To SCALE or Not to SCALE a profit?

[jem]

I am not sure what you consider macro and micro. I guess we agree that you can look at scaling as sort of a 2 part trade and decide to look at separate risk reward setups for each part of your trade as suggested by the other poster. But I think you are saying that you must look at your trades as a group and consequently taking 10 cents now on this trade just because you have it could be stupid. If that is what I meant I would agree with you but that is not what I meant.


I think you may have misinterpreted what I said. (It was not perfectly written)

I was saying saying take a look at a trade before it begins. I have certain setups that I like. Now they may be worth trying to pull 50 cents out of, but if,as I am about to make a trade a program rolls on-- I buy double and throw out the first part as sort of a very very low risk trade. Also, even with no programs under certain circumstances I expect to have almost instant profits but I am not really sureif the trade is any more likely to follow through to my target. While not double up and flip out some and get positive slippage.

You see it is all about your rs and not about scaling. It is all how you see "it". I thought the people arguing that scaling is not smart because it throws off Rs by accepting less than full profit but assuming full risk--- were not taking into effect that scaling reduces risk. Trading is all about managing your trade. And I believe that when you tally up the factors you scale. Also in reality the R concept is a bit of a deception and I will get into that in a future post. Lets just say that I know you commisso as the lover of all things zen (just an exaggeration) realize that the concept of risk on a trade is the same as pointing at the moon.

So lets us look at the pros of scaling.

Not all setups are equal.
Scouts
smoothness of equity curve
psychological benefits of having less losers.
Profiting from short term setups that boost your trades profitability like programs and selling into momentum.

Ability to step up to larger size faster
availability and turnover of capital for other trades and premium setups

Less big losers

Trading is really dynamic and while you are holding for your target it may actually be advisable to put more on or take some off relative to the true risk reward atmosphere of the market. In the end you do your tally and make sure on the whole you are making more than you are risking. But to argue against scaling is to misunderstand what is really going on when you are trading. I am not saying that you are currently arguing against scaling commisso). If you like to be rigid go for it but it is not better just different.

 

[nitro]

Originally posted by AAAintheBeltway


What about exiting winners?

I have not modeled exiting winners, but I can tell you this with conviction based on experience: In NEARLY every case that I have tried to scale out of a winner [usually for the lame psychological that I want to see my P/L go up and I want to "ring the register" for fear of losing] I have always regretted it - all it does is get you out of a good position too soon. The only thing that should determine your exit is stop loss, a coil [if you are good enough to put the position back on without hesitation] or a clear change in trend/momentum [ Some will tell you that if a stock's daily range is x% and the stock has already moved that much, take action - I am not opposed to that, but I have not modeled that either.]

People invent these things so that they can get instant gratification to their psychology of being right. But I can tell you with no uncertain terms - There are only two ways to make a living trading:

1) You got to do size
2) If you do not do size, then you have to let trades work
3) [I said two, but the above two are axiomatic, this is an obvious theorem] IMHO, the best traders do both 1 and 2, even if it means that you are wrong more times than right, as long as the rights are much bigger than the wrongs and you have strict stops when you are wrong.

If you only do 1) without 2), then you have to trade frequently, i.e., you are probably scalping, in which case you better be more right than wrong.

Doing 1) is, IMHO, a lot easier than 2) [for me - that is probably why I gravitate towards a scalping style currently] - this is also probably axiomatic - the greater the size of the trade, the more likely we are to jeopardize "letting the trade work."

nitro

 

[Pabst]

Originally posted by nitro

I have not modeled exiting winners, but I can tell you this with conviction based on experience: In NEARLY every case that I have tried to scale out of a winner [usually for the lame psychological that I want to see my P/L go up and I want to "ring the register" for fear of losing] I have always regretted it - all it does is get you out of a good position too soon. The only thing that should determine your exit is stop loss, a coil [if you are good enough to put the position back on without hesitation] or a clear change in trend/momentum [ Some will tell you that if a stock's daily range is x% and the stock has already moved that much, take action - I am not opposed to that, but I have not modeled that either.]

People invent these things so that they can get instant gratification to their psychology of being right. But I can tell you with no uncertain terms - There are only two ways to make a living trading:

1) You got to do size
2) If you do not do size, then you have to let trades work
3) [I said two, but the above two are axiomatic, this is an obvious theorem] IMHO, the best traders do both 1 and 2, even if it means that you are wrong more times than right, as long as the rights are much bigger than the wrongs and you have strict stops when you are wrong.

If you only do 1) without 2), then you have to trade frequently, i.e., you are probably scalping, in which case you better be more right than wrong.

Doing 1) is, IMHO, a lot easier than 2) [for me - that is probably why I gravitate towards a scalping style currently] - this is also probably axiomatic - the greater the size of the trade, the more likely we are to jeopardize "letting the trade work."

nitro

Great post Nitro. I am firmly with you and Commisso on this issue. I spent most of my career being a "true" scalper, i.e. a local who bought the bid and sold the offer. Adding to losers and peeling out of winners was my schtick. Very rewarding in the micro. Obviously low commissions and a tremendous positive expectancy are paramount for successfully employing such a strategy. However few traders are so skilled as to lay claim for any expectancy high enough to justify an even R/R. In the macro those "consistent, steady returns" that every novice yearns for are death waiting to happen.
True story. A guy was relating to me how he started way back in the MidAm with Rich Dennis. They would each enter a trade off the same setup. As the fellow I met would be getting out with a small profit, Dennis would be doubling up. I remember the phrase he used, "Rich would push the pedal to the metal on every trade." Of course RD was in his formative stages on his way to 200mil. Later in the evening a friend who was with me, said that the guy who had told us the story about Dennis had also been a pretty good trader. But on that night, many years after his prime, the fellow with the good tale, was toiling behind the bar pouring drinks.

 

[jem]

I can point to analogies to support my side too. In fact Richard Dennis closed a fund in which I was invested. You could say he put the pedal to the metal and blew it up. After a decent run he lost 50% of the fund rather quickly. (a few months). I was told he had so many redemptions he had to close the fund. Luckily I scaled out of half of the account when he was profitable to buy a house so I wound up being net positive with him. That is the absolute honest truth. I invested through ( I cut this because I do not want to adverstise this risky stuff) the money was syndicated by a firm in Greenwich and placed into the Dennis fund.

But that is besides the point. I agree I am not going to get rich scalping and I am stretching out my trades. The question is does scaling out necesarily change your expectancy and is it foolish to give into the psychological temptation?

 

[Pabst]

Jem, my take is that nothing is easier than scaling out of a nice trade in the midst of a big trend day. Feels great. You make a few bucks, pat yourself on the back and say hey, I wasn't a pig, I took something off the table every step of the way, and still caught most of the move. The hard thing is saying to yourself this is a heaven sent opportunity to pour it on. I'm risking everything I'm up, and doubling down. If this puppy is setup the way I think it is I'll unload MOC, because about 5 days a month it closes on some fuck you extreme high or low of the day, taking no prisoners. Granted a very hard thing to do, and I'll admit, maybe once a quarter I have the gumption to trade like that. But I've traded every style there is, and I am completely without doubt that in the macro doing the "hard thing", whether in trading or life, produces the most fruits of success.

 

[jem]

This seems to be another one of those threads where we will wind up agreeing. I would like nothing more on trend days to then to double up hold half all day and double up on a pull back and then hold that new half with my old half. I have even gone so far as to have learned how to identify those days. Instead of not making money on those days I am now making some. I expect to go big soon. However, my only real point that I was making was this.

When the trade starts you have 1r risk and say you look to 4r gain. Now if you scale out on the way to your profit objective with half your postion you will only get say a total of 3r. This seems to most people to mean that you still risk 1r and now you get less on your winners so you have screwed up you balance.

I was only pointing out that since you took half off for profit you now are in a 1/2 risk position. (A version of marginal this equals marginal that) So by scaling you are not screwing up your Rs. Then the smart poster pointed out you should look at a scale as two seperate trades. I find that logic to be perfect. That was the whole point of my post.

Now I also explained why scouts are good by I really do not care to belabor the point. And commisso's point about scaling in is being studied, Although I think you can also argue it is just a new trade. But, I see what he is getting at. In the end it is my thesis that it is all a great big R mix. And in reality, since no one really knows what his true risk is and the market may not function as it has in the past; I believe the whole R discussion is just representative of what is going on but it is not reality. Most people can not make money because there is another component to trading that is sort of left out of the R equation. But at this point in time my kids are a sleep and it is a perfect time to test out my concepts of how to make more money with less work in tradestation.

 

[vladiator]

Originally posted by bone
Why would you risk 20% of your account on one trade, Mr. one-lot wonder? Is that the way to go? Do you walk up to a roulette table and put your entire stake on red or black?
</b>

Since you have no idea of what my strategy is about, I suggest you keep your smirks of critiques to yourself. I take 100% (if not more) exposure positions overnight and close them in the morning. It IS the way to go for me. And it has nothing to do with roulette. Stop loss at 5%? It's IRRELEVANT in my case b/c it can open much lower than that. And NO, I would not have made 25% instead of ten. B/c the stop would have just filled me in at the very lowest tick.

I'm an Engineer by training, so I understand statistics way better than you.

So what? Am I supposed to be impressed or what? I have a PhD in finance (with a cognate in stat) and don't think my advice is idiotic. Yours seems to be though. Way better? Not sure about that... Especially given your statements.

I am offended by jags like you giving such idiotic advice. You're no different than Jack Grubman giving advice about WorldCom. Nobody risks 20% of their account on a trade and stays around for very long. I am a professional trader - I do this for a living. I support a family of five... big house, 5 cars, vacation home, private tuition for three kids. I do it by making smart decisions, minimizing my risk, and trading size.

I'm glad it works out for your strategy. It will not for mine. And you have no basis to ascertain that what you do is correct just b/c it seems to have worked out OK so far. I'm up VERY handsomely since I started out, thank God, and have no intention of changing anything. Especially when it comes to rushing to take losses. In your case you may be advised to. In my case, if it does seem to go against me, it most cases it is only a temporarty fluke.

Your problem is that you're right and the market is wrong. You advocate huge capital drawdowns and promote it to others. Criminal on your part, AMATEUR.

Not at all. And you did get it right. I AM 100% contrarian. When I take positions, I go completely against the prevailing movement. If that means that I assume the others are wrong, you can interpret it the way you want. It yields me about 1-2% a day and I'm very content with that (in some cases the market IS wrong. You just have to know when and have the balls to act on it.)
You are the one who started your arrogant and belittling remarks. And correcting my spelling was a very cheap shot. English is my 4th language. I'd love to see you spell in your 4th language, dude. Besides, I was in a hurry.

 

[vladiator]

OK, just to set the record straight. In what I do, I identify a set of stocks to be held overnight. I buy them pre close and sell next day after the open. That is done completely independent of what happens overnight and is remarkably consistent. By averaging out the prices on both ends, I dramatically reduce the volatility and get a much tighter distribution of returns. The buys and sells are always done at the same predetermined times and all the other factors you enumerate there (like "if it doesn't go my way after the first lot, I get out") are irrelevant. If I get the first 1000 shares and it keeps going up I get more, if it goes down, I still get more. Thus, I get the average price I'm shooting for over the period I need to get long over. I do the same on exit. B/c the rule is based 100% on sound theory (and not mindless data snooping), I tend to get on average what I expect to be getting. Had I just gotten into and out of the positions in one single trade and at a radom point within those time intervals, statistically it would not make any difference given an infinite number of iterations and potentially unlimited resources (b/c the expected fill prices would be equal to averages over those time frames). By gradually getting into/out of positions, I make sure I remain close to those averages in any given case and reduce the likelihood of outliers.
If your engineering-trained brained doesn't understand that, then you wasted your tuition money.:eek:

 

[nitro]

Originally posted by vladiator
OK, just to set the record straight. In what I do, I identify a set of stocks to be held overnight. I buy them pre close and sell next day after the open. That is done completely independent of what happens overnight and is remarkably consistent. By averaging out the prices on both ends, I dramatically reduce the volatility and get a much tighter distribution of returns. The buys and sells are always done at the same predetermined times and all the other factors you enumerate there (like "if it doesn't go my way after the first lot, I get out") are irrelevant. If I get the first 1000 shares and it keeps going up I get more, if it goes down, I still get more. Thus, I get the average price I'm shooting for over the period I need to get long over. I do the same on exit. B/c the rule is based 100% on sound theory (and not mindless data snooping), I tend to get on average what I expect to be getting. Had I just gotten into and out of the positions in one single trade and at a radom point within those time intervals, statistically it would not make any difference given an infinite number of iterations and potentially unlimited resources (b/c the expected fill prices would be equal to averages over those time frames). By gradually getting into/out of positions, I make sure I remain close to those averages in any given case and reduce the likelihood of outliers.
If your engineering-trained brained doesn't understand that, then you wasted your tuition money.:eek:

Vlad,

This is a bizarre "system." I understand what you are saying - somehow it feels like there is a fundamental mathematical error in your analysis...However, given your credentials, it hardly seems likely...

Let me think about it....

BTW, is this an NYSE or NASDAQ or both system?


nitro

 

[Trend Fader]

OK... good job guys.. we got some good insights into scaling vs non scaling. The most important thing I realized was the consideration of scaling as seperate trades in itself.

But this is all opinion. No hard facts.

Does anyone here have any statistical evidence they can share.. in regards to scaling vs non scaling?

Can someone backtest a bunch of different trading systems using various strategies and we can compare the results? THe most important thing I would look for is not the % return or risk to reward.. but the slope of equity curve and drawdowns.

I would be the first to do this.. but I always fudge up my backtests so I would not deem my results as credible.


--MIKE