Set it and Forget it, Dollar Cost Averaging | Page 7 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Wide stop loss, Dollar cost averaging or doing nothing 18 replies

Best Indicator For Dollar Cost Averaging Into Crypto 0 replies

10,000 MA Dollar Cost Averaging 6 replies

The Dollar Cost Averaging Pivot Point Trading Method 914 replies

Fast Averaging versus Slow Averaging 1 reply

Attachments: Set it and Forget it, Dollar Cost Averaging
Exit Attachments

Set it and Forget it, Dollar Cost Averaging

  • #121
  • Aug 21, 2012 7:20am Aug 21, 2012 7:20am
In an effort to make things as complicated as I can for myself...here's what I'm trying...

I implemented the trade both ways last evening. Placed pending orders every 50 pips. So I had two additional orders opened to the buy side during the night.

Now instead of closing the buys that hit +50, I decided to move them to BE and then trail them from that point so I wouldn't lose the profit but could maximize it if price continued to climb.

So two of those just closed at about 74 & 80 pips respectively. 154 pips instead of 100 pips. So that worked nice that time.

I still have one buy open for about 25 pips of floating profit. And of course I still have the two sell orders open which combined are about -160 pips.

It will be interesting to see how this nets out with price movement.
  • #122
  • Aug 21, 2012 11:53am Aug 21, 2012 11:53am
Quoting spyderman
In an effort to make things as complicated as I can for myself...here's what I'm trying...

I implemented the trade both ways last evening. Placed pending orders every 50 pips. So I had two additional orders opened to the buy side during the night.

Now instead of closing the buys that hit +50, I decided to move them to BE and then trail them from that point so I wouldn't lose the profit but could maximize it if price continued to climb.

So two of those just closed at about 74 & 80 pips respectively. 154 pips instead of 100 pips. So that worked...
Spidey,

Nice job on the buy side account. How did you handle placing trades on the sell side account? Did you stack the pending orders the same way on the sell side?

I assume your initial buy and sell trades were placed simultaneously? Did you have a specific criteria for making your initial trades?

Keep up the good work!
  • #123
  • Aug 21, 2012 12:45pm Aug 21, 2012 12:45pm
Quoting Bobcat2
Spidey,

Nice job on the buy side account. How did you handle placing trades on the sell side account? Did you stack the pending orders the same way on the sell side?

I assume your initial buy and sell trades were placed simultaneously? Did you have a specific criteria for making your initial trades?

Keep up the good work!
I 'tried' to follow what Jaws was saying. I placed my initial sell order (along with buy, both random). Then I placed another sell at the beginning of the next day. (I'm planning on placing my orders shortly after the new daily bar which for me is 5:00 EDT. I don't want to get up at midnight. ) Those sells of course are still open and in hefty DD.

I've also placed buy and sell pendings at 50 pip intervals. I adjust the pending price to reflect the entry price on the standard daily entry for buy/sell.

So now I've had 3 positions close for 74, 59 & 80 pips. Another buy is riding at 31 pips. Sells are running -270. So I'm still in the hole taking everything into consideration. See what happens when price reverses.
  • #124
  • Aug 21, 2012 1:00pm Aug 21, 2012 1:00pm
Quoting spyderman

So now I've had 3 positions close for 74, 59 & 80 pips. Another buy is riding at 31 pips. Sells are running -270. So I'm still in the hole taking everything into consideration. See what happens when price reverses.

If I still had those buy trades open they'd now be in profit a total of 334 pips. So what I'm going to try and do is trail these orders by a percentage of profit, namely 80%. So if the trade is 50 pips in profit it will trail 10 pips behind...100 pips in profit, 20 pips behind, etc.
  • #125
  • Aug 21, 2012 2:24pm Aug 21, 2012 2:24pm
hi all

i feel like an expert in averaging ...and i can say that it's a good way to make money , until ......you get burned.

it will always happen if you don't have a lot of cash to invest , otherwyse you can count on a standard percentage on your account , let's say 3-5 % per month .

i have an ea , that i post here , play with it in all the way you want , you will see that it's good to collect a drop in the ocean every day, but that drop is yours for sure ...if you have money enough not to suffer drawdown .

some explanation on the settings .

- use with money management off if you increase the lot size .
- lotsx : increase lot size
- reverse signal = always true .
- first trade 0= sell 1 = buy
- margin : space between each trade

let me know if you have good settings .

thanks
Attached File(s)
File Type: mq4 Up Down Touch EA mod 2.mq4   75 KB | 491 downloads
  • #126
  • Aug 21, 2012 2:32pm Aug 21, 2012 2:32pm
another good way of averaging is using two correlated cross

buy eurusd usdchf

if it's in profit well, if not buy another set and so on .

drawdown would be less, and eurusd is not always faster than usdchf , so if a long trend happen and you are caught in countertrending , you can also be saved from usdchf running opposite drection faster than eurusd .....
  • #127
  • Aug 21, 2012 11:04pm Aug 21, 2012 11:04pm
for my own long/short system I have a larger TP of 150 pips for each individual trade.
that way the long or short side wont be closed in 1 hit.

the advantage of a gradual closing of positions and a higher TP means that the long and short side are far more hedged for longer.
  • #128
  • Aug 22, 2012 12:11am Aug 22, 2012 12:11am
As a student of averaging and backtesting, I seem to always come away with either broken ideas or broken accounts. Probably one reason why I've basically left forex for the foreseeable future. But enough about me.

Check out 2010:

http://www.forexfactory.com/attachme...1&d=1345611388

It's not a spectacular year, nothing too obvious except a two very prominent trends. So I thought, why not backtest this.

The set up:

Using TrueFX free tick files (only way to backtest is w/ tick data, granularity is key to robust results).

Random time of day entry--I basically tell software to start X ticks into the file, given I don't even know then TrueFX timestamps their data (UTC, I think), it doesn't really matter. In this run, it trades at 3:48AM whatever time zone...

I go long and short at the same time every day.

Target is 50 pips from average position price.

I have not calculated net gain loss, I was more concerned w/ how many entries are needed. Jaws expects ~20.

The Results:

That downtrend from Jan to June requires 133 positions--YES, 133 buys are needed until the position is exited in August. It takes 2 months just to get to the average buy price.

AND

The uptrend from June to Sept requires over 30 positions short before it is closed.

AND

There is a period where both averages are so far removed from current price that there is 133 positions long AND 23 positions short simultaneously such that an entire month can go by without a single closing trade.

AND

Just to finish the year, you end up with 41 positions long and 3 positions short on 12/31/2010. (Plus or minus, I have not checked every trade by hand, but the few I checked matched output).

OUCH...

Sloppy text file is attached for those that want to see the output and at least eyeball results...

Don't shoot the messenger...BroncoCap

EDIT: Looking at the text file shows that a week in February is missing...free tick files, get what I pay for, I guess. Probably effects the results a bit, but the takeaway remains the same I believe.
Attached Image (click to enlarge)
Click to Enlarge Name: Euro2010.jpg Size: 111 KB
Attached File(s)
File Type: txt Euro2010.txt   59 KB | 376 downloads
  • #129
  • Aug 22, 2012 12:49am Aug 22, 2012 12:49am
Who needs sleep???

Tested 2011 using same criteria and settings in prior post. Here is the year's price action:

http://www.forexfactory.com/attachme...1&d=1345614458

As you can see, a bit choppier, beginning of the year and end of the year were main trends, while chop had wide range leading to more hit targets.

Feb through May had many 7-15 position runs, but ultimately retracted to targets sooner than expected just looking at the chart. Intraday range was key to 2011 coupled with slightly reduced yearly high-low range (I presume).

Worse period is end of 2011 where system would be flat in the short account, but have 46 positions long, yikes.

Anyways, your mileage may vary...BroncoCap
Attached Image (click to enlarge)
Click to Enlarge Name: Euro2011.jpg Size: 113 KB
Attached File(s)
File Type: txt Euro2011.txt   58 KB | 308 downloads
  • #130
  • Aug 22, 2012 1:12am Aug 22, 2012 1:12am
Before anyone asks, here are two results for 2010 using 100pip and 20pip targets. 100 pip is painful as it went 133 positions + 41 positions the other direction, logically. 20 pips is interesting as the most dual positions I saw was ~100 + 10. Therefore you could look at adding size to the opposite side if one side gets lopsided. Say after 20 positions, double the opposite, after 40, do 3x or 4x...

Anyways, way, way too late here and 1000 other things I should've been doing rather than testing, but you never know...

Cheers!
BroncoCap
Attached File(s)
File Type: txt Euro2010.100pipTarget.txt   61 KB | 306 downloads
File Type: txt Euro2010.20pipTarget.txt   56 KB | 301 downloads
  • #131
  • Aug 22, 2012 3:31am Aug 22, 2012 3:31am
E/U is not a good pair for a long/short system, too trendy and the yearly range is far too high.

you should look at low range pairs like A/C, A/N, U/C, E/G etc. or a combination if that range is even lower.
  • #132
  • Aug 22, 2012 6:23am Aug 22, 2012 6:23am
Bronco,

Was curious if your testing is just allowing for 1 trade each way per day or is it placing a order ever 50 pips. That will greatly effect the outcome.

Also extracting more pips from the profitable trades by trailing will increase the returns.
  • #133
  • Aug 22, 2012 7:29am Aug 22, 2012 7:29am
Quoting yonnie
E/U is not a good pair for a long/short system, too trendy and the yearly range is far too high.

you should look at low range pairs like A/C, A/N, U/C, E/G etc. or a combination if that range is even lower.
Yup, I do EURUSD for the spread, liquidity, etc, but for this idea, no go..
  • #134
  • Aug 22, 2012 7:30am Aug 22, 2012 7:30am
Quoting spyderman
Bronco,

Was curious if your testing is just allowing for 1 trade each way per day or is it placing a order ever 50 pips. That will greatly effect the outcome.

Also extracting more pips from the profitable trades by trailing will increase the returns.
This was 1 trade per day. I've done x pips (also known as grid trading) with similar results. BUT, I did not do it using two sub-accounts, I'll see if I can dig out those results...
  • #135
  • Aug 22, 2012 8:37am Aug 22, 2012 8:37am
Okay,

Jaws, I'm anxious to hear you weigh in on this.

It seems we've been having a lot of discussion on direction and positions open this way or that. For me, we've maybe gotten away from the core issue.

So for my clarity, if no one else's, here's what I see.

The main issue, of course, is can this strategy be profitable. I don't care what direction, or how many trades you take...at the end of the 'day', are we profitable.

Secondly. Can our account handle the margin requirements in order to achieve that profitability.

Addressing the second issue first. It's going to depend to a large degree on the individual brokers we may be using. Brokers handle margin differently. Some brokers offset the margin you are using for a long position by any short positions you have on that currency. So obviously you can have a much greater amount of positions open as long as they 'hedge' each other. Others don't allow that.

Some brokers figure your margin on your account balance, whereas some figure it on your account equity. I have a live account with MB Trading. Doesn't matter if I have a million dollars (that would be nice) in floating profit, that is not usable against margin until I close those orders. (Pretty smart way to figure it IMO)

I have other live accounts with brokers that adjust your margin as your open trades go in and out of profit. So everyone will need to examine the strategy through that lense.

Now for the first point...
  • #136
  • Aug 22, 2012 8:54am Aug 22, 2012 8:54am
Can this be profitable?

Obviously if we enter one short and one long position each day with no s/l and no t/p, it doesn't matter what price does, the net effect will be zero (obviously for the purpose of the discussion we are not counting spread and lag time between placing the buy and sell).

Even if we close a buy order at 50 pips tp and immediately enter another buy, the net effect will be the same (again not counting spreads, etc.)

So Bronco has run several tests and mentions the huge amount of positions opened before a retracement closed them. I don't think that is really the issue. Those many positions should have been offset by the many positions that would have closed at t/p. So the account balance (and the resulting increase in margin available) would have been increasing. Unfortunately those gains would be offset by the large dd of those open positions.

So the only way to be profitable trading both ways is for your open positions in the direction of price action to out number the positions the other way. That's where stacking interday trades helps.

If we place a buy and sell at the beginning of the day and then price moves 150 pips north in a day we would have hopefully had three additional buy positions opened. But if we closed each of those at 50 pips profit, we'd still be at a BE scenario. 3 buys closed at 50 pips each and 1 sell open at -150 pips.

So we have to let those winners run to be profitable. If we were able to hold those positions open then we would be sitting at +150 pips:

1 buy +150
1 buy +100
1 buy +50
1 sell -150

That's why I'm trying to trail y orders once they hit that 50 pip profit point. I don't see any other way to be profitable.

Sorry about the length here but I wanted lay all the points out and see what everyone thought.
  • #137
  • Aug 22, 2012 9:10am Aug 22, 2012 9:10am
I've been looking at posts and trying for solve the issue with huge runs and massive DD and here is what I came up with.

I think I found a way to not only survive, but thrive through a hellish death spiral. Trade in both directions, use multiple limit orders (200+ pips out) to catch big moves and use a 150 pip from average Sl to close all positions and limit losses during a big move. Then when you get a huge move like that September 2008 you make a killing on the winning side and losses are minimal as long as you set up enough limit orders to cover the entire move. After a loss wait till the next day to trade that direction again.

So what do you think.
  • #138
  • Aug 22, 2012 9:12am Aug 22, 2012 9:12am
Another aspect that has to be factored in, and I think Jaws has made this point, is that your entry orders every 50 pips need to be reset in line with your automatic daily buy/sell orders. So if today you enter a buy and sell order at 1.24750 then your next interday entries would be at 1.25250 and 1.24250.

Otherwise if you had a big move to the upside of 150 pips in one day you wouldn't have any sell orders in place coming back down on subsequent days.
  • #139
  • Aug 22, 2012 9:12am Aug 22, 2012 9:12am
Quoting spyderman
Can this be profitable?

Obviously if we enter one short and one long position each day with no s/l and no t/p, it doesn't matter what price does, the net effect will be zero (obviously for the purpose of the discussion we are not counting spread and lag time between placing the buy and sell).

Even if we close a buy order at 50 pips tp and immediately enter another buy, the net effect will be the same (again not counting spreads, etc.)

So Bronco has run several tests and mentions the huge amount of positions opened before a retracement closed...
Good analysis, but note. My results were Jaws' original system. Open only 1 trade per day at a given time. Therefore, the system I tested is a likely loser because losses (aka many positions) form an exponential negative equity curve whereas gains are linear. That is to say, winners only give a fixed amount (x pips) whereas losses are unlimited.

Said another way, think of it like this. In any given week, there will be 10 total trades, 5 long and 5 short. The absolute max gain would be both sides hitting targets and accts flat over the weekend for 10 winners (at 50 pips) = 500 pips. Worse that could happen is a straight trend w/ no pull backs. You would have 5 winners in the direction of the trend (250 total pips), but an unrelated and unbounded loss related to the counter-trend trades...therein lies the rub.

Now your system of opening a new trade may overcome the linear vs exponential issue. I'll get back to you...eventually...

But as I say, I was more concerned w/ max open positions rather than net PnL. I will add that to my software soon, but not now...work calls...
  • #140
  • Aug 22, 2012 9:13am Aug 22, 2012 9:13am
Quoting spyderman
Can this be profitable?

Obviously if we enter one short and one long position each day with no s/l and no t/p, it doesn't matter what price does, the net effect will be zero (obviously for the purpose of the discussion we are not counting spread and lag time between placing the buy and sell).

Even if we close a buy order at 50 pips tp and immediately enter another buy, the net effect will be the same (again not counting spreads, etc.)

So Bronco has run several tests and mentions the huge amount of positions opened before a retracement closed...
I went the other way, I decided to limit losses and leave the profitable end alone.
3 traders viewing now