Scaling In and/or Out

[brownsfan019 17]

In my ever pursuit of maximizing gains and minimizing losses (if possible) I'm wondering if anyone is currently trading with a scaling in and/or out methodology. As I've mentioned previously, I like to keep things simple (one entry, one exit) but I had a little idea today that if I scaled into my entries, perhaps losses could be kept smaller.

 

I suppose it really comes down to the exact entry method you use, but for me, I typically enter with Buy Stop Limits or Sell Stop Limits. By using this type of entry, I'm wondering if I could keep losses smaller in cases where my full contract lot does not get filled.

 

Let me explain...

 

In this chart I am assuming that we have a long entry and put 1/3 of contracts at 3 levels (just one tick apart) on the ES:

 

 

So in order for the trade to fully 'confirm' 3 price levels would need to be taken out to the upside. Obviously if going long, you'd like to see price moving up. The problem with this is that your overall net fill is going to be the average of all 3 (mid-point level in this example) and you could have your full boat riding at one tick better. But in my mind, that one tick better for the winners could be offset by taking much smaller losses. Let me explain further:

 

In this chart I am assuming that my entry gets ticked in and 1/3 of my contracts get filled and then a stop loss occurs:

 

 

The reasoning behind this is that I have seen trades where with a buy/sell stop order that price can just peak it's head through enough to trigger my order and then retreat. Since I am mainly trading the ES, it's realistic to consider that it can hit one level of my entries and then retreat. So why be exposed for the full amount at one level... That's the idea currently.

 

The issue here is that on the winners you are going to wish that you got it all in at that one price level. But by spreading your orders in thirds over three levels, your net average is simply the middle price level. And assuming that you just go one tick above each previous tick on the ES, it doesn't appear to be a big sacrifice in consideration of the winners, but could result in considerably less in the losers.

 

I'm going to have to examine this some more but thought I'd open it up for some discussion here.

[brownsfan019 17]

OK, I figured out why some traders may not scale in - trying to do the math on how your trades play out at 12:30am is not easy. :o

 

I'll have to come up with an Excel sheet or something on how to run this.

 

Example:

• Short this AM with average price of 50.50 on the ES (short 3 levels as illustrated above)
  
• 1/3 of the contracts are taken out at 48.00 & reversed long here as well
  
• the other 2/3 to reverse long do not fill
  
• the new 1/3 at 48.00 stops out at 46.00
  
• meanwhile still have 2/3 short from 50.50
  
• initiate new short of 1/3 at 45.50
  
• cover and go long 1/3 at 45.50, 1/3 at 45.75, 1/3 at 46.00
  
• cover and short 1/3 at 45.00
  
• cover and go long 1/3 at 47.00
  
• meanwhile still long 2/3 with average at 45.75
  

 

Anyways... you see where this is going all day.

 

Wed was a great example for me to look at this as a good handful of my losing trades would have never filled at 3 levels. Quite a few only filled 1/3. What I am seeing is:

 

1) Winners guaranteed to fill completely.

2) Losers may only fill partially.

 

This accomplishes a few things:

 

1) Commissions are lowered

2) You may take some 'shake out' trades, but not on the full load. It's kind of like testing the water before diving in.

3) Winners can actually net quite abit more b/c the shake out trade only took out a portion of your contracts. So in my short example this AM, I was short at 50.50 with a full load. Had I exited fully on my next setup at 48.00, that's +2.5. Not bad. The catch is that at that entry where I covered my short, I also went long, which stopped out for -2.00. Now if I was fully out on my short (which I was) at +2.75 (not scaling in) and then getting fully stopped on the next trade for -2.00 your net is easily calculated at +.75 before 6 full round turns. See how the numbers can really change?

 

Off to bed here but for our math whizzes on the board, any help on how to easily calculate this would be great and much appreciated.

[walterw 10]

Nice Thread here Brown, its not easy to scale in as it is to scale out... very few people (retail traders )have this skill developed... I think one nice way to scale in is as you get pullbacks against you with your timing entry criteria still present... now ES its not easy as you dont have much range... so there are not too much fractions there to be able to scale without being so near to your stop... thats why I never been able to be good on ES... its a very professional traded future... cheers Walter.

[brownsfan019 17]

After some sleep and playing out some of yesterday's trades, one item jumped out:

 

If you are using your exits as a place to also initiate a new position (as I am) the math can get even more sketchy.

 

Example - if short a full 3/3 and you get a reversal signal and get ticked in at ONE level, you stay short, but only for 1/3 of your contracts.

 

Let's say you are short at 1520, 1519.75 and 1519.50. A full 1/3 at each level.

 

You then take a reversal long at 1518.00. Only one level gets filled, but since this is a reversal long, your order should be resting for 2/3 of your contracts (you would need to have your orders lined up for 2/3 at each level). So you would stay short, but for only 1/3, which kind of defeats the purpose.

 

From what I can see now, this seems like a very viable strategy if you do not flatten and reverse a position at a new entry level. Just depends on how you trade really.

 

Looks like it's back to the simple method here.

 

Just when I thought I might have found a way to complicate things. :o

[Nick1984 10]

How many contracts are you trading? I would scale in only if i were to trade a heap of contracts. This is how professionals build up their portfolio so as to attract as minimal attention as possible and vice versa when they sell it off.

 

Also on what time frame are you thinking of scaling in your buy orders? If your going in and out pretty quickly then is it worth all your trouble? Maybe if your swing trading then you can scale in your position on pull backs it can work but for day trading time frame seems like too much of a headache!

[brownsfan019 17]

How many contracts are you trading? I would scale in only if i were to trade a heap of contracts. This is how professionals build up their portfolio so as to attract as minimal attention as possible and vice versa when they sell it off.

 

Also on what time frame are you thinking of scaling in your buy orders? If your going in and out pretty quickly then is it worth all your trouble? Maybe if your swing trading then you can scale in your position on pull backs it can work but for day trading time frame seems like too much of a headache!

 

Nick - the headache aspect of this does not concern me if it betters my results. And that was the initial thought behind this. Again, I think it could warrant attention even if just trading a 2 lot. Since the majority of my trades are done with a reversal trade this not only is a headache, but the math behind it doesn't always work in my favor which was my goal to begin with.

[ItalianSharp 10]

I believe going all in at once is a bit unwise given the nature of the markets, which will almost always allow you to get in at a better price.

 

Let's say you want to go long 3 contracts in YM.

 

You get in with one contract at 13000. You keep the remaining 2 just in case you can get a better price within your stop loss range.

 

In this case, if you lose straightaway you only lose on 1 contract.

 

If you win straightaway, you'll win on 1 contract...yet it'll be a very easy and almost stress-free trade because it went your direction immediately. And I like that even if it's only a 1-contract position!

[brownsfan019 17]

I believe going all in at once is a bit unwise given the nature of the markets, which will almost always allow you to get in at a better price.

 

Let's say you want to go long 3 contracts in YM.

 

You get in with one contract at 13000. You keep the remaining 2 just in case you can get a better price within your stop loss range.

 

In this case, if you lose straightaway you only lose on 1 contract.

 

If you win straightaway, you'll win on 1 contract...yet it'll be a very easy and almost stress-free trade because it went your direction immediately. And I like that even if it's only a 1-contract position!

 

Good point IS, but here are some things to consider under that strategy:

 

1) You are hoping the position GOES AGAINST you, so you can get a better price, but not too much so your stop is not endangered... Seems counter-productive to me - you want to lose initially, but not too much.

 

2) On your winners where you do not enter more, it will be too easy to kick yourself in the butt thinking your profit could have been double or triple in your example.

 

3) You would need a 'generous' stop level in order for this to work. Since you want price to go against you so you can buy/sell more, but not 'too' much, your initial stop level needs to be at an area where you can enter more and not get quickly stopped. This can work, but I personally have stops of 4-5 ticks on average in the ES.

 

4) On trades that stop out, you could easily have a FULL LOAD of contracts getting stopped out and on winners might only be able to get 1/3 of your contracts into the trade... All of a sudden those winners need to be fairly large to cover the losses even if the losses are 'small' b/c you averaged down, but it was on a full boat. I personally think averaging down is a loser's game as illustrated here.

 

The idea could work depending on your trade setups and rules. I think your stop loss area would have to be larger (and too large for me personally) and your mentality has to accept a very counter-intuitive way of thinking.

[jperl 11]

Good point IS, but here are some things to consider under that strategy:

 

1) You are hoping the position GOES AGAINST you, so you can get a better price, but not too much so your stop is not endangered... Seems counter-productive to me - you want to lose initially, but not too much.

I think there is some confusion here about the difference between scaling in vs. averaging down. They are not the same thing. In averaging down, you load up the boat on your first entry. Let say you are comfortable trading a max of 8 contracts with a loss of 5 pts/contract or 40 pts. If the market moves against you, you then overload the boat above your 5 pt stop increasing your load to 16 contracts. If the market continues to move against you, you are now in deep doodoo. You are trading beyond your comfort range. You can't stand the possibility of taking a loss so you add more contracts and you are really sweating it. But of course the market continues its relentless pursuit against you. Eventually you lose a bundle. This is what averaging down is all about.

 

Now consider scaling in.

Let say you are normally comfortable trading no more than 8 contracts with a stop loss of 5 points, 40 pts total . At entry however, you only trade 4. Reason- you think you may have to scale in. The market moves against you 5 pts. Your down 20 pts., so you are not uncomfortable (remember your risk tolerance is 40 pts). You reconsider the trade. If it still has possibilities on the long side, you load the boat. Loading the boat means adding 4 more contracts. The market could drop another 2.5 pts. before you would be out 40 pts total (4x7.5 +4x 2.5=40). But interestingly enough, the market only has to revert 2.5 pts to the upside, for you to break even. So in reconsidering the trade at the 5 pt down entry, you have to decide whether the probability for the market going up 2.5 pts is greater than the probability for it going down 2.5 pts. If it is not, don't scale in. If it is, pull the trigger.

 

The difference then between scaling in vs. averaging downs has to do with planning and an understanding of your risk tolerance. Scaling in is not a helter skelter event which averaging down is. You should always be undertrading, in the event that scaling in or reversing becomes a necessity.

 

As an aside, every day trader should have scaling in as one of his or her strategies. And in addition, every day trader needs to learn how to reverse a trade when the situation warrants.

JERRY

[brownsfan019 17]

Jerry - good points. The initial topic was to discuss scaling in and it looks like Sharp brought up averaging down. I'm with you - I think averaging down is a loser's game over time.

 

But to say that everyone needs to scale in I think is incorrect, as I've demonstrated in this thread. You mention that traders need to be able to reverse, and we agree there, but to scale in and be able to reverse quickly is not one in the same. As I mentioned above, the numbers can get skewed quickly if you scale in AND reverse when necessary.

[Dogpile 10]

just thought I would add my current thinking:

 

I do like going in on 2 trades rather than 1 --- scaling in...

 

I always use 'buy-market'/'sell-market' on stops for ENTRY (sends a market order if it trades at my 'trigger price').... so this makes this clean when your order is sent during a buy or sell program that is executing. 'joining in' on this with a stop-market is a beautiful thing..... I am generally playing for a profit goal of just a few points...

 

so for example,

 

say I have staggered 2 different 'short-market' orders on NQ at 1960.00 and 1959.00... say they are each for 5 contracts each...

 

now the market fills the first one at 1960.00 -- I instantly put a 'buy to cover' limit order for 5 contracts at 1957.50. why?

 

because it will fill my other order at 1959.00 and give me 'exposure' of 10 contracts at that point. I will be averaged in at 59.50 and general scalping target is 3 points (1956.50)...

 

so my real risk is that the market fills all my orders and instantly reverses between 57.50 and 59.00...

 

if my first order is a bad trade, I am somewhat unlikely to fill on the 2nd part of the order without also filling the buytocover limit order given the small distance from the 2nd entry to the first exit point.

 

if my order is a good trade, I will fill on both and might make my first profit within a minute or two of the second entry... this fill reduces risk exponentially.

 

many times, I will take my second position and manage the exit on a discretionary basis. I might watch the first push and move down my stop to the top of the 3-minute bar that the entry triggered. good trades tend to work right away while bad trades tend to reverse right away --- so I have found this to be effective. your worst analysis of the market will tend to fill only on the first order.

 

one other scaling in method is simply using signals that are similar but on different contracts (ie YM)-- and use a similar style to exit... get out of the first entry relatively quickly after the second entry 'triggers'...

 

futures have so much leverage that it is important to control your risk tightly... I find that having multiple orders out (say 4 orders across 2 different contracts) is a really nice way to control risk when you begin trading bigger size and the dollar values get big... you only need a couple of points a day with size to make a living...

[BlowFish 10]

Hows this for a technique :- lets assume the Russel is in a reasonable tight range and you are anticipating a move down out of it.

 

Sell 2 units at the top of the range and cover 1 at the bottom lets say at +1.2 for arguments sake Move stop to -1. If price approaches the top of the range sell 2 more and rinse and repeat.

 

I have not used this technique for a while but it is possible to build a large multiple position that is essentially risk free (if you count the whole exercise as a single trade). long tight boxy ranges are good for this.

[malverd 10]

Jerry,

 

I know you mention that PVP is used as a stop loss point.

 

Let's assume a long position. Say your risk tolerance would allow you to place your exit point below the PVP.

 

If price closes below the PVP, but hasn't hit your risk tolerance stop AND the PVP point hasn't changed, should you stay in the trade?

[malverd 10]

Jerry,

 

I haven't seen you mention the following scenario:

 

For long:

 

The VWAP is above the PVP and lower SD1, SD2, etc. is also above the PVP.

 

I know you mention bounces off of the upper SD1 and VWAP are high probability trades.

 

But I watched you video on the symmetrical YM trade and obviously price was not above VWAP on that trade.

 

Would it be a high probability trade to take bounces off of the lower SD1 and SD2?. I realize that price is not above VWAP.

 

Also in that YM trade, you mentioned that you were using no stop loss. I assume that you were using a “stop” based on your full risk tolerance. If the trade had not worked out and the VWAP went below PVP, would you have exited the trade immediately?

 

Thanks

[Tradewinds 12]

Hows this for a technique :- lets assume the Russel is in a reasonable tight range and you are anticipating a move down out of it.

 

Sell 2 units at the top of the range and cover 1 at the bottom lets say at +1.2 for arguments sake Move stop to -1. If price approaches the top of the range sell 2 more and rinse and repeat.

 

I have not used this technique for a while but it is possible to build a large multiple position that is essentially risk free (if you count the whole exercise as a single trade). long tight boxy ranges are good for this.

 

So, the reason it is "Risk Free", is because you took profit, and even if subsequent orders went to a loss, the original profit covers the loss. The real issue is identifying the "long, tight, boxy range". There needs to be a way to anticipate and verify that price pattern. I think this strategy would be excellent for a sideways market with a downward bias.

 

No matter what strategy is being employed, there needs to be some underlying basis for the application.

[Tradewinds 12]

Two questions:

 

• What percentage of the time is the entry price being misjudged by an amount that would warrant the scaling in?
  
• Are you good at judging the direction the market will go in, but not be so good at getting the entry price right?

 

I'm wondering if the answer to those two questions determines whether averaging in might make any sense or not.

 

It might be possible to be good a judging the direction the market will go in, but not be so good at getting the entry price right.

 

My opinion is, that averaging in, is making the assumption, that the optimum entry point is often misjudged. If the optimum entry point was rarely misjudged, then averaging in would be counter productive. If you were real good at picking the entry price, and tried to average in, not all the orders would fill. Unless you intentionally picked bad entry prices. That wouldn't make any sense.

 

So my first question would be, "What percentage of the time is the entry price being misjudged by an amount that would warrant the scaling in?"

 

In post #4 you stated:

"Looks like it's back to the simple method here."

 

So I'm assuming that you gave up on this idea.

 

And in post #2 you brought up the point about how the exit orders fill, as opposed to the entries. It seems you were trying to both average in, AND scale out. If you averaged in, but then took profit on ALL the orders at the FIRST opportunity, that would eliminate the problem of having the order go against you with to many orders in the market.

 

But no matter what the strategy, it will always come down to how well we are reading the market and anticipating what it will do.

 

Again, the temptation is to average in because the entry is being misjudged. If the entry is being misjudged more than not, then averaging in might compound the mistake. It could make things even worse. I suppose it might be possible to be good at judging the direction the market will go in, but not be so good at getting the entry price right. That's not to bad a situation to be in.

 

If the entry is being anticipated well more often than not, then averaging in on the good entries either gives you worse prices than you could have gotten, or minimizes the size of the entry. Either way, potential is being hurt. So that doesn't make any sense.

 

If you are real good at judging what direction the market will go in, but not so great at picking a good entry price, then averaging in, would probably be good. Instead of 3 orders at 3 different levels, I would consider entering 2 orders, the smaller order at the least good entry, and the second, bigger order at a really good price. So 1 entry of 1 and a second entry of 2 contracts. Instead of 3 entries of 1 contract.

 

From experience, when I misjudge the entry price, it probably means that I misjudged what the trend was going to do. In that case, averaging in, just makes things even worse. So my gut feeling is, that even in the best case scenario, "averaging in" is tricky.

[BigAl 10]

For me I always trade in multiples of 2. Either 2, 4 6, 8 or 10 cars.

 

I take 1/2 of the position at my 1st target. 10 ticks for YM, 4 for ES, 12 for NQ and 5 TF.

 

Place the pstop with the ticks gained (from breakeven to risking 50%) from the 1st target and let the 2nd target trade. In fast moving or ranged bound markets, I'll play it fairly tight.

I average about 70% profitable, 20% breakeven and 10% non-profitable

 

Trade Well!

 

BigAl

[malverd 10]

What methods are you using for 70% profitable?

 

Thanks

For me I always trade in multiples of 2. Either 2, 4 6, 8 or 10 cars.

 

I take 1/2 of the position at my 1st target. 10 ticks for YM, 4 for ES, 12 for NQ and 5 TF.

 

Place the pstop with the ticks gained (from breakeven to risking 50%) from the 1st target and let the 2nd target trade. In fast moving or ranged bound markets, I'll play it fairly tight.

I average about 70% profitable, 20% breakeven and 10% non-profitable

 

Trade Well!

 

BigAl

[BigAl 10]

What methods are you using for 70% profitable?

 

Thanks

 

I trade the reversals of the morning time zones of 10:00, 10:30, 11:00, 11:30

12:00 and 12:30. I use volume spikes and 512 tick bars for the YM. The closer the spike is to the time zone the better the signal. Watch the market internals. The weekly open, overnight open/high/low, that day's open and previous day's high/low. Candlestick formations and patterns and Market Profile Value High/Value Low.

 

The key is getting 10 tick at the 1st target. Much higher percentage than trying for 15 or 20 ticks.

 

See the image below.

 

Trade Well,

 

BigAl

[AuctionMarket_Trader 10]

I scale out .....never scale in....

[ericmoles 10]

For what it is worth, I've heard that scaling in and scaling out can improve results, but that you don't want to do both in the same trade.

 

If you scale in, your avg. entry price is higher (assuming long for this example). When you scale out, your avg. exit price is lower. A higher entry and lower exit price means less $ won in your trade.

 

Testing would need to be done on a system to see if any of what I've just said/heard affects your results.

[RichardCox 10]

I just finished an article on this topic here:

 

http://www.traderslaboratory.com/forums/forex/18483-fibonacci-trade-scaling.html