Position sizing and adding to winners.

bitstream_ryder said:
1)Should one vary position size based on confidence of market direction.
Short answer is a resounding "No". Short unsophisticated answer is : Suppose you feel confident that the market is bullish but got stopped out of your long position due to bad entry for 1% of equity, so you double up for the next long entry. You get stopped out again and your loss is now 3% of equity. Do you double down again?

A series of losses WILL happen, you just don't know when and this could cause a major dent in your account, but that's a small problem. The bigger problem I find is how those escalating losses affect your trading psychology and performance in the subsequent trades (usually negative impact).
More...

That's why you have a daily stop limit to manage your risk. I'm not suggesting to double down to infinity. I'm not even suggesting to double down, but varying bet size based on estimated probability/confidence in the trade.

Further, you're assuming every entry has the same probability of success, yes? I believe that's not the case.

2)Should I add to winning positions
I have never successfully back-tested or live traded a strategy where adding to winning position was positive in the long run. I could write a short book on why this is a bad idea but I'd rather you back-test or forward test this to convince yourself.

With all this being, there are going to be trading situations and instruments where 1) and 2) will make sense but I have yet to come across a situation where this makes sense.

I suspect you don't keep a trading journal or if you do, it might not be detailed enough.

Hope this helps.
More...

Did you see the basic example I posted here?

Position sizing and adding to winners.

Theoretically, all in/all out will always be more profitable assuming you're actually buying the low and selling the high and can predict that outcome accurately in advance.

The idea of adding to winners, at least the way I see it, is that you increase your size as you're getting further confirmation that your anticipated scenario is taking place.

Of course - not all trades are suited for this. Ideally, you'd need a strong move for this.

If you're scalping for ticks - it's probably not suitable.
 
virtusa said:
I checked, starting with a smaller size and adding while profits rise brings me less money than going in full size at once.

I also know why. I wait till the trend is strong enough before entering.
More...

Yes.

Assuming a very good system or being able to predict a move from inception to end - going all in/out will always be more profitable.
 
Laissez Faire said:
That's why you have a daily stop limit to manage your risk. I'm not suggesting to double down to infinity. I'm not even suggesting to double down, but varying bet size based on estimated probability/confidence in the trade.

Further, you're assuming every entry has the same probability of success, yes? I believe that's not the case.



Did you see the basic example I posted here?

Position sizing and adding to winners.

Theoretically, all in/all out will always be more profitable assuming you're actually buying the low and selling the high and can predict that outcome accurately in advance.

The idea of adding to winners, at least the way I see it, is that you increase your size as you're getting further confirmation that your anticipated scenario is taking place.

Of course - not all trades are suited for this. Ideally, you'd need a strong move for this.

If you're scalping for ticks - it's probably not suitable.
More...

Thank you for taking the time to respond. I think it's only possible to evaluate the probability of success of a setup based on numerous observations. ie. if i trade a particular setup, my records show me that it will be successful 60% of the time. It would be rather difficult to predict outcome of each particular trade.

The idea of adding to winners as you are getting further confirmation is a very sexy idea and one which I have explored to some extent but have never got it to work either in back-testing or live trading. This is usually because the longer the trend has been in place, the greater the chance a squeeze, correction ect ..will take place and impact the outcome.

This method definitely works better before 2010(approx) and trend length seems to be getting shorter and less frequent in today's context. You can back-test the turtle trading strategy which i think is conceptually similar to yours to convince yourself.

"Theoretically, all in/all out will always be more profitable assuming you're actually buying the low and selling the high and can predict that outcome accurately in advance."

I only decide directional bias and try to catch a portion of a potential move, perhaps 30% to 50% of the daily range on a retrace going all in. Why? I can confidently say that there is a much higher probability that price will move within 50% of the daily range. I think it is impossible or perhaps I am not skilled enough to pick tops and bottomso_O

Again I have voluminous thoughts and data on this and the above descriptions is very brief. Anyway I encourage you to back/forward test your idea. Just saying that I have never got it to work after testing it somewhat extensively and thought it may save you some time in answering your question.

Good luck to you.
 
My most profitable trades are almost always ones I've scaled into.

Because it's all about law of large numbers, taking a lot of small positions with minimal upfront risk.

Think of it like dating...you go on lots of cheap first dates and scale into second third dates with a subset of all the first dates.

You don't go on first date at a fancy steakhouse and get married a week later lol. Unless you get lucky.

And trading isn't about luck, it's a numbers game
 
Well like I said, there may be situations can be profitable. For me personally, I' have yet to see evidence in terms of back-testing or verified trading records to support scaling in. I guess it depends on what your trading, the time frame, R:R objectives ect... If your testing and trading records supports it, then go for it. In short you have opinion on both sides. Which leads to the logical conclusion which is... Test the idea for yourself :D
 
Last edited:
Scaling in to winners for me is based entirely from price/volume action. I don't factor in probabilities, predictions, or confidence in the trade. I only care about having asymmetrical risk-to-reward, the reward is based on direct market feedback - the open profit.

I add on when I first have a substantial running profit & a healthy correction develops with constricting price/volume. My stop will get me out with a profit, it's just a matter of letting the profit run or hitting the stop - either way the position will be a win.

Leveraging profits to have the potential in making larger gains without increasing risk is the sweet spot in trading. When price/volume is making a strong trending move is the ideal condition to be most active in scaling into.

SGBX broke out on volume far above it's avg. The 4 cyan arrows are the trades scaled into.

upload_2020-4-13_14-59-53.png
 
Last edited:
comagnum said:
Scaling in to winners for me is based entirely from price/volume action. I don't factor in probabilities, predictions, or confidence in the trade. I only care about having asymmetrical risk-to-reward, the reward is based on direct market feedback - the open profit.

I add on when I first have a substantial running profit & a healthy correction develops with constricting price/volume. My stop will get me out with a profit, it's just a matter of letting the profit run or hitting the stop - either way the position will be a win.

Leveraging profits to have the potential in making larger gains without increasing risk is the sweet spot in trading. When price/volume is making a strong trending move is the ideal condition to be most active in scaling into.

SGBX broke out on volume far above it's avg. The 4 cyan arrows are the trades scaled into.

View attachment 224876
More...

This is the only way the math works. Everyone knows it. People used to regard pyramiding as a "well kept secret" on ET.... :vomit::vomit::vomit: So to have the runway to pyramid, all us independents have to be trend traders. And when you think about it, all we do is copy what the big guy, fat cats, "professionals" do. They manage huge amounts, find a trend, and scale into it. Us daytraders (really ought to daytrade entries, and swing trade exits), do the same thing on a shorter time frame. That said, most who "make it", are doing the same thing as everyone else. The irony of it all. Hence, "there are no secrets".
 
Hotcakes said:
This is the only way the math works. Everyone knows it. People used to regard pyramiding as a "well kept secret" on ET.... :vomit::vomit::vomit: So to have the runway to pyramid, all us independents have to be trend traders.
More...

This is different than pyramiding, I am managing each scale-in independently of the prior trades. Each trade has to have a lower risk entry point & has to be profitable or it will get cut - regardless of how much unrealized I have from prior scale-ins.

Your typical pyramid scheme is you add on when your up by a certain amount, & repeat - I think that is a good way to turn a winner into a loser .


upload_2020-4-13_19-20-22.png
 
You must log in or register to reply here.
Back
Top