To start with, I am a great admirer of POP. All that he said just make sense to me and this strategy of all in/scale out is based on one of POP's sayings: "Let the market tell you your position is proven correct, but never let the market tell you that your position is wrong."
So here is how I implement it when I take any position:
1. From my past trade records, I had noticed that whenever I won, the wins were quick and price would move in my favour in not more then 5min or say 2-3 bars(I trade based on 3min charts).
2. Also, as time passes by and price does not move in my favour, the probability of me being stopped out keeps on increasing.
3. Now whenever I get my valid trade setup, I enter with my max allowable position size on that setup. Now I presume that I am on wrong side and keep scaling out until I am proven correct on that particular position.
4. There are times when I don't even get time to scale out of my position as then market quickly moves in my favour. Also if this does not happen, I keep scaling out if price is below my entry price and time is ticking by as per my plan. Also a thing to note that, I don't wait for my SL to get hit as I am already out of my position well before my SL being hit(never let the market tell you that your position is wrong.)
So, this is what helps me to win big when I win, and lose small when is lose.
This is more of a intuitive part and it works most of the time for me.
Hope this was of some help to you.
The following 3 users say Thank You to Phantom for this post:
Thats exactly what i was thinking , i noticed that scaling-out when it goes wrong hasn't been discussed properly , most of the discussion is about scaling-out when taking profits , what about scaling-out when taking losses as well ? Example : Lets say i trade 3 lots Dax , and my first SL at 10 points , second SL at 20 and 3rd SL at 30 points away .
The following user says Thank You to Surprise for this post:
if I have a scale in methodology and add to my positions in thirds how would you manage the trade after the first two scales are made? Additionally, where would you begin to take profit after you are scaled in at full size? Look forward to hearing some responses....
The following user says Thank You to mabr0408 for this post:
I trade longer term, but the following scenario applies to all types of trading.
A stock gaps past my intended buy point. Not very happy with that, but since I don't know whether this will be a major runner, I buy some stock. Then during the day, as it descends, I buy some more. As long as it holds the breakout point, I am comfortable adding. My stop-loss is still based below the breakout point and I don't adjust it downwards with any additional buys.
In the above example, I get a lower average cost and less risk to my account. End of the day, my probability of being stopped out is still the same as if I had put on the full trade immediately, except that I would lose slightly less if it does happen.
Depending on how you scale in could lead to bigger profits than going all in. Also, depending on how you scale out, you could also have bigger profits than just going all in. Although, I will admit that scaling in/out will most likely smooth the equity curve.
The following user says Thank You to grausch for this post:
Attached is @Big Mikes trade data (which he has given permission to repost here) that he previously posted to the forum.
Mike has a real edge, and an integral part of that is how he manages the trade by scaling in, averaging down, adding to and scaling out. He mixes it up based on his current read of the market.
In using his entry and exit points I've made some basic assumptions so I can test out the different entry / exit methods.
The assumptions are:-
- He opened a trade when he thinks he has an opportunity to make money
- He fully closed the trade when he thinks he no longer has an opportunity to make money
That last assumption is a big one. In reality Mike scales out at structural targets and when he perceives his odds of further profit are waning. Perhaps that invalidates the results. Perhaps not. I think it's safe to say he doesn't leave a trade open when the conditions on which he opened the trade are now invalid. And if they are still valid he has the opportunity to make more money.
In any case, the goal was to determine what effects entry/exist & trade management has on the outcome (based on the 2 assumptions above) - and if so, what is it.
The s/sheet is set up with various adjustable parameters - so its possible to adjust the opening size, how many points before you add, scale out or average down etc.
Regardless of starting AIAO size, # contracts added, initial scaling in size, or points in your favour before adding - both "AIAO" and "All in - Scale out" eat ass.
Averaging down - which most people curse as the devil is actually pretty darned good (unless I've ballsed up my formulas)
I ran through some Monte Carlo's using the parameters in the screen shot below to look at the distribution of returns. Te results are based on 200 trades with 5000 simulations. The more profitable methods had bigger distributions - but the downside wouldn't be enough to stop you trading those methods.
The following 15 users say Thank You to sixtyseven for this post:
I’d like to be persuaded to favor the scale in – scale out method as it’s promoted by many successful traders here. The study posted shows Ave Size for the various scale-in/out methods. It ranges from 8.9 to 11.5 contracts. The AIAO method uses 9 contracts to be comparable.
The problem I have is that the study doesn’t show the maximum number of contracts used for the scale-in/out methods. If you’re willing to tolerate a maximum of say, 15 contracts for scaling then your risk tolerance is 15 contracts… period. Shouldn’t the AIAO method use the maximum tolerated ?
This observation is far from academic for me. It’s the real # 1 reason I actually trade AIAO on a per “strategy” basis. I do “scale” by stacking multiple setups/strategies as opportunities for their setups present themselves. Wow, I think I answered the question for myself. My maximum is distributed across my trade setups.
Anyway, maybe the study should have the maximum listed anyway and show the results for AIAO just to complete the picture.
The following user says Thank You to timendaGain for this post:
I think maximum tolerated risk is a red herring, rather expectancy should be the focus. I've attached a quick spreadsheet to demonstrate why I think this.
2 different cases.
Case 1: Assumes a 'BE' stop if price goes your way, at which point you add.
Case 2: Leaving the stop where it was if price goes your way.
I've used a binomial tree to calculate the probabilities of various edge sizes and the resulting likely-hood of a full loss, full win, BE, or full loss if price first moves your way. I didn't bother running the tree all the way out - so the total probability of outcomes adds only to 99.6%
Adding - and exposing yourself to more potential risk - also exposes yourself to more potential gain. That's the nuts of it. And in reality if you possess an edge, and take on more risk it will pay off. If you don't have an edge - you will get smashed. The distribution of outcomes will be wider though, so you'll have some knarly equity swings both good and bad. But in the long run - if you possess a true edge - everything will be ok.
Obviously your total account size etc, needs to be considered when deciding on trade size, I'm not saying go crazy here. But I think it's wrong to look only at maximum total loss. Here's why....
In case 2, you occassionally get really smashed, and lose a lot more than what you do with AIAO. BUT it's only occasionally and that's the very important key - approx 15% of the time for a simple risk 5 get 10 situation. The outsized winners more than make up for it.
In case 1, if you move yourself to BE - you will have the same maximum loss - even WITH more contracts. BUT, your win % goes down, which on the surface may be troubling. This is because you will lose when price comes back to your entry point (which I've called breakeven - for the purposes of doing a direct comparison with AIAO)
I've included commissions of $5 RT.
I also added in a comparison of scaling in, and adding. So that the maximum total loss is the same is AIAO. That doesn't look pretty because you are exposing yourself to much less total risk, and therefore much less total gain. In the scenario with the parameters I previously posted the Scale-in & add method worked out quite ok - but that was because you took on more average contracts - which meant more exposure to bigger occasional losses. In this case - to keep maximum total loss at a certain amount means the majority of your losses are small, and so are your winners. With only that occasional full size loss you were happy taking with the AIAO.
Last edited by sixtyseven; July 2nd, 2015 at 05:39 AM.
Reason: Added in commissions to calculations
The following 2 users say Thank You to sixtyseven for this post: