On the topic of BS in trading forums... | Page 5 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Forums for Trading Futures 1 reply

Who knows any good automatic trading forums? 7 replies

Attachments: On the topic of BS in trading forums...
Exit Attachments

On the topic of BS in trading forums...

  • #81
  • Jun 3, 2010 3:43am Jun 3, 2010 3:43am
Quoting mikkom
There are programmers and programmers. Check what an average salary is for an usual fund tech guy compared to industry standard.
I dunno, low-mid 6 figures?

But that's the commercial sector isn't it, I can't see Joe Trader paying much more than what I paid, although I could be wrong!
  • #82
  • Edited 4:13pm Jun 3, 2010 4:02pm | Edited 4:13pm
@pipmutt
“Don't keep us in suspense, who is it!”
I expect that it would be a violation to post names or URLs in this part of the forum. So for the moment, I’ll just say that he’s an occasional attendee at out local TA group, and not a regular FF contributor. He’s been trading forex profitably for around 3 years, and has been approached by a sufficient number of others who’d like him to manage their money, that he’s now decided to formalize, and grow, this arrangement. Whatever his reasons, he believes that I can add value to his business, and has offered me a stake in it. I see it as a worthwhile potential opportunity; nothing ventured, nothing gained.

“As for programming, it's an extremely competitive marketplace. ....... so far I haven't seen anything for sale that isn't either snakeoil and/or available free (or extremely cheaply) elsewhere.”
I agree with you. For example, there’s a currency meter called ‘Accustrength’ (Google it) that you can buy for around $800, but why would anybodyspend that kind of money when you can download one (like the one that I wrote - shameless plug, LOL) for free? Anyway, I do have a couple of handy MT4-based tools that I plan to offer inexpensively, but I see this initiative as little more than a sideline; I have nothing to lose financially, by doing so.

However, my primary reason is simply to declare my intentions (see post #72). I only mentioned the ‘Commercial Member’ thing because I wanted to post a summary of my forex learning to date, as a kind of ‘farewell’ to this part of the forum. It’s really not a big deal to me, but it’s certainly provoked some discussion.

I share your skepticism toward vendors in general, a topic that I’ve posted on many times. I've certainly spent my share of money on worthless junk; hopefully I've learned my lesson.

One more word on programmers: IMO there have been many generous folk who've written indys and EAs here for free, whose time and effort has not been recognized, and that as a result, programming expertise is now very underrated around FF. No offense to your friend, but if he thinks that several hours worth of work is worth $20, then he's helping to perpetuate that impression. Where I live (New Zealand) software design and development is regarded as a very respected profession. During the 1990s I worked on contracts here that paid anything up to NZD 80 per hour, and there were project leaders on much higher rates than this.


@Rabid
“Have you considered a fixed exit strategy, rather than a variable strategy based on ranging/trending? What criteria do you use to determine your exit?”
My exits are a work in progress, but here’s my current approach.

I realize that scaling out doesn’t provide an edge, but (like Kiwi_trader said) it’s a way of smoothing the overall result. I also realize that advancing SL to BE doesn’t necessarily provide a statistical edge, either; I do it nonetheless, for all of the 'obvious' reasons.

My trade generally consists of 5 single unit positions. That gives me 5 potential exit points. If the longer term (generally daily) trend appears strong, I’ll let 2-3 units run longer. When I exit the first position, I move SLs on all others to BE.

I use 3 technical tools: (1) climax volume up/down (green and red in the attached screenshot) candles; (2) areas of obvious prior S/R (e.g. double tops and bottoms); (3) an ATR based trailing SL (purple line) on each of the H1 and D1 charts. No special reasons for these indys, or the way in which they’re calibrated; as a discretionary trader I’m merely trying to find a balance between taking some profit at higher probability points of reversal, and letting some of it run in the hope of catching a high RR trade. In regard to the latter, the goal is to find low risk entries (a principle I learned from DIBS, but I don’t use IBs, though) on a low TF chart (usually M15/M30), and then exit part of the position on a D1 TF scale. In this way it’s sometimes possible to get a 200-300 pip return for 10-15 pips risk. (Back in March, when EUR was weakest and AUD strongest, I managed to exit 2/5 of a position on EURAUD at 47:1 RR, by using this approach – probably a once-in-a-lifetime event, and too bad it was only microlots, LOL).

The exits are designed to complement the fact that I trade the strongest vs weakest currencies, seeking out the (potentially) strongest and ‘cleanest’ trends.

I’m sure much of this flies in the face of a rigorous mathematical approach, and I’ll have all of the proponents of KISS likewise haranguing me, but that’s my current exit plan. LOL.

[Note: sorry if it’s hard to distinguish between the red and brown candles in the screenshots – not sure what happened during the screen capture process].
Attached Image(s) (click to enlarge)
Click to Enlarge Name: Rabid exits1.jpg Size: 143 KB
Click to Enlarge Name: Rabid exits2.jpg Size: 135 KB
Click to Enlarge Name: Rabid exits3.jpg Size: 162 KB
  • #83
  • Jun 3, 2010 4:22pm Jun 3, 2010 4:22pm
Quoting hanover
I realize that scaling out doesn’t provide an edge, but (like Kiwi_trader said) it’s a way of smoothing the overall result. I also realize that advancing SL to BE doesn’t necessarily provide a statistical edge, either; I do it nonetheless, for all of the obvious reasons.
Nod. I've been wondering how best to handle scaling out with an edge. It occurred to me that in a faster moving market, it makes more sense to scale out on higher pip amounts. Kindof like scaling out on an efficiency ratio thing.

After working on it a bit, it occurred to me that maybe it's best to scale out over set units of time, rather than pips. That way in a fast moving market you capture more pips, in a slower moving market you have a natural exit.

I advance my SL to BE, too. The question I've ran into, time and time again, is how long/many pips to wait.

Quoting hanover
I use 3 technical tools: (1) climax volume up/down (green and red in the attached screenshot) candles; (2) areas of obvious prior S/R (e.g. double tops and bottoms); (3) an ATR based trailing SL (purple line) on each of the H1 and D1 charts.
Nod, S/R provides stacked order areas. ATR is interesting, but it requires optimization and lags the market. I've had problems in that area. It working for you?

Quoting hanover
In regard to the latter, the goal is to find low risk entries (a principle I learned from DIBS, but I don’t use IBs, though) on a low TF chart (usually M15/M30), and then exit part of the position on a D1 TF scale. In this way it’s sometimes possible to get a 200-300 pip return for 10-15 pips risk. (Back in March, when EUR was weakest and AUD strongest, I managed to exit 2/5 of a position on EURAUD at 47:1 RR, by using this approach – probably a once-in-a-lifetime event, and too bad it was only microlots, LOL).
Hehe. Yeh, had a few of those. In earnest, it's 47:1 RR in terms of pip sum, but not a 47:1 RR in terms of actual dollars because of the scaling. I've never liked pip sum measurements, personally.

I've been playing with your currency strength indicators. It looks like it triangles over several pairs to determine strength. Are you familiar with efficiency ratio?
  • #84
  • Jun 3, 2010 4:54pm Jun 3, 2010 4:54pm
First off, Hanover, great post and a whole bunch of wisdom in that... and anyone who thinks it long winded, clearly hasn't understood it properly. Good luck with the future endeavours too.

Re exits, I've been sweating over these for some time now, settling with ATR-based exits after extensive testing and trading. Some interesting things came out of this:
- the markets trend more than they range on a 24-hour basis. On an hourly basis they range more than they trade. Interesting huh? The definition of trending I'm using is to have greater than 50% body vs total of wicks.
- this means there is a natural bias on the longer timeframes to trending and RR should reflect that
- you can predict the daily range to within 30% on average (with a basic ATR) so it makes a natural, statistically valid aim
- I can't improve ATR. Not with EWMAs or anything. If anyone else can, I'm all ears. I'm wondering about overnight option vols...
- If you've got a good entry on a reversal that price has moved away from quickly, if price reaches it again, it'll still be a good entry point, ergo don't move to break even. Obviously this depends on your trading style. I happily move to b/e on a breakout
  • #85
  • Jun 3, 2010 6:17pm Jun 3, 2010 6:17pm
Quoting Rabid
Nod. I've been wondering how best to handle scaling out with an edge......
Time based exits: I haven't done a lot of work in this area. My (over-simplified) view is that markets are the result of collective sentiment, which creates orderflow, which creates price, then price results in a change of sentiment, and this chicken-and-egg process then repeats itself. Given that most traders think about price more than time (price equates more directly to profit, the ultimate goal), then my gut tells me that price would advantageously be the #1 consideration. However, I'd be very interested to see research that illustrates how time can be used effectively, especially as my understanding of Gann, time cycles, etc is not great.

I think there are two main approaches to exits: either we (1) set optimized targets and/or trailing SL distances, based on past stats (daily ranges, average move lengths, etc), or (2) use a more dynamic approach, i.e. following price and adjusting exit points as key levels are created or destroyed. In terms of providing an edge, the latter ultimately stands or falls on how 'predictable' the price action is, i.e. its behavior in terms of trend continuation, or stalling/reversing at certain levels. If we had access to more 'orderflow' type data, e.g. market depth, separate bid/ask volume across the entire market, then IMO a dynamic approach would become even more advantageous.

I tend to favor the dynamic approach nonetheless, however, I'm sure that the 'optimized targets' can be made to work - but again, I've devoted less time into it, so I can't really offer intelligent comment. Philmcgrew has performed some extensive research into optimized exits - he's posted graphs other than this one, but I can't recall where. Another reason I prefer the dynamic approach is that it's universally applicable, whereas optimization presumably requires using a different value for each pair. It also raises questions as to which parameter(s) should be fixed, and which variable; largest possible sample size vs obsolescence; how often one should re-optimize; and a whole lot of other statistical concepts that go way beyond my high school math education.

Re optimizing ATR: because of my strong beliefs in orderflow and S/R levels**, I suspect the best way to use ATR would be on some kind of dynamic S/R basis, i.e. simply advance the SL to each new swing point.

[**BTW, when I refer to S/R I include prior highs/lows, trendlines, pivot levels, fibo retracements, round numbers, etc - any levels with potentially higher than normal visibility - explained in detail here]

I'm simply using an ATR based trailing SL because I'm trying to find strongest trend setups, and hence let part of the position ride, thereby allowing the exit method to complement the strategy as a whole. The way I see it, giving pips back due to a lagging indy is an occupational hazard that I must embrace, if I want the opportunity to pick up some high RR trades. I know that's stating the obvious, but ultimately I can't see any way around it. I'm sure that my ATR SL EA could be optimized much more effectively, but I'm happy to trust it to manage the last (and most time consuming) part of my exit while I head off to bed!

Here's an idea that I'm currently experimenting with. RR can be magnified greatly by the initial SL point; for example, if I can find technical justification (i.e. high number of different S/R types between entry and SL point) to use a 5 pip SL instead of 10 pips (disregarding spread), then I'm doubling the RR on all scale-outs, regardless of their outcome. I recently read an interesting e-book on 'trapped traders', a concept that might help to explain why there are many occasions where S/R holds exactly to the pip (especially on the majors). Anyway, whatever the underlying reason, (obviously) the bottom line is that unless a 5 pip SL gets triggered at least twice as often as a 10 pip one, the former is preferable. And there's no reason why we can't hedge our bets (smoothing, yet again) by having multiple SL points.

Re the formula behind the strength indy: it's merely a weighted average of the prices in all of the pairs that you specify, for a given currency, in the parameters. Explained in detail here.

Efficiency ratio: it's possibly a concept that I understand, but I'm not familiar with the terminology. Please feel welcome to explain, if you have the time; many thanks.
  • #86
  • Jun 3, 2010 6:46pm Jun 3, 2010 6:46pm
Quoting triphop
First off, Hanover, great post and a whole bunch of wisdom in that... and anyone who thinks it long winded, clearly hasn't understood it properly.
Even the longest forum post (wherever it might be) can't possibly be as 'long winded' as, for example, a 600 page book on market microstructure. LOL.

"the markets trend more than they range on a 24-hour basis. On an hourly basis they range more than they trade. Interesting huh? The definition of trending I'm using is to have greater than 50% body vs total of wicks...... this means there is a natural bias on the longer timeframes to trending and RR should reflect that"
Yes, this is very interesting. James16 suggests that newbies start by trading the daily charts upwards, supposedly because candle patterns (pinbars etc), S/R etc, are more reliable there, and trends are 'cleaner'. I've noticed that also, but there are plenty of intelligent folk here who believe that all charts are created equal.

Two possible reasons that arguably support Jim's view: (1) the longer the time period, the greater the opportunity for a significant cross-section of the trading population to participate, giving a more comprehensive market view, i.e. a kind of 'collective averaging'; (2) broad economic trends (perceived strengths or weaknesses in a nation's economy; inflation; the global economy in a state of risk appetite or risk aversion; carry trade; etc) are going to stay in place, and play themselves out, over a long period of time, which will ultimately be reflected in longer term technicals.

Of course (and before the scalpers jump on me, LOL) I'm NOT trying to imply that intraday trading is futile.

"If you've got a good entry on a reversal that price has moved away from quickly, if price reaches it again, it'll still be a good entry point, ergo don't move to break even. Obviously this depends on your trading style. I happily move to b/e on a breakout"
Yes, yes and yes! I've had so many trades stopped out because I moved to BE too quickly, only to find price revisit the same S/R several candles later, forming a double top/bottom. (Hopefully I've learned my lesson now). One of the many reasons why I have such a healthy respect for the validity of S/R.
  • #87
  • Jun 3, 2010 9:03pm Jun 3, 2010 9:03pm
Quoting hanover
Time based exits: I haven't done a lot of work in this area. My (over-simplified) view is that markets are the result of collective sentiment, which creates orderflow, which creates price, then price results in a change of sentiment, and this chicken-and-egg process then repeats itself. Given that most traders think about price more than time (price equates more directly to profit, the ultimate goal), then my gut tells me that price would advantageously be the #1 consideration. However, I'd be very interested to see research that illustrates how...
Microstructure is fine, but as a retail trader we're outside of that perspective.
Traders definately think about time, just look at how the markets behave around
index openings.

I'm not suggesting time cycles or gann or anything voodoo like that. I'm simply
saying that when price moves, it moves at a rate. Pips/minute or pips/hour or
whatever unit you choose to use. This means that when the market moves fast, each
unit of time is worth more. When the markets move slowly, each unit of time is
worth less.

I haven't done a lot of research in this area, but I do have time-based exits in
my rule list. I have found that slow periods tend to be more prone to whipsaw,
so it makes sense not to ride thru them. Trades that don't reach my profit targets
end up getting closed if they go sideways too long. What would happen if you got
into a trade based on strength/weakness, and then had that change mid-trade?

Quoting hanover
I think there are two main approaches to exits: either we (1) set optimized targets and/or trailing SL distances, based on past stats (daily ranges, average move lengths, etc), or (2) use a more dynamic approach, i.e. following price and adjusting exit points as key levels are created or destroyed. In terms of providing an edge, the latter ultimately stands or falls on how 'predictable' the price action is, i.e. its behavior in terms of trend continuation, or stalling/reversing at certain levels. If we had access to more 'orderflow' type data, e.g....
So you've essentially got an exit signal approach? Whenever the market gives the
red flag, you exit? Doesn't that mean that the success of your exits depends on the
accuracy of that signal? I would be concerned that a good entry could be destroyed
by a bad exit.

Quoting hanover
I tend to favor the dynamic approach nonetheless, however, I'm sure that the 'optimized targets' can be made to work - but again, I've devoted less time into it, so I can't really offer intelligent comment.
I tend to use self-optimizing targets. Atleast with mechanical systems. Price hits
X, I calculate the odds of price going to different spots. I calculate the size of
the average failed trade, and use that to place a stop. If the RR between the
projected target and stop is too low, then no trade is taken.

Quoting hanover
Philmcgrew has performed some extensive research into optimized exits - he's posted graphs other than this one, but I can't recall where. Another reason I prefer the dynamic approach is that it's universally applicable, whereas optimization presumably requires using a different value for each pair. It also raises questions as to which parameter(s) should be fixed, and which variable; largest possible sample size vs obsolescence; how often one should re-optimize; and a whole lot of other statistical concepts that go way beyond my high school math...
I look at it differently. I think if you have to optimize a parameter, you're already
in a losing position. From my experience, the markets often behave chaotically.
That means any parameter you use is subject to changing randomly over time. I
prefer parameters that automatically adjust from market data.

Quoting hanover
Re optimizing ATR: because of my strong beliefs in orderflow and S/R levels**, I suspect the best way to use ATR would be on some kind of dynamic S/R basis, i.e. simply advance the SL to each new swing point.
My concern with orderflow is that, from our perspective, it's essentially random. Sure
there is a pattern to it, but we can't see that as retail traders. We can't see when
someone is planning to put a large trade in. All we can see is price has reversed at a
previous level, and so orders may be stacked there. What kind of initiative trading
results from those orders... we can't know in advance.

Quoting hanover
The way I see it, giving pips back due to a lagging indy is an occupational hazard that I must embrace, if I want the opportunity to pick up some high RR trades. I know that's stating the obvious, but ultimately I can't see any way around it.
Nod, takes us back to symmetry. A few high RR trades (long tail) or a several lesser
RR trades? There's no right answer there. Obviously the "hit it outta the ballpark!"
trades are way more sexy and attractive, tho.

Quoting hanover
Here's an idea that I'm currently experimenting with. RR can be magnified greatly by the initial SL point; for example, if I can find technical justification (i.e. high number of different S/R types between entry and SL point) to use a 5 pip SL instead of 10 pips (disregarding spread), then I'm doubling the RR on all scale-outs, regardless of their outcome.
I hadn't really thought about it in terms of stops, but that makes sense. If cutting
your SL in half doesn't double your lost trades, then there's some asymmetry there.
Only question then is: If you cut your SL in half and increase your losses by 30%,
what's your net win/loss? Seems you could cut your SL in half as long as two constraints
are true: Your win/loss is high enough, and you don't double your losses between
the tests.

Quoting hanover
I recently read an interesting e-book on 'trapped traders', a concept that might help to explain why there are many occasions where S/R holds exactly to the pip (especially on the majors). Anyway, whatever the underlying reason, (obviously) the bottom line is that unless a 5 pip SL gets triggered at least twice as often as a 10 pip one, the former is preferable. And there's no reason why we can't hedge our bets (smoothing, yet again) by having multiple SL points.
I gave up, a long time ago, trying to understand "why" things happen in the markets.
At the time something's happening, the "why" doesn't seem to matter.

And if scaling out works for you, that's really all that matters.

Quoting hanover
Re the formula behind the strength indy: it's merely a weighted average of the prices in all of the pairs that you specify, for a given currency, in the parameters. Explained in detail here.
Nod, read that. Basically you're measuring the slope of price changes in terms of
percent change over multiple pairs and averaging the results.

Quoting hanover
Efficiency ratio: it's possibly a concept that I understand, but I'm not familiar with the terminology. Please feel welcome to explain, if you have the time; many thanks.
Basically it's a measure of price movement to end price change. Take a period, take
the high-low, then take the AbsoluteValue(open-close). Then divide the open-close by
the high-low. The closer the value is to 1, the more efficient (stronger sentiment)
that market is.

Now I don't like the requirement for a fixed period. Instead, I tend to use an
average pips measurement to determine the stop and start, but you get the basic
idea.
  • #88
  • Jun 3, 2010 9:06pm Jun 3, 2010 9:06pm
Two other things support James's view that people should start with longer timeframes:

- measurements of noise vs signal favour long timeframes supporting anecdotal evidence that the patterns are clearer
- longer timeframes give traders more time to make decisions and recover from the emotional impacts of poor ones or from simple losses

I recall somewhere that it takes around 30m before your brain moves from significantly impacted by a felt loss back to fairly rational - think about the last time you were furious with a teenager. So 1m timeframes and 5m trades put you squarely within the flood of adrenaline.
  • #89
  • Jun 3, 2010 11:09pm Jun 3, 2010 11:09pm
Quoting hanover;3775823<br /> <br /> [I
"If you've got a good entry on a reversal that price has moved away from quickly, if price reaches it again, it'll still be a good entry point, ergo don't move to break even. Obviously this depends on your trading style. I happily move to b/e on a breakout"
[/i]Yes, yes and yes! I've had so many trades stopped out because I moved to BE too quickly, only to find price revisit the same S/R several candles later, forming a double top/bottom. (Hopefully I've learned my lesson now). One of the many reasons why I have such a healthy respect for the validity of S/R.
Can you and/or Rabid comment on this trade management technique of getting a free trade without moving SL. I'm quoting myself from another thread. Thanks in advance.

========================================================
You can probably reduce the number of losing trades with this method. First divide up your position size in thirds; for example only trade with 3, 6, 9 mini-lots. When the trade moves in your favor by 0.5R take profit on 2/3rds of your position, keeping your original S/L, and you now have a free trade to trail. For example, suppose you enter a trade with a 20 pip stop-loss using 3 mini-lots. You book profits on 2 mini-lots when your position is +10 pips in your favor ( which is 0.5R ). This gives you a profit so far of $20 ( $2 a pip at +10 pips ). Now if the trade goes sour and hits your S/L it will be a breakeven trade ( your remaining 1 mini-lot will have lost 20 pips at $1 a pip in that scenario ). There are pros and cons to this MM strategy. Take it or leave it.

Instead of taking 2/3rds of your position off at 0.5R, you can take off ½ your position instead at 0.5R, but then you have to move your S/L up halfway between your initial S/L and your entry price. For example, suppose you enter a trade with a 20 pip S/L using 2 mini-lots. You book profits on 1 mini at +10 pips in your favor and then you move you S/L up or down 10 pips towards your entry price, in order for you to secure a free trade. Check my math on these. This is sort of confusing for me too.
=======================================================
  • #90
  • Jun 3, 2010 11:34pm Jun 3, 2010 11:34pm
Quoting triphop
- I can't improve ATR. Not with EWMAs or anything. If anyone else can, I'm all ears. I'm wondering about overnight option vols...
There is a lot of discussion about this in "volatility trading" scene. Read about black scholes, implied volatility and later theories etc.
  • #91
  • Jun 3, 2010 11:42pm Jun 3, 2010 11:42pm
Quoting mikkom
There is a lot of discussion about this in "volatility trading" scene. Read about black scholes, implied volatility and later theories etc.
Nod. Black-Scholes has a few problems tho.

Modeling this kind of stuff is not an easy task.
  • #92
  • Jun 4, 2010 12:49am Jun 4, 2010 12:49am
Quoting hanover
One more word on programmers: IMO there have been many generous folk who've written indys and EAs here for free, whose time and effort has not been recognized, and that as a result, programming expertise is now very underrated around FF. No offense to your friend, but if he thinks that several hours worth of work is worth $20, then he's helping to perpetuate that impression. Where I live (New Zealand) software design and development is regarded as a very respected profession. During the 1990s I worked on contracts here that paid anything up to NZD...
Sure, but the thing is there are people happy to work for that kind of money. In that part of the world $20 is probably regarded as a decent rate for the hours spent on the job, and the $100 paid is an average monthly income!

Bottom line, why pay exorbitant rates when you can get the same job done for a fraction of the cost and at the same time help support struggling young programmers in a disadvantaged country? That's the beauty of the globalized world we live in, opportunity and competition, and everyone benefits!
  • #93
  • Jun 4, 2010 12:58am Jun 4, 2010 12:58am
Quoting pipmutt
Bottom line, why pay exorbitant rates when you can get the same job done for a fraction of the cost and at the same time help support struggling young programmers in a disadvantaged country? That's the beauty of the globalized world we live in, opportunity and competition, and everyone benefits!
Legal reasons. Someone in the third world cannot be held to a contract like an NDA. Someone in the US can be.
  • #94
  • Jun 4, 2010 12:59am Jun 4, 2010 12:59am
Quoting mikkom
There is a lot of discussion about this in "volatility trading" scene. Read about black scholes, implied volatility and later theories etc.
I know... it's pretty grizzly. I started a while back here:
http://dspace.iimk.ac.in/bitstream/2...Volatility.pdf
My maths wasn't up to scratch to reproduce a Garch model, so I modelled EWMA - and it came out as slightly worse than a simple ATR. Typical. It does seem logical IVs should be better predictors and at least they're easy enough to get hold of. But unless I can prove it (to myself) then it seems like a lot of work with no obvious upside.

Emerson - the first sentence hits the nail on the head. It reduces losing trades, but at the expense of profits as you're scaling out before target. Way too complex for me but model it and see...

Hanover - I know some people believe all charts are created equal (I give it 2 posts before someone posts one and says spot the difference, aggh) but if they'd just put the charts down, you can easily tell the difference between OHLC daily and hourly by measuring the ratio of noise:move. Reading Rabid's post - never knew it had a name... efficiency ratio. Nice.

Rabid - I know this wasn't addressed to me, but (in spite of the nods/winks etc around here) orderflow isn't a complete black box - eg IFR data is pretty specific about what levels have interest and stacks of orders.
  • #95
  • Jun 4, 2010 1:28am Jun 4, 2010 1:28am
Quoting Rabid
Legal reasons. Someone in the third world cannot be held to a contract like an NDA. Someone in the US can be.
There's an easy solution to that one when it comes to EA's and such, just disguise/change parameters.

Copyright and patent are a minefield anyway, infringements are a daily occurrence, people are plagiarizing others work every day and getting away with it.

But let's keep this in context here, we're talking about indicators for MT4!
  • #96
  • Jun 4, 2010 1:36am Jun 4, 2010 1:36am
Quoting pipmutt
Sure, but the thing is there are people happy to work for that kind of money. In that part of the world $20 is probably regarded as a decent rate for the hours spent on the job, and the $100 paid is an average monthly income!

Bottom line, why pay exorbitant rates when you can get the same job done for a fraction of the cost and at the same time help support struggling young programmers in a disadvantaged country? That's the beauty of the globalized world we live in, opportunity and competition, and everyone benefits!
Sorry, I misread your post, somehow I didn't see the part about his being from an Eastern European country.
  • #97
  • Jun 4, 2010 1:40am Jun 4, 2010 1:40am
Quoting triphop
I know... it's pretty grizzly. I started a while back here:
Interesting link. Ran into something the other day, reminds me of that a little. I'll attach it.

It's not entirely surprising that an EWMA would fair worse than a simple ATR tho. ATR just averages range over a time period, the EWMA would put more emphasis on recent events. Considering that volatility in forex (at least on a pair by pair basis) tends to regress pretty strongly to the mean, an EWMA might be putting too much emphasis on noise.

Quoting triphop
Reading Rabid's post - never knew it had a name... efficiency ratio. Nice.
I wish I could take credit for it, lol. I first encountered it in a book by Perry Kaufman entitled "Smarter Trading." It uses the efficiency ratio to build an adaptive moving average. It's a worthwhile read, especially when people bring out the "the markets are fractal, every timeframe is the same!" argument. If you're bored sometime, do an hour by hour study of this. Some hours tend to be much more trendly than others.

Quoting triphop
Rabid - I know this wasn't addressed to me, but (in spite of the nods/winks etc around here) orderflow isn't a complete black box - eg IFR data is pretty specific about what levels have interest and stacks of orders.
Hrm. Care to elaborate on that?
Attached File(s)
File Type: pdf 0911.0454.pdf   1.5 MB | 798 downloads
  • #98
  • Jun 4, 2010 1:43am Jun 4, 2010 1:43am
Quoting pipmutt
There's an easy solution to that one when it comes to EA's and such, just disguise/change parameters.
Doesn't disguise the logic. Given X parameters and Y average options per parameter, there exist only X*Y possible outcomes.

Quoting pipmutt
Copyright and patent are a minefield anyway, infringements are a daily occurrence, people are plagiarizing others work every day and getting away with it.
Which is why it's useful to hire someone full time, and pay them accordingly. If they're under a work for hire agreement, they can't do this w/o running into some serious problems.

Quoting pipmutt
But let's keep this in context here, we're talking about indicators for MT4!
In all fairness, he didn't say that. There are numerous brokerages that offer API access to clients. He could be developing in C++ for all we know.
  • #99
  • Jun 4, 2010 2:06am Jun 4, 2010 2:06am
Quoting emerson
Can you and/or Rabid comment on this trade management technique of getting a free trade without moving SL. I'm quoting myself from another thread. Thanks in advance. .......Check my math on these. .......
As far as I can tell, your math is 100%. The stoploss always needs to be advanced in such a way so that the total pips still at risk does not exceed the total pips safely 'locked in'. Of course you'll need to take the spread into account.

If you're asking as to which approach is the more profitable, IMO it depends on contextual specifics, i.e. based on technical and/or statistical criteria that apply to each individual trade. I don't believe that one exit technique necessarily provides an edge over another, in every conceivable situation.
  • #100
  • Jun 4, 2010 2:07am Jun 4, 2010 2:07am
Quoting Rabid
Doesn't disguise the logic. Given X parameters and Y average options per parameter, there exist only X*Y possible outcomes.
It's the application of that logic which makes or breaks a strategy, apart from which a single indicator might only be part of a strategy, it's a bit like someone copying Porsche's fuel injector design and expecting to build an engine of comparable quality.


Quoting Rabid
Which is why it's useful to hire someone full time, and pay them accordingly. If they're under a work for hire agreement, they can't do this w/o running into some serious problems.
Didn't Microsoft and Goldmans both run into problems just recently, employees (presumably under contract) running off with proprietary information?


Quoting Rabid
In all fairness, he didn't say that. There are numerous brokerages that offer API access to clients. He could be developing in C++ for all we know.
I guess we'll see what's on offer and who's offering it, eventually.
0 traders viewing now