Pyramiding Positions

[pavilion103]

I am new to trading and am currently trading a simulator. I haven't touched on pyramiding as yet but want to now get some advice.

I trade at the end of each day and trail my stops (currently 2xATR). I want to determine two points:

1. When is the best time to pryamid (e.g. when I can use the profit to pyramid and if I exit the position I break even? or do I pyramid when it looks like an another opportunity I would have entered anyway if I wasn't already in.... etc)

2. How much should I be adding to subsequent trades in comparison to my initial entry.


I've very quickly circles some entry points in this chart. Does this look reasonable? Or are the entries too close together/too far apart.

 

[captain black]

1. When is the best time to pryamid (e.g. when I can use the profit to pyramid and if I exit the position I break even? or do I pyramid when it looks like an another opportunity I would have entered anyway if I wasn't already in.... etc)

2. How much should I be adding to subsequent trades in comparison to my initial entry.

I notice you're using Amibroker; one of it's strengths is the ability to accurately test various scenarios such as pyramiding. You'll no doubt get a lot of varying opinions on things such as pyramiding, stop placement etc. which are great for introducing new ideas but at the end of the day all that really matters is the numbers and the best way to skew the numbers in your favour is to rigorously test any ideas such as pyramiding. Use Monte Carlo testing, walk forward testing etc. to validate your results.

The guide to pyramid code is below:

http://www.amibroker.com/guide/h_pyramid.html

Enjoy

 

[Tanaka]

Good subject Gordon, I'll be interested to see how others approach pyramiding.

Just one observation, those circles are all done in hindsight, do you have rules about how to spot those situations when live?

I have only ever played a little with pyramiding, I have a couple of EOD systems that are profitable so I tested simple % profit based pyramiding, with these systems I found that a 3 position pyramid adding to position after 2.5% profit and closing all positions at sell signal on average netted me 15-18% more profit. These results like I said are based on a couple of systems that I developed so probably won't be achieved in other peoples situations.

I would like to develop a more sophisticated pyramiding system to be backtested but I'm no programmer.

 

[tech/a]

(1) I pyramid as quickly as I can---see (2).
(2) The pyramid opportunity must scream buy me.(Normally a small tight consolidation after a buy of a breakout).
(3) I position size it as a separate trade and set buy and sell stops as an individual trade.
(4) As the trade progresses the sell/trailing stop will become the same.
(5) I'm happy to pyramid more than 2wice.
(6) If I happen to see a stock being bought heavily I will buy a larger than average risk (My average is 05 to 1% of Portfolio equity) 3-5% and close the whole Pyramid trade on first sign of weakness (I trail stops a few ticks back---you need to be near a screen).
This can be very lucrative.

 

[tech/a]

Here is a current example HUN

Note the shrinking volume on the small consolidations and the increased volume (Effort) on the breakouts.




QNR is another if you wish to take a look.

 

[Tanaka]

Here is a current example HUN

Note the shrinking volume on the small consolidations and the increased volume (Effort) on the breakouts.

Good example tech! I took a position on HUN on Wednesday for the exact reason didn't add to it on friday because I was AFK stopped out 3 times last week on other trades and this one has already made up for the stop outs I may add that is my pick for the April tippings, aren't I just awesome :

but going back to pyramiding I agree with tech, if it has moved the way you had planned for, jump on it straight away.

 

[Synergy]

(3) I position size it as a separate trade and set buy and sell stops as an individual trade.

I don't consider this to be pyramiding but simply holding more than one position in a stock. You are doing nothing but taking another trade as you would do with any other trade you take. At least until you treat them as the same position.

I think I've posted on pyramiding several times in the past.

If you are a systems trader then I'd assume you have a reasonably low risk profitable entry. The entry is important. It's something you've worked on to provide the best results. Buying when you buy provides the best results.

So why buy somewhere else? Why buy when you know it doesn't produce the best results? If the pyramiding entry produced the best results, why not move all of your entries to that point in the trade?

I feel that if pyramiding works for you, it's because something else can be improved. Either your initial entry is too early in a trend or before a trend starts. Perhaps it's too high risk. Or maybe you are not getting enough buy signals and simply have spare cash to throw at ideas like these.

Either way, any new buy order needs to be evaluated on it's on merits at that exact point in time. Whether you already hold a position in that stock or not. For me, the best entry to a trade is where I've work out to be the best entry. Not before or after that point.

 

[markhocky]

...So why buy somewhere else? Why buy when you know it doesn't produce the best results? If the pyramiding entry produced the best results, why not move all of your entries to that point in the trade?...

Sorry, have to disagree there Synergy.

You may not always be able to wait for the absolute best trade, because it will decrease your trade frequency with a corresponding reduction in overall return. I'd be more than happy to enter in a 'good' trade provided i expected it to be the best allocation for my stake at that time (i.e. no 'great' trades available). If it later shows the signs of being a 'great' trade then you should add to your position. I guess this is analogous to Tech's example of a consolidation after a breakout signifying an improvement in expectations.

Which basically leads to my philosophy on pyramiding; your position size should be increased/decreased in line with increases/decreases in expectations of reward vs risk.

How to determine the changes in expectations (or change in trade characteristics) are many and varied. My belief is that as long as you have some evidence that it works (e.g. like Tanaka and his backtest results), then go for it.

Oh, and +1 vote from me on great discussion topic.

 

[Synergy]

Yeah it is a good discussion topic, and I'm trying to be open minded about it... I'd love if I could agree with you because I could use it to increase profit

I'd be more than happy to enter in a 'good' trade provided i expected it to be the best allocation for my stake at that time (i.e. no 'great' trades available). If it later shows the signs of being a 'great' trade then you should add to your position.

Which basically leads to my philosophy on pyramiding; your position size should be increased/decreased in line with increases/decreases in expectations of reward vs risk.

The difference in thinking seems to be that you put an expectation on a trade before you buy it, where I do not. I'm happy to enter a good trade also, but all trades are good trades before I buy them. They give me an average return above 0%.

I have great trades too, but only once I've sold them for a nice profit.

You are happy to add to a trade once it shows signs of becoming a great trade. Would it show signs of becoming great trade if you didn't already hold the stock. Are you assessing the pyradmid entry as a new position?

I think trade frequency could be a reason for different opinions. I can understand that if you have free cash in a good market and no buy signals, then it makes sense to put funds to use if you're happy enough with your own analysis. I have a slight excess of buy signals so rarely have spare funds unless the market is dodgy.

Also, when you pyramid, do you plan ahead to pyramid and take a small initial stake, or do you take a full position initially and then add to that?

It makes more sense to me to take a full position where you know you'll be most profitable (your normal entry point). Then if you pyramid reduce the risk with a tighter stop.

 

[tech/a]

Different strokes--folks.
I'm not arguing that my method is best just that its a method I have used for years.
Served me well.
The OP asked for peoples ideas and we have had a couple.

 

[markhocky]

The difference in thinking seems to be that you put an expectation on a trade before you buy it, where I do not. I'm happy to enter a good trade also, but all trades are good trades before I buy them. They give me an average return above 0%.

Sorry Synergy. Don't think i explained myself very well in my last post. When i say expectation i mean having an average return above zero (not profit target or similar).

I measure this expected return throughout the trade, so if it increases at some point during the trade (i.e. my expected average return is higher), i will increase the position size, and vice versa. The position size is determined by these expectations, and is not a function of what the current position size is, or whether the trade is currently active.

This has to be balanced against the negatives of additional transaction and slippage costs though, which has made it difficult for me to find a workable method in practice.

Anyway, this is just my thinking on the subject. Not saying it's the best approach, just providing an alternate view.

 

[peter2]

Pyramiding has an unexpected psychological benefit that is worth mentioning. Once a trade is started and progressing well less experienced traders are probably thinking about the exit and realising their profit. If your trading plans include pyramiding then you will be waiting for a setup to pyramid rather than a reason to exit. This is an important psychological benefit. Pyramiding keeps you in a positive mindset about the trade. It is very easy to drift into a negative mindset and think about the exit which may then be premature.

Letting the winners get bigger and increasing your average win is the second hardest skill to develop. Learning to take your losses and lower your average loser is the hardest. I would encourage less experienced traders to include conservative pyramiding strategies into their trading plans to help develop a positive mindset while the trade is open.

 

[pavilion103]

I came across an interesting example below. I am not sure if this is the best way to do it however.
This one says to pyramid every 0.5R price increment and trail the stop accordingly (i.e. stop moves 1.5R from the initial stop also).



I would be interested to see if people agree with my thoughts
1. It would be much better to use a technical entry point rather than a predetermined 0.5R increase.
2. It would also be much better to have a stop that makes sense rather than an arbitrary figure.

Another important question: Is it generally better for the size of trades to be smaller as each pyramid position is taken? or can each position size be the same?

Also when the position is pyramided, is the stop for the total position generally the same (i.e. does the stop of previous pyramid positions get moved to the stop of the last pyramid position)?

 

[skc]

I came across an interesting example below. I am not sure if this is the best way to do it however.
This one says to pyramid every 0.5R price increment and trail the stop accordingly (i.e. stop moves 1.5R from the initial stop also).

View attachment 42981

I would be interested to see if people agree with my thoughts
1. It would be much better to use a technical entry point rather than a predetermined 0.5R increase.
2. It would also be much better to have a stop that makes sense rather than an arbitrary figure.

Another important question: Is it generally better for the size of trades to be smaller as each pyramid position is taken? or can each position size be the same?

Also when the position is pyramided, is the stop for the total position generally the same (i.e. does the stop of previous pyramid positions get moved to the stop of the last pyramid position)?

There are many ways to pyramid but the over arching risks need to be in check. E.g.

- You can stagger your stops but the risk should be 2% or whatever figure you have if all are stopped out.
- The overall position size should be kept in check to manage gapping risk. You wouldn't want to be pyramiding MEO 5 times when they drill a duster...

 

[pavilion103]

There are many ways to pyramid but the over arching risks need to be in check. E.g.

- You can stagger your stops but the risk should be 2% or whatever figure you have if all are stopped out.
- The overall position size should be kept in check to manage gapping risk. You wouldn't want to be pyramiding MEO 5 times when they drill a duster...

So the whole idea of pyramiding is for greater potential gains, with the same risk as a single trade (e.g. 2%)?

I guess this isn't the market to be attempting this strategy anyway.

 

[pavilion103]

And is it best to wait untilt I am at least in a break-even position before I pyramid?

Do most people then enter as they would a separate trade? (but maybe trade a smaller position size?)

 

[Wysiwyg]

Another important question: Is it generally better for the size of trades to be smaller as each pyramid position is taken? or can each position size be the same?

Makes me think this example is hypothetical to exaggerate the gains that can be made.
The quantities remain the same with the initial parcel of shares cost $9799 and the last parcel of shares cost $15813. If the stop loss was tightened to compensate then a stop out becomes increasingly likely.

As I think, retrospective, hindsight and hypothetical shows what could have been, success wise.