How to Pyramid Entries to Increase Initial Risk/Reward 10-fold | Page 3 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

What's your TP? Fixed reward vs dynamic reward 5 replies

Precision, low-risk entries for discretionary systems. 36 replies

Pure Risk to reward trade System - 1 to 3- Totally random entries 178 replies

Rationale behind "Risk Per Trade" and "Risk/Reward Ratio"? 8 replies

Risk Reward Ratio and Account Risk 30 replies

Attachments: How to Pyramid Entries to Increase Initial Risk/Reward 10-fold
Exit Attachments

How to Pyramid Entries to Increase Initial Risk/Reward 10-fold

  • #41
  • Aug 26, 2013 12:42am Aug 26, 2013 12:42am
Quoting 2+2=4ex
As has been discussed in other threads, pyramiding is nothing more than a collection of individual trades. Stacking another trade on top of an already existing trade does not change the dynamics of your exposure as opposed to closing original trade and opening a new trade of larger position size. In a hypothetical scenario of all things being equal, risk is proportional to reward. Of course not all things are equal in the real market. Your success will depend on finding these inequalities (also known as inefficiencies) and then exploiting them....
Exactly right, and well explained.
  • #42
  • Aug 26, 2013 4:49am Aug 26, 2013 4:49am
Quoting yonnie
who cares what you call the trades: point is that in a well established trend it is very smart to open a new position after each meaningful retrace and the trend resuming. with the older positions at least @ BE under the last swing, you have only 1 trade at risk, 1 @ BE and the other positions in profit should price hit the stop. ofcourse the exit price is relative to the average entry price, but if I have 10-15 positions in a good trend, the profit can be quite large. I`m in good company with Jesse Livermore.Smile
You don't have a single trade a risk, all of them are! Retracements can be deep. All of the positions were at risk before trailing the SL. You will accumulate several losers in the way you wouldn't have had with the first position only.

More important: BE is a loser! Think about it. You took risk for no reward. Either you lose or you make no profit. This is a losing game long term. All those BE trades impair your expectancy (your edge).


If it is indeed smart to enter in each retracement is it really smart to try and close all of them at the end? The answer is probably no.

Yes "the profit can be quite large" but the risk increases faster. Up to twice as fast. Your profits grow in N(N+1)/2 ; your risk grows in a.N. 0<a<1, unknown. At worse the RR is halved. That's what I call hidden over-leverage.

I proved a grid is worse than "set and forget". A pyramid is an un-evenly scaled grid, so I think it is bad too. But the grid step is agnostic to the PA while the pyramid isn't...

So far I proved that for almost any winning strategy:
- Trailing SL to BE is bad
- Taking partial profit at any time is bad
- Taking partial profit at any time and trailing SL to BE is bad
- Scaling out of a loser is bad
- Adding to a winner in a grid is bad
- Adding to a loser is bad, even when you can limit the risk to a constant.

I'm still missing the formal proof the pyramid is bad. Unfortunately the OP left his own thread before showing anything.
No greed. No fear. Just maths.
  • #43
  • Aug 26, 2013 6:50am Aug 26, 2013 6:50am
The original poster isn't going to post here again until he gets his trade explorer back up to snuff. Massive losses last week for him. He trades the account like any gambler would and as such, I wouldn't expect anything profound to come forth from the thread starter.

With kind regards,
MK
  • #44
  • Aug 26, 2013 9:04am Aug 26, 2013 9:04am
I think I found a way to prove that adding a position to a winning trade lowers the expectancy of a strategy in the general case. A fortiori, adding more than one makes things worse. If I've time this week-end I'll post the demonstration in Expectancy Management.
No greed. No fear. Just maths.
  • #45
  • Aug 26, 2013 11:21am Aug 26, 2013 11:21am
wow pipmeup, you prove everybody wrong..........do you have actually time to trade?
  • #46
  • Aug 26, 2013 11:41am Aug 26, 2013 11:41am
Quoting PipMeUp
{quote} You don't have a single trade a risk, all of them are! Retracements can be deep. All of the positions were at risk before trailing the SL. You will accumulate several losers in the way you wouldn't have had with the first position only. More important: BE is a loser! Think about it. You took risk for no reward. Either you lose or you make no profit. This is a losing game long term. All those BE trades impair your expectancy (your edge). If it is indeed smart to enter in each retracement is it really smart to try and close all of them at...

I have found most of this statements to be true.kudos
wo-yoy! wo-yoy! wo-yoy! wo-yoi! wo-yoy-yoy-yoy!
  • #47
  • Aug 26, 2013 12:14pm Aug 26, 2013 12:14pm
Quoting yonnie
wow pipmeup, you prove everybody wrong..........do you have actually time to trade?
I would by far prefer proving a single person right, especially the OP. It's actually much more difficult to prove a strategy correct. But if maths say you are wrong, you are just wrong. Like it or not, accept it or not. If people prefer trading wishful thinking, their business. It's already hard to find an edge, better not destroying it with a wrong trade management!
If you have a problem with your time table, try the daily TF <:°)
No greed. No fear. Just maths.
  • #48
  • Aug 26, 2013 4:09pm Aug 26, 2013 4:09pm
Before, I thought adding positions to winning trade is a good technique. I created an EA and test the result with 20+ years data. The result is disappointing, first and single entry give better result.

I agree that each trade is responsible for it's own risk. Second trade is made because the first one is on green pip will not make the second one have better chance to be green also. In fact, RRR will become worse for the second trade because we enter it at higher price.

My advise is keep it simple. All we need to do in "method" phase is to find entry time when price chance to up or down is consistently not 50:50 and take the trade according to it. Once you find this, test it on different pairs in long term, 10+ years data is needed. Do not concentrate on rising your PF, consistently profitable in long term and many pairs is more important, higher PF is only bonus. You should take attention on your largest drawdown too. If you see satisfied result on most pairs, go to second phase, "Money management". Anti martingale position sizing and diversification is the key to explode your profit.
  • #49
  • Aug 26, 2013 4:38pm Aug 26, 2013 4:38pm
to add a trade because your last trade is in profit is silly.
an EA would never differentiate between trends and ranges, retraces and continuation like a trader would, so the results of an EA would be useless to me.

I trade only when I see a strong trend on the daily chart, like the AUD fall recently and add only when price retraces and confirms that the trend is about to resume. Only trading A/N at the moment.

and you want me to have only 1 trade in A/N in all those months?

do I get it always right? ofcourse not, but what system does.

as long as you trade a system consistently and make consistent profits, what is there not to like about it.
  • #50
  • Aug 26, 2013 9:06pm Aug 26, 2013 9:06pm
wtf... 48 replies...

insane.. this has taken on life of its own i see..


as you all can see i have been far too busy to finish the original post...

maybe one day. i try on weekends but... weekends are often just as busy as the weekdays.

id like to get some charts up here soon but no promises of anything. i just need to find a little time




anyway, carry on discussion!
Be hopeful in a winning position, and fearful in a losing position.
  • #51
  • Aug 27, 2013 2:43am Aug 27, 2013 2:43am
Yonnie is going to be happy. After checking my maths I realized that I proved one solution of pyramiding to be wrong but not in the general case. I need to prove this pyramid is the best possible. I don't know how to.
The pyramid I tested guarantees that if the first position is a winner all the others will be as well with no edge required. No magic, you just need to set the next SL at the same place as the first one, never trailing. If the first is a winner all the other will be. Conversly if the first one is a loser all the additional positions will also lose. But this is anyway true for any pyramid (remember BE=loser).

@Yonnie: "[...] you want me to have only 1 trade in A/N in all those months?"
Absolutely NOT. Just the next positions shall not try to reach the end of the trend. The best I found so far is to keep the trend trade running as long as the trend lasts and catch the next swings taking profit as near as possible of their respective top. Never trade the counter-trend swings (I can prove this last point ;-) ).

What people shall realize is that if Rag2RichesFX can show a valid pyramid method the implications are AMAZING.
No greed. No fear. Just maths.
  • #52
  • Aug 27, 2013 3:01am Aug 27, 2013 3:01am
Quoting PipMeUp
Yonnie is going to be happy. After checking my maths I realized that I proved one solution of pyramiding to be wrong but not in the general case. I need to prove this pyramid is the best possible. I don't know how to. The pyramid I tested guarantees that if the first position is a winner all the others will be as well with no edge required. No magic, you just need to set the next SL at the same place as the first one, never trailing. If the first is a winner all the other will be. Conversly if the first one is a loser all the additional positions...
PipMeUp, if you enter second trade with same SL with first one, then you will get bigger SL and smaller TP for second trade. This only reduce your PF and increase largest drawdown.
  • #53
  • Aug 27, 2013 3:39am Aug 27, 2013 3:39am
Quoting Special One
{quote} PipMeUp, if you enter second trade with same SL with first one, then you will get bigger SL and smaller TP for second trade. This only reduce your PF and increase largest drawdown.
Absolutely! (Except it's not about profit factor but expectancy)
That's exactly why I'd like to show this is the best possible pyramid! Any pyramid is worse than the best one. If the best pyramid is inefficient, any pyramid is inefficient. ;-)
No greed. No fear. Just maths.
  • #54
  • Aug 27, 2013 5:37pm Aug 27, 2013 5:37pm
Quoting PipMeUp
Yonnie is going to be happy. The pyramid I tested guarantees that if the first position is a winner all the others will be as well with no edge required. No magic, you just need to set the next SL at the same place as the first one, never trailing. If the first is a winner all the other will be. Conversly if the first one is a loser all the additional positions will also lose.
phewwwwwwwwww, I just scraped through I think.......

I cant see that traders would place their stop on all positions at the same place as the first trade.
suppose you scale in @ 1.3000 - 1.3500 - 1.4100 - 1.4450 I would not want to lose all those pips and exit at my first stop loss of 1.2800.

I don't understand you pmu: on the one hand you say to exit near the top of the swing - on the other hand you say the above.

sorry but my brain cells are not really well developed.
some even say to leave your brain at the door in trading!
  • #55
  • Aug 28, 2013 2:23am Aug 28, 2013 2:23am
Did you read my previous post? If I can prove this is the best way to pyramid and you consider it stupid (which it is), you shall consider even more stupid to trade any other pyramid (because worse). I just fail at proving it is the best. That's why I'd like R2RFX to post his idea to see if he has found better: a counter example proof that I'm wrong. I'd be happy to be wrong because if it is the case, as I wrote earlier, the implications are amazing.

Quoting yonnie
some even say to leave your brain at the door in trading!
You can find them in the "Martingale makes you rich" kind of threads
No greed. No fear. Just maths.
  • #56
  • Aug 28, 2013 2:32am Aug 28, 2013 2:32am
Quoting PipMeUp
{quote}So far I proved that for almost any winning strategy: - Trailing SL to BE is bad - Taking partial profit at any time is bad - Taking partial profit at any time and trailing SL to BE is bad - Scaling out of a loser is bad - Adding to a winner in a grid is bad - Adding to a loser is bad, even when you can limit the risk to a constant. I'm still missing the formal proof the pyramid is bad. Unfortunately the OP left his own thread before showing anything.
I want to support something here. Please think a moment WHAT a set to BE is: It is in most cases a trailing stop - nothing more. It's named "BE" BUT always keep in mind the market don'T care if your are "BE" or +5 pips +10 or whatever. You always lower your expecatancy of winners with this strategy(and the expectancy of losers of course). The bottom line is that in 90% of all cases setting to BE makes no sense. Most people set BE too fast. BE makes only sense if you are hitted almost your TP. Then you can sqeeze a bit more out of your RR ratio.

*confirmed*

Taking partial profit I have same opinion. It's not worth. Tested it also.

Adding to a loser is bad: Well...it depends. On imperfect market it can make sense. Why ? cause if you have an Edge you can avg. your RR ratios. If you are not sure when you RR is best you can avg. the best RR's by this method. So if the first entry RR performs weaker and the 2nd performs better you get avg. good result.

formal proof the pyramid is bad: I cannot give you the formula but I know why pyramiding does not work. The high advantage of a pyramiding trade (one risk, xx chances of win) has two huge flaws: First your avg. in rises. Please be aware !!! : If you avg. in on 10 positions rises only 10 pips, your risk increases 100 pips !! (to the first position) This is often not mentioned.
2nd flaw is: Setting to BE. IF you think your are smart this way you already lost. I prooved various times (on practise) that you miss this way too many winners because of down swings. If you DON'T go to BE, well... then you loose up to let's say 500 pips to the initial position (why ? cause all your positions now turn negativ and EVERY pip you loose you loose xx positions*xx pips !)

Another negative aspect on pyramiding is that on RR 1:1 (roulette) you definetely lose over time. This is caused by busting huge winning trades and missing huge winners which are needed to equal the losses before. that's why RR most times sink on using pyramiding instead of using one position.

thanks for reading, thanks for the thread, helps me !
  • #57
  • Aug 28, 2013 5:05am Aug 28, 2013 5:05am
Another negative aspect on pyramiding is that the BE point follows you. A small retracement pushes you down by a few pips multiplied by a huge lot size.


Quoting JackintheBox
Adding to a loser is bad: Well...it depends. On imperfect market it can make sense. Why ? cause if you have an Edge you can avg. your RR ratios. If you are not sure when you RR is best you can avg. the best RR's by this method. So if the first entry RR performs weaker and the 2nd performs better you get avg. good result.
There is one case where you are correct. I couldn't find a system matching the requirements: a good win/loss and a big R:R at the same time. A position can only make money linearily in the number of pips. An averaging down will therefore generate no more than a profit linear in the number of positions. You cannot make better. The best possible averaging down method will keep the risk constant. So the best possible averaging down method will increase the R:R linearily in the number of positions. You can achieve this. Your initial SL is 1 unit of risk. Add when you're half way from the entry to the SL. Add another one when you're half way from the previous entry and the SL. And so on. Your risk is 1 + 1/2 + 1/4 + ... This is less than 2. Worse case you double the initial risk. 2 is a constant. The potential reward is R + R+1/2 + R+3/4 + ... which is linear in the number of positions. Your R:R grows linearily with the number of positions.

The probability of an event X is greater than the probaility of "X and something else". That's simply because you add a constraint. Thus if you have a probability P of winning, your probability of winning after the next position is open is smaller than P. The probability of the next position to be a winner is always strictly smaller than the previous one. But the R:R is improved this may compensate.

With 2 additional positions (3 positions open) the improvement of expectancy is strictly smaller than (2R+13/4)P-2. This has to be positive. Knowning that the bigger R the smaller p, I have yet to find a strategy where it would work.

Also note the paradox that you make more money when you timing is bad... This makes me think that the required system doesn't exist.
No greed. No fear. Just maths.
  • #58
  • Edited 10:11pm Aug 28, 2013 8:43pm | Edited 10:11pm
Quoting PipMeUp
Did you read my previous post? If I can prove this is the best way to pyramid and you consider it stupid (which it is), you shall consider even more stupid to trade any other pyramid (because worse). I just fail at proving it is the best. That's why I'd like R2RFX to post his idea to see if he has found better: a counter example proof that I'm wrong. I'd be happy to be wrong because if it is the case, as I wrote earlier, the implications are amazing. {quote} You can find them in the "Martingale makes you rich" kind of threads
Forget about R2RFX providing that proof. If you were following his trade explorer after he started this thread you'll see clear evidence that whatever he was doing virtually blew him out. He's taken the trade explorer offline now to hide the results, but at one point he was up something like 411% and in 3 days it was down to 150% and continued to decline from there before he made the trade explorer private. I know you are an on the ball sort of chap so surely you didn't overlook this, but I post it here for others that may stumble across this "how to" thread.

With kind regards,
MK
  • #59
  • Aug 29, 2013 2:52am Aug 29, 2013 2:52am
I noticed. He was +426% at the highest ;-)
But a DD doesn't prove his method wrong. His newest thread, however, brings discredit...
No greed. No fear. Just maths.
  • #60
  • Aug 29, 2013 5:43am Aug 29, 2013 5:43am
It wasn't just DD, it was pure gambling with reckless abandon.

All my best,
MK
0 traders viewing now