Forex Factory (https://www.forexfactory.com/forum.php)
-   Trading Discussion (https://www.forexfactory.com/forumdisplay.php?f=11)
-   -   What works, and what doesn't (https://www.forexfactory.com/showthread.php?t=560393)

bloodpoodle Oct 4, 2015 9:08am | Post# 1

What works, and what doesn't
 
2 Attachment(s)
I have done backtests of multiple systems trying to find what works and what doesn't work. The one thing I found is market likes to be efficient and doesn't like quick moves which are highly inefficient. Market likes to range 75% of the time So these quick inefficient moves often get faded right back into range.
This is what my research has shown. If you have other research consisting of 1000 back tests in all market conditions please share! I'm looking for statistical proof of what works and what doesn't. Please post any research of 1000+ backtests in all market conditions.. Flat, uptrending on daily and downtrending on daily.

What dosn't work consistently:
1. Trendlines - Most rookies start with drawing trendlines.. They don't work. Too many false signals.
2. Multiple indicators... I haven't found one system with consistent results yet. That works in all types of market Bull, bear, Flat.
3. over complicated systems.
4. trading good news or bad news.. If it were as easy as going long on good data... we would all be rich..
5. Chart patterns found in most books. They work some of the time.. But fail a lot.. When they do fail.. usually a good clue.
(When you challange me on this provide Data that show's you don't take multiple Losses to get one win on a highly consitant basis.)

What does work...
1. Market structure...

a. A FAST MOVE UP IS WEAK!... Never chase... Look to fade... if there is opportunity. 75% of the time these quick moves are short lived.
Click to Enlarge

Name: Screenshot4.png
Size: 268 KB

b. A SLOW GRIND IS STRONG.. You don't want to be the first to go short against a slow grind up. Reason: lots of bears stuck in the grind. that now have to buy to exit
Name:  Screenshot1.png
Views: 29695
Size:  108 KB

This is not a holy grail... Usually if price moves fast up.. it can retrace just as fast. If it grinds up slowly its usually trapping a lot of people on the wrong side as it keeps poking up higher and higher.... And price won't want to come back and let the trapped people out who are on the wrong side of the market...
Somtimes you might get 2 or 3 legs up after a fast move before price decides to retrace.... A lot of its based on reading the liquidity of the market. and the best way to read liquidity.. Is to Know the time of day and when major data is out...

2. Liquidty
Reading liquidty...
Time of day.. Liquidty may be super high at london open then die off into london lunch.. Then surge again at USA open.
Time of week.. Liquidty may die down into a major event such as a few days before Non-farm payroll might be slow.
Time of year.. Summer and Christmas are usually lower liquidty as traders are on vacation.
This takes a lot of chart time to "get the feel" for when liquidty will change..
When liquidty does change Bulls may stop pushing... The bids start drying up and price may reverse..
to get the feel start marking on your charts and taking note of liquidty changes at different times of day and at data releases.

3. A major change in monetary policy will reverse a currency.. (might take a while to kick in)
If Aud is in major up trend and the RBA says it will start a rate cut cycle... Aud will fall.
If USA stops printing money as starts to hike interest rates Dollar will rise on the weekly charts.
I tend Not to over-complicate this by analyzing every bit of data... Its a Broad view... That doesn't change on one or two data points.
Most traders over Complicate this. For me has to be a major change Announced by a central banker. ONE BAD NFP or one GOOD EUROZONE DATA POINT WONT CHANGE MY LONG TERM VIEW. Yellen saying Rate hike is off the table would change my long term view of the USD.

FerruFx Oct 4, 2015 9:20am | Post# 2

What dosn't work: 1. Trendlines - Most rookies start with drawing trendlines.. They don't work. Too many false signals. 2. Multiple indicators... I haven't found one system with consistent results yet. That works in all types of market Bull, bear, Flat. 3. over complicated systems. 4. trading good news or bad news.. If it were as easy as going long on good data... we would all be rich.. 5. Chart patterns found in most books. They work some of the time.. But fail a lot.. When they do fail.. usually a good clue.
If you believe these 5 points don't work, you are obviously at the beginning of your trading learning process.

What doesn't work for you is not necessary true for everyone ...

bloodpoodle Oct 4, 2015 9:28am | Post# 3

1 Attachment(s)
A trade I took on Thursday... Based on a weak run up in N/U
Click to Enlarge

Name: Image1.png
Size: 354 KB
Notice the Market structure... Thin run up gets faded...
Can you see other quick runs on this chart? How did they hold up over time?

Atokys Oct 4, 2015 9:29am | Post# 4

What dosn't work: 1. Trendlines - Most rookies start with drawing trendlines.. They don't work. Too many false signals. 2. Multiple indicators... I haven't found one system with consistent results yet. That works in all types of market Bull, bear, Flat. 3. over complicated systems. 4. trading good news or bad news.. If it were as easy as going long on good data... we would all be rich.. 5. Chart patterns found in most books. They work some of the time.. But fail a lot.. When they do fail.. usually a good clue. What does work... 1. Market structure......
What do you mean when you say that something 'works'? Everything that 'works' or 'doesn't work' is entirely subjective and based on your beliefs about markets.

bloodpoodle Oct 4, 2015 9:53am | Post# 5

1 Attachment(s)
Click to Enlarge

Name: Screenshot1.png
Size: 74 KB
Entered this trade after 2nd leg retrace... it was up near highs on daily chart.. We were going into End of quarter and Non-farm payrolls were coming up in a few days.. I saw that Liquidity was drying up into the big NFP event in 2 or 3 days.. So I figured 1st leg wouldn't retrace based on dying liquidity. Entered when second leg retraced. I exited when I saw the Third Leg run really fast and thin.. I figured it would retrace so I exited. And it did retrace as I expected.
QUICK PIP RULE. USUSALLY IF YOU GET A QUICK SPIKE START LOCKING IN THE PROFITS.

bloodpoodle Oct 4, 2015 10:52am | Post# 6

1 Attachment(s)
Click to Enlarge

Name: Screenshot1.png
Size: 97 KB

This chart shows the liquidty surge during London... Bulls keep pressing for a while as London can be strong. Going into Usa you see the bid drying up.. and price falling back into Asian range... When Liquidty picks up again and USA open its the bears turn to push.. But even that doesn't last long.. The move down was too quick and inefficient and had to retrace..

bloodpoodle Oct 4, 2015 11:15am | Post# 7

1 Attachment(s)
Click to Enlarge

Name: Screenshot1.png
Size: 67 KB

Market structure is important... If you don't use MARKET STURCTURE, it needs to be in your vocabulary.
chart showing that quick inefficient moves don't hold and are good opportunities!
The gray boxes are London open to close and not that important to this chart. You can ignore them.

Dovilo Oct 4, 2015 1:16pm | Post# 8

What dosn't work: 1. Trendlines - Most rookies start with drawing trendlines.. They don't work. Too many false signals. 2. Multiple indicators... I haven't found one system with consistent results yet. That works in all types of market Bull, bear, Flat. 3. over complicated systems. 4. trading good news or bad news.. If it were as easy as going long on good data... we would all be rich.. 5. Chart patterns found in most books. They work some of the time.. But fail a lot.. When they do fail.. usually a good clue. What does work... 1. Market structure... a. A FAST MOVE UP IS WEAK!... Never chase... Look to fade... if there is opportunity. 75% of the time these quick moves are short lived. {image}

Long term trendlines are like one of few things that actually works...

Xela Oct 4, 2015 2:03pm | Post# 9

I have done backtests of multiple systems trying to find what works and what doesn't work.

Please excuse my making what might come across as a somewhat pedantic point, but strictly speaking it isn't possible to tell from backtesting what works and what doesn't: only what worked and what didn't. The tense is rather significant.


I'm looking for statistical proof of what works and what doesn't.

That can, by definition, come only from forward testing.


Most rookies start with drawing trendlines.. They don't work.

If, by "work", you mean "consistently produce profit on their own without taking a range of other trading-skills into account", then no: they don't (nor, probably, does any other individual technique or parameter you could mention). Personally, I still find some trendlines enormously valuable in my trading, in spite of that.


Multiple indicators... I haven't found one system with consistent results yet.

Well, I'm not an indicator trader, myself, but clearly many people have. There are loads of beginners' textbooks (such as Van Tharp's Trade Your Way to Financial Freedom and Tushar Chande's Beyond Technical Analysis) which list and illustrate and discuss some multiple indicator systems, together in some cases with decades of backtesting results and a year or two's forward testing results, all analyzed and explained in great detail. They're still "not my choice", though.


Chart patterns found in most books. They work some of the time.. But fail a lot..

Not my experience, I must say. I've learned many from books that work a significantly high proportion of the time. I'm thinking of the books of Joe Ross, Bob Volman, Al Brooks and Lance Beggs, in particular, but there are others, too. I wouldn't presume to generalise about "most books", but the price action/chart pattern textbooks I've read have been enormously valuable to me, so much so that I don't pretend that I'd be making a living at all without having studied them so assiduously.

bloodpoodle Oct 4, 2015 2:13pm | Post# 10

1 Attachment(s)
Whutevuh dude... Go trade your trendlines
Price can move in one direction for a long time... But price will move at different speeds based on liquidty.. thus never respecting the trendline enough to be consistant.

Click to Enlarge

Name: Image10.png
Size: 60 KB

bloodpoodle Oct 4, 2015 2:26pm | Post# 11

At lease nobody saying I'm wrong about what does work...

vasily Oct 4, 2015 2:35pm | Post# 12

when the case is forex, everything works, until is doesn't...!

vasily Oct 4, 2015 2:42pm | Post# 13

"slow grind up" is where it should be shorted. but there is always possibility of grind zone will be broken up and it will hunt you as shorted
when you wait for it to break then short where will give you better risk reward ratio then it might be the start of impulsive reverse move

you can never know what will follow the spike. some short it some long it. most works then one day it hunt you bad

I have done backtests of multiple systems trying to find what works and what doesn't work. The one thing I found is market likes to be efficient and doesn't like quick moves which are highly inefficient. Market likes to range 75% of the time So these quick inefficient moves often get faded right back into range. This is what my research has shown. If you have other research consisting of 1000 back tests in all market conditions please share! I'm looking for statistical proof of what works and what doesn't. Please post any research of 1000+ backtests...

Mariodrugs Oct 4, 2015 5:35pm | Post# 14

{quote} If you believe these 5 points don't work, you are obviously at the beginning of your trading learning process. What doesn't work for you is not necessary true for everyone ...
I have the same feelings like bloodpoodle, does it work for you FerruFX ?

For me only the best is good flag pattern, supply ressistance, SR zones works 50 / 50.

ruma29 Oct 4, 2015 5:49pm | Post# 15

I have done backtests of multiple systems trying to find what works and what doesn't work. ....
There is no AI yet for trading. Humans with all their emotions are heavily involved in trading...and finally without understanding trading itself there is no way for consistent success in trading.....I might be wrong on this subject, but that is my opinion.

Mariodrugs Oct 4, 2015 6:02pm | Post# 16

bloodpoole so when do you enter ? where is your sl ? and Tp ?

cakrajaya Oct 4, 2015 6:38pm | Post# 17

I have done backtests of multiple systems trying to find what works and what doesn't work. The one thing I found is market likes to be efficient and doesn't like quick moves which are highly inefficient. Market likes to range 75% of the time So these quick inefficient moves often get faded right back into range. This is what my research has shown. If you have other research consisting of 1000 back tests in all market conditions please share! I'm looking for statistical proof of what works and what doesn't. Please post any research of 1000+ backtests...
market structure absolutely important thing, but statement "do not work" on several listed mention above need careful. first meaning work and do not work. how the work? maybe should understanding about the meaning work. hammer absolutely not work to saw out a wood, to saw we need a saw

hanover Oct 4, 2015 6:40pm | Post# 18

.........but strictly speaking it isn't possible to tell from backtesting what works and what doesn't: only what worked and what didn't. The tense is rather significant......... That can, by definition, come only from forward testing.
I've read this type of remark many time before, but -- unless it's referring to the fact that backtests don't properly take (what is now unknown) costs and slippage into account; or any dangers involved in curve fitting a system to maximize its performance over past data -- I must respectfully disagree, because (everything else being equal) any forward test that was started X days ago (and up to today) will yield the same entries and exits, and therefore P/L, as a backtest that's conducted today across the same period.

To whatever extent backtesting is impotent because markets change their color over time, and hence past data doesn't reflect what will happen henceforth, forward testing is equally so, because your forward testing has likewise been conducted over what is now historical data.

fxnikolaev Oct 4, 2015 7:18pm | Post# 19

Newbie talk!!!! And nothing more ! This thread has to be in rookie talk! Plz some one report it. Thx!!

hanover Oct 4, 2015 7:29pm | Post# 20

3. over complicated systems.
...... and also systems that are simple enough to be coded into an EA.

Between 2009-2012 I coded maybe 20-30 EAs by handpicking the seemingly best ideas from the dozens of requests I received from (in some cases, the most highly followed) FF members, who all insisted that their strategies were profitable. When this proved not to be the case, I modified these EAs, adding their suggestions for filters to eliminate entry signals that resulted in losses, different SLs, TPs, trailing stops, MMs ..... you name it. As far as I know, NONE of these EAs are still being run today, a complete waste of time for everybody involved.

Hence I think it's safe to say to anybody that "if your system is simple enough to be coded as an EA, then it almost certainly won't be long-term profitable." No matter how you slice it, an experienced human trader will always have more intelligence than even the most sophisticated EA.

And, moreover, it stands to reason that if the markets could be conquered by a single, objectively definable system, that everybody would use such a system. Then, given that there must ultimately be a willing seller for every buyer, markets would either cease to exist (LOL), or would necessarily adapt in such a way as to make the system unprofitable. That's one reason why IMHO profitable retail trading involves nuanced knowledge -- and preferably knowledge that's somehow grounded in how heavyweight forces are currently driving price. A simple TA-based system can remain profitable only in so far as it exploits heavyweight-generated inefficiencies whose uniformity and repeatability can be guaranteed to last throughout the lifetime of the system.


© Forex Factory