Four Week Rule - from the J.J.Murphy book

[Shadower]

Hi all,

first timer here, techie by nature, interested in tech anal for my job so be gentle with me.

I'm going through the John Murphy classic and, well, I'm stumped.

Chapter nine, moving averages. Lags, trendfollowing, etc, get that fine. Then page 215ish, the four week rule.

Briefly :-

Buy when the close is higher than any close in the previous 4 weeks.
Sell when the close is lower than any close in the previous 4 weeks.

A couple of monkeying about bits about getting out at 2 weeks and stuff. But.....

I've just put this again the FTSE100 for a year and pretty much it's buying high and selling low. Shurely shome mishtake ? At the first buy signal ( I'm seeing it at around the end of July ) your supposed to cover your shorts and go long. thats covering from selling at a little under 3800 and buying them back at close to 4400. Is this a mis-print or should I just go back to filling the printers ?

Damn confused.

Sorry for a basic on here, seem to be about the best tech anal site I can find so far.

Cheers

D

 

[oatman]

.......or writing books :cheesy: ?

 

[Helenqu]

I think it may just be that the market has changed significantly from when the book was written in 1986. Markets change all the time. Set ups that worked ten years ago don't work now. The fun bit is finding the ones that do work now

 

[Skimbleshanks]

Sounds to me like you're describing Donchian channels - as used by the Turtles. It's an 'always in the market' strategy and therefore it will give you an anomaly if you just pile in. You need to wait for the extreme, and then pile in on a breakout of the 20 day (four week) channel / envelope. You then reverse your position when the price breaks down through the bottom of the channel, although this is a long term strategy. A shorter term uses 20 day envelopes to enter and 10 day envelopes to exit (ie, not reversing).

Perhaps you're mistaking channels and envelopes with moving averages?

 

[Shadower]

Aye, see what you mean, the Turtles stuff does look like this, although I guess with some other funky stuff. The book also has it as a constantly 'on' trading system, unless you do the 20 day in 10 day out bit.


I've put a graph in of how I'm reading it, so I don't think I'm getting the two confused but.

According to the rules, buy when > 20 day high, sell when < 20 day low.

Hopefully the attached graph will show the FTSE 100 from June 02 to June 03, and I've labeled in 5 distinct direction changes.

The first, which trigger the initial position as the sells from 15th Aug 02 to 24th Aug 02, four distinct new lows, closing at 3974 down to 3626. From the rules, these would be keep until the buy signals come in, starting 9th Sept at 4190. So, just to cover out of my shorts I'd be giving up minimum over two hundred points.

The rest of the signals seem ( and I'm probably reading these wrongly ) that I'm buying at the absolute peak of the market and selling when it's rock bottom.

Again, it could just be that this isn't the market to be using this system in, or am I just not reading this right. If this kinda system is very market specific ( and by that I mean the state of the market rather than the market itself ) then should it really be in Murphy's book, which I've been led to believe was basically the A-Z of TechAnal.

Confused and Dumb. Mostly dumb.

Cheers for your time all.

D

PS - I just cant chart loading working - but it's a FTSE 100 OHLC chart, I'm sure you'va all seen one at some time ;-)

I'm getting sells mid July, buys through august, then sells through Sept, all the worst times to do so.

 

[wysinawyg]

The problem is that this kind of breakout system only works with big trends. The FTSE has been pretty choppy for the past year so it isn't really surprising its made losses over that time. The idea of this kind of system is to make huge gains in long runs (up or down) and accept that you'll take hits when the markets are going sideways.

I don't think it is a great idea to trade this kind of EoD system on just one market. You need a whole bunch of different markets (i.e. not just 10 different equity indexes, but commodities, metals, bonds etc.) trading at the same time on the basis that each system is profitable over a long (10-20year?) period and that trading them all together means you remain broadly profitable all the time.

Murphy is a great A-Z of TechAnal, but think of it more as an actual A-Z (i.e. covering everything even where you won't care about it) rather than a set of directions that show you how to get where you want to go.

wysi