Higher Lows and Lower Highs

[Roberto]

I'm putting this here rather than under "forex" because although it's been described to me as a successful, simple system for trading Cable (GBP/USD) it strikes me that its discussion-potential is actually far wider than simply the world of forex. (The moderators may feel differently, in which case of course they may move it).

A colleague of mine has sent me the very simple system below and asked for my observations. It would be remiss of me not to offer it up for some discussion first, particularly as I haven't got adequate back-testing facilities to see whether there might be anything in it worthy of more detailed analysis.

It's been alleged that this system is consistently profitable with very reasonable drawdowns, which seems very surprising to me. In fact I can envisage it being more potentially worthwhile, with the numerical parameters modified to suit the volatility of the instrument concerned, _outside_ the world of forex, where volume can easily be monitored and some sort of "volume filter" added to it for trade-selection purposes.

Here's the whole thing:-

Using a 15-minute bar chart, enter a long trade at (or when approaching) the end of a bar if there have been 3 consecutive higher lows, and enter a short trade at (or just before) the end of a bar if there have been 3 consecutive lower highs. Initial stop-loss = 35 points and take-profit limit order = 115 points. If a bar closes with 35+ pips of profit, close half the position and move the stop to breakeven on the other half. If a bar closes with 50+ pips of profit, move the stop to 12 pips of profit. If a bar closes with 70+ pips of profit, close the trade completely. No stop-and-reverse if the stop-loss is hit at any stage, or if a "contrary signal" appears. Don't enter any new trades when a trade is already in play.

Comments, anyone?

Back-tests, anyone?

 

[Spreadbetteur]

Using a 15-minute bar chart, enter a long trade at (or when approaching) the end of a bar if there have been 3 consecutive higher lows, and enter a short trade at (or just before) the end of a bar if there have been 3 consecutive lower highs.

Roberto

Not sure I get it - If there have been three higher lows are you entering at the end of the 3rd or 4th candle ?

Also which end of the bar are you talking about ?

 

[Roberto]

You are entering officially at the start of the 4th bar (or candle) wherever the price happens to be at that moment. Alternatively, you can enter just before the end of the third bar, i.e. a minute or two earlier than at the start of the 4th one, if you're "sure" that the last minute's action of the third bar won't suddenly prevent it from being the third consecutive higher low or lower high, as the case may be. It makes no difference to the system whether the bars are rising or falling (although I would imagine that in practice if there are three higher lows they're likely to be rising, and vice-versa). You are always entering with a market order, in other words, not a limit order.

 

[Thirteen]

yes - sorry but its stupid. it assumes volatility never changes.
where do all the numbers come from? do u know why they are these numbers?
if u dont understand it u wont have confidence in it and so u will lose money with it.

 

[Roberto]

it assumes volatility never changes.

Does it? Would you still think that if it turned out to have roughly or exactly the same results at times of very different volatility? Or do you mean that your guess from looking at it without back-testing is that it would do better at some times than at others, depending on the volatility?

where do all the numbers come from? do u know why they are these numbers?

Sorry; I don't. It's only second-hand (at best) to me at the moment. I'll try to find out more and report back.

if u dont understand it u wont have confidence in it and so u will lose money with it.

I thought that one of the purposes of completely mechanical systems like this was to remove completely things like "psychology" and "confidence" from the equation?

 

[Thirteen]

1/ yes it does. i know because many years of trading tells me so. backtesting? whats the point? im not interested in what the market did last week/year. are u? why?
2/ you buy on smiths tip u gotta sell on smiths tip.
3/ no. when your mechanical system that u dont really understand losses 75% of your account r u telling me u still wont have any emotional involvement? will u know how to fix it given that u dont understand it? will u have the confience to even try? (but it worked on last months data right! wots wrong????)

a system built on the back of a fag packet and passed by chinese whispers isnt going to bring prosperity to any1 now is it.

was this post a joke or did u really think this was 'going to be the one'

 

[Thirteen]

by the way - i can tell u exactly how this sytem will pan out.....

some times it will work, somtimes it wont. over any long period (more than a year) however it will fail. if u do your own work u will learn why this is & u will learn more (better than a post expecting someone else 2 do the work 4 u eh)


some 'bright spark' will doubtlessly post data to the prove me wrong, but they will only post dats for a +ve year, not all the negative ones.

 

[LION63]

It is sad that some individuals rubbish ideas without even looking into them or understanding what is being said, in such situations it would probably be better to keep one's counsel. If one has decided that a particular strategy is not for him/her it would be better to ignore the thread and leave it for others that might find it worthwhile exploring the information that is being offered and shared.

Otherwise, we will end up with a forum that contains useless posts of limited or no benefit to anyone.

 

[Roberto]

backtesting? whats the point? im not interested in what the market did last week/year. are u? why?

I'm not, particularly. The point, since you ask, is that given that I've been asked to advise someone about a "potential system", I'll know more easily what to say if it's never had a good run at all than I will if it's quite often had a very good run. I don't think anyone should trade a mechanical system that's never had a good run at all. And you can't know that without backtesting.

will u have the confience to even try?

Not without it having been tested, no.

was this post a joke or did u really think this was 'going to be the one'

Neither. If you read my post you'll see I've already said I'd be surprised if it worked. But don't let a little matter like that get in the way of some rudeness and aggression, will you?

 

[Roberto]

some times it will work, somtimes it wont. over any long period (more than a year) however it will fail.

That was my feeling about it, yes. And now we know that it's yours, too. It's just a shame that we had to put up with your customary style of reply first in order to find that out. But thank you for your opinion anyway.

 

[FetteredChinos]

morning all. interesting idea Roberto. (ps do you post with an extra "V" over on moneytec??)

if i recall. i had a brief test of this on hourly bars a while back. (not sure if i can still find the spreadsheet. will have a look later)

from what i found, it was that 3 higher lows was usually bearish, and 3 lower highs usually bullish.

then again, i wasnt looking at profit targets or stops, as per your original post. just looking at momentum of the following bars.

as a result, and much like i hate to agree with 13, i dont think this system has too much longevity.

then again, i did only test on hourly bars over 2 years of data. things might have been different on 15 mins. but i wouldnt bet on it.

sorry to be a damp squib, but perhaps someone can prove otherwise?


FC

 

[BBB]

It is sad that some individuals rubbish ideas without even looking into them or understanding what is being said, in such situations it would probably be better to keep one's counsel. If one has decided that a particular strategy is not for him/her it would be better to ignore the thread and leave it for others that might find it worthwhile exploring the information that is being offered and shared.

Otherwise, we will end up with a forum that contains useless posts of limited or no benefit to anyone.

Actually, I must say that the method posted also seems doomed to failure. I can understand where 13 is coming from in his explanations on why this is so. Ok, so the manner isn't ideal I will agree.

Its not a case of 'that a particular strategy is not for him/her ' its a case of the system will not work for anyone. Period. As you rightly say Lion, 'It is sad that some individuals rubbish ideas without even looking into them or understanding what is being said' !

I'm sorry to say that this type of post is the very reason why I seldom bother with Trade2Win these days and gets me hopping mad (ok I appreciate that 13's attitude is sometimes OTT). Would you rather we all spend weeks and weeks finding out the obvious, going down the same old road that has been travelled so well on every other system discussion? or is the purpose of Trade2Win to allow you, the not so experienced members to learn more quickly from others experience?

May be I misunderstood Trade2Win. It used to be a great place where like minded people came to discuss real trading ideas. It is now little more than a talking shop for those 'hobby traders' still in search of the ideal way to enter the market. The great shame is when someone who does seem to know what they are talking about (stevet, grey1, et al) leave because of general ignorance of others, and the need to keep reinventing the same old wheel. As a result the active membership consists of a regular churn of newbies who eventually fade out, and the same old vendors.....

Rant over.

 

[TheBramble]

Why does everyone keep calling me al....?

 

[FetteredChinos]

sorry Ernest Theodore..

 

[FetteredChinos]

oh Al, what have you done in the Other Place?

 

[Roberto]

(ps do you post with an extra "V" over on moneytec??)

No, I am not "Robertvo". I occasionally read there (but don't like it much); don't post. Has Robertvo been discussing something similar there? If so, I'd be interested to read it, actually ... might account for a bit ...

if i recall. i had a brief test of this on hourly bars a while back ...

from what i found, it was that 3 higher lows was usually bearish, and 3 lower highs usually bullish.

Really? I must admit that would surprise me even more.

as a result, and much like i hate to agree with 13, i dont think this system has too much longevity.

Yes indeed ... it pained me too to agree with him, but I had in fact already suggested as much.

i did only test on hourly bars over 2 years of data. things might have been different on 15 mins. but i wouldnt bet on it.

Well, even if they were, the moral of the story might be simply how arbitrary and unreliable these things are.

sorry to be a damp squib, but perhaps someone can prove otherwise?

That was sort of what I was wondering, but with very little optimism, I must say.

Thanks for your replies, all.

 

[java]

Joe Ross has a method a bit like this in one of his books - one of his methods of determining trend change is to start counting lower highs/higher lows. When he gets to 3 successive lower ones he will go short on the break low of the bar that made the third lower low. I haven't really tried it in a determined way but the principle seems useful for helping to pick direction.