwards from the mean. You might see things differently. The easiest thing is to pick two lows from a bottom of an up trend and a the highest point between those lows for the top of that up channel. At least you'll be consistent. Clarity comes from realizing the mean and moving outwards. This way you can see price even turning a bit earlier or later at the extremes as not being a big thing.



You never know, but the exceedingly higher number of stop outs should alert to something being not right. Keep in mind though that the mean reverting market isn't always mean reverting. There has to exist a clear mean first. That same mean is the answer to your first query. AMT looks at the median to determine the extremes. Just step back until things clear up. Staying out happens more often than staying in.

Gringo

Note: The term mean & median is being interchangeably used to show coherence of thought. Both aren't the same thing.

So, more appropriately, it's technically called "median-reversion"? And if there isn't a borderline painfully obvious median, then there's no channel and no compelling trades if one prefers to trade median reversion?

 

[YoungBuller]

As for the "correct" trend channel, this is also addressed in the above as well as throughout the book. I also point out that as trendlines don't provide support or resistance, the mean matters more than the exactitude of the trendlines in that the distance traders can and are willing to travel from value is arguably of more importance to the trader than value itself. Gringo addresses at least some of this as well.

I understand that completely. The channels are there to alert the trader to "overbought" and "oversold" conditions. A reversal short taken an extended distance away (above) from value is more likely to pay off and pay bigger than one taken near or below value. The key is to recognize that price is relatively extended away from value and in an "overbought" condition whether I'm looking at the 1st channel, 2nd channel, or 3rd channel. Yes?

 

[dbphoenix]

The only channel that matters is the one in which one is trading. If he's doing five-year-old replay, then the pertinent channel is the one in force at the time. The one that is to be used now is the one that is now in effect.

We left "value" a month ago, which provided the cue for examining the daily and subsequently shorting the reversal, either off the daily or the hourly. That short is gone, as is the subsequent long. The short after that is still in effect, and trading will continue until the trader exits the market according to the rules, at which point he'll look for another cue.

 

[YoungBuller]

The only channel that matters is the one in which one is trading. If he's doing five-year-old replay, then the pertinent channel is the one in force at the time. The one that is to be used now is the one that is now in effect.

We left "value" a month ago, which provided the cue for examining the daily and subsequently shorting the reversal, either off the daily or the hourly. That short is gone, as is the subsequent long. The short after that is still in effect, and trading will continue until the trader exits the market according to the rules, at which point he'll look for another cue.

1. I'm with you.

2. By "we left value a month ago", I'm assuming you mean the bounce off the median at ~4730. Are you referring to a reversal off that median and the "subsequent long" would be the bounce from ~4455?

 

[dbphoenix]

Yes.

 

[YoungBuller]

Yes.

Why would you short a reversal off the mean? I thought you only traded at Extremes? Or are you just including all possible hypotheticals?