Trend Following-Trend Commandments

Just for the record. Here is my definition of a trend.

Trend is a general description of a sustainable slope of any interpolator plotted on a set of historical data.

There are no objective methods to predict the begging of a trend or its end. It's just a convenient non-formal verbalization of an observed tendency. Trends could only be useful in a presentation of the past data to acknowledge a DC component in it (i.e. BIAS).
 
Price Phsycis is the most well articulated theory of market movement. Every theory needs some lengthy explanation, and after reading all of the tools in the CFA curriculum there isn't a better theory for price action than that.

Jack Hershey's system can be tested now, but I haven't heard anyone reporting on its' performance.

I have other strategies like pairs trading, which with the old version turns $10k into less than $150k, and the new version $10k to $535k. Same values, different sell discipline. It's night and day.

When I look at strategies, the performance metrics are usually self-evident. It's very easy to tell with enough quantitative training if a strategy is worthwhile or not.

I don't see why Mr. Covel won't share some of his price action theories and strategies. Most of the threads appear to extoll the virtues of trend following, but without any discussion other than who is a trend follower there isn't really any room to analyze results. I'm one of those people that accepts output from models readily, provided I don't see any problems with the backtests. I find analyzing results is a better way to approach the viability of trading strategies, and, until I have some investors, there will continue to be realistic gaps in some of my trading activities.

Mr. Covel as a sponsor shouldn't have any problem posting performance summaries or even trade lists. The possibility of reverse engineering a strategy from price histories is very remote, and I consider it a theory only, not something that can actually be done. This might be what is keeping him from posting but you can't get anything to reverse engineer from just performance summaries so if he's concerned about that and that is probably too much credit to him then he doesn't necessarily have to post a trades list.

I don't know why he wouldn't do that anyway.
 
Quote from bwolinsky:

Price Phsycis is the most well articulated theory of market movement. Every theory needs some lengthy explanation, and after reading all of the tools in the CFA curriculum there isn't a better theory for price action than that.

Jack Hershey's system can be tested now, but I haven't heard anyone reporting on its' performance.

I have other strategies like pairs trading, which with the old version turns $10k into less than $150k, and the new version $10k to $535k. Same values, different sell discipline. It's night and day.

When I look at strategies, the performance metrics are usually self-evident. It's very easy to tell with enough quantitative training if a strategy is worthwhile or not.

I don't see why Mr. Covel won't share some of his price action theories and strategies. Most of the threads appear to extoll the virtues of trend following, but without any discussion other than who is a trend follower there isn't really any room to analyze results. I'm one of those people that accepts output from models readily, provided I don't see any problems with the backtests. I find analyzing results is a better way to approach the viability of trading strategies, and, until I have some investors, there will continue to be realistic gaps in some of my trading activities.

Mr. Covel as a sponsor shouldn't have any problem posting performance summaries or even trade lists. The possibility of reverse engineering a strategy from price histories is very remote, and I consider it a theory only, not something that can actually be done. This might be what is keeping him from posting but you can't get anything to reverse engineer from just performance summaries so if he's concerned about that and that is probably too much credit to him then he doesn't necessarily have to post a trades list.

I don't know why he wouldn't do that anyway.
More...

I categorically oppose any discussions of actual results as they have very little to prove or disprove. I would rather focus on the theoretical premise without any mentioning of the actual trading results. Even if somebody posts the results there is no objective way to verify that those results were actually achieved by using a method in question or not.
 
Quote from MAESTRO:

I think the best way to start "debugging" the myth of trends is to start reading these types of books:

http://www.amazon.com/Optical-Illusions-Science-Perception-Illusion/dp/1554071518/ref=pd_sim_sbs_b_1

http://www.amazon.ca/Seeing-Deceiving-Psychology-Visual-Illusions/dp/0470265221
More...

Good point, though I'm not quite so sure that sustained shifts in supply/demand that unfold as trends don't exist, myself.

While I haven't read any of Covel's pamphlets, I hear he dropped out of the Turtle experiment put together by Dennis and Eckhardt, who were successful trend followers. As far as I know, their methods weren't based on analyzing visual trend patterns so much as being poised to exploit sustained movements while cutting losses in unfavorable conditions.

One of the former turtles out there actually published the simple systematic rules in a pdf floating around out there for free.
 
Quote from MAESTRO:

Just for the record. Here is my definition of a trend.

Trend is a general description of a sustainable slope of any interpolator plotted on a set of historical data.

There are no objective methods to predict the begging of a trend or its end. It's just a convenient non-formal verbalization of an observed tendency. Trends could only be useful in a presentation of the past data to acknowledge a DC component in it (i.e. BIAS).
More...

Well, yes there is an objective method to do that, but it isn't mysticism, it's pure quantitative analysis. When you look at a chart, you have only the enclosed spreadsheet to use to define trend, and this is what Price Physics teaches.
 

Attachments

Quote from bwolinsky:

Price Phsycis is the most well articulated theory of market movement. Every theory needs some lengthy explanation, and after reading all of the tools in the CFA curriculum there isn't a better theory for price action than that.
More...

"Price Physics" is bunk, plain and simple. I have a good enough background in physics, math/stats, finance/econ and trading system development to make that call. Maybe it's changed since 2005 or so when I bought the manual and observed the chat room for a bit, but I doubt it. People tried to trade with clueless discretion or else tried to make it mechanical (I was in the latter group, and tried many variations).

Like anything, you can curve-fit some nice-looking strategy out of it. However, until it's proven (read: audited) in real-time over statistically-significant number of trades and multiple market conditions, it's nothing more than theoretical mumbo jumbo. Sorry if that burst any bubbles.
 
Quote from jsp326:

"Price Physics" is bunk, plain and simple. I have a good enough background in physics, math/stats, finance/econ and trading system development to make that call. Maybe it's changed since 2005 or so when I bought the manual and observed the chat room for a bit, but I doubt it. People tried to trade with clueless discretion or else tried to make it mechanical (I was in the latter group, and tried many variations).

Like anything, you can curve-fit some nice-looking strategy out of it. However, until it's proven (read: audited) in real-time over statistically-significant number of trades and multiple market conditions, it's theoretical mumbo jumbo. Sorry if burst any bubbles.
More...

No, it's not bunk. That's just ignorance of what we're talking about. No you probably don't have a good enough background in any of those fields to "make that call."

There is a significant amount of work involved to make it reality. I have read the manual and I guess I used it far more effectively than you. It is 100% mechanical so I don't know what discretion has to do with anything.

Curve fits are not the same as optimizations. You and many like you confuse the two when this is at the heart of the scientific method, a basic tenet in education. There have been hundreds if not thousands of examples I've seen of the method, and if that's not enough I don't know how much you need, a hundred thousand? A million? No, you don't need that much.
 
Quote from bwolinsky:

Well, yes there is an objective method to do that, but it isn't mysticism, it's pure quantitative analysis. When you look at a chart, you have only the enclosed spreadsheet to use to define trend, and this is what Price Physics teaches.
More...

That's fine. Automate it and let it roll in real time. Set up a journal here or at Collective2. Let's see actual trades if the rules are that clear.
 
You must log in or register to reply here.
Back
Top