How do you scale-in and why?

[fader]

Quote from RoughTrader:

In a very general sense, I feel scaling in does not provide much benefit.

Let's say you want to take on a position with 3 contracts. Without scaling in, the worst case loss is simply

Loss = 3 * Stop Loss Amount

However, as you are scaling in, your entries become farther into the position, and hence farther away from the stop loss. Your worst case stop can be much more than the above loss, assuming the profit target or stop-and-reverse point was never hit.

And yes, I understand that there will be times where the position will be stopped out with a favorable excursion less than the various scale-in locations, and these smaller losses will mitigate the effect of the larger losses accrued by a favorable position reversing itself into a loss.

This neutralizing effect becomes apparent when you backtest these types of entry techniques.

This is just what I have observed in my own research, others may find differing results.

RoughTrader

as i mentioned above, i am looking at a different form of scaling-in, entering against the current price movement.

however, this second mention of scaling into trend has prompted me to take a fresh look at the Turtle rules.

it's very interesting, according to the rules, the Turtles scaled in equal increments and had the same stop loss for each scale-in piece.

so, they kept a constant level of risk for each scaled entry, which is what i am assuming as a goal, i.e. i am not seeing a rationale (or i don't have one) to increase / decrease risk (or risk / return for that matter) during the process of scaling.

hence, what i am concluding is that the goal is to keep a target level or risk (and/or risk/return) and then assume a probability distribution and solve for the scale-in size which maintains the desired level of risk.

 

[fader]

Quote from jficquette:

Actually I feel that you should expect higher returns on the shares gotten that are the closest to the end price, assuming the end price is statistically significant. I mean the end price has to be way out there in odds of being hit. Say less then .1% of the time.

Let 1 equal the odds of getting hit. Find the additional move that will generate .5 the odds, then find the additional move that will generate .5 of that move. So 1x.5x.5. So small shares at 1, bigger shares at .5 and bigger shares at .25. The odds of being hit on the last shares need to be very low. Say .1% of so.


You find the increment to add at based on the moves that are half the prior move's probability. In other words don't use a method like add every dollar up etc.


John

you are considering only one type of probability, the probability of getting filled on a scale-in piece; i am also considering the probability of getting stopped out and the amount of stop loss which are all variable for each scale-in level.

one would have to assume a probability distribution and then model these variables, so far i have put together some preliminary scenarios but i am seeing only more risk from scaling-in at an increasing rate which seems to be what most people do;
i'll have to look at some more material, there must be a problem with my logic.

 

[Newc2]

Important thread.

If you have a problem in taking profits ie always taking profits too early, then scaling in makes a great positive difference to the end result. Not mathematically but psychologically.

By scaling in we are always looking to buy as opposed to sell.

 

[systematictrader]

i am working on adding a scaling-in component to my entries.

in my approach, i define a starting point and an ending point of scaling-in and a rate of scaling-in between these points.

based on an arithmetic solution applied to my trading strategy, it appears that an optimal rate of scaling-in is proportional - i.e. at the starting point, the position is 0% and at the ending point the position should be at 100%, so at 50% i should have one-half of the position on.

i have heard and read people say: "i put on half now and half X points higher/lower" or "i put on a small piece now and scale-in the rest at better prices later"... - while that is interesting, i am more interested to know why you scale-in the way you do, what is the rationale?

also, any recommendations for relevant books or other material on the subject would be appreciated.

u should also focus on scaling out, exit price is more important thats entry price, in matter fact u can even make money from random entries if ur exits are good enough

 

[W.J. Feathermaker III]

Interesting timing regarding the re-emergence of this thread. Similarly low levels of volatility in 2006 as now...a "drift up" kind of market, etc, etc...

 

[Overnight]

This thread wins the award for oldest-one revived I think. Last post before Sunday was about 10 years 9 months ago. Bravo Newc2!

 

[Buy1Sell2]

i am working on adding a scaling-in component to my entries.

This is a good strategy only when you don't currently have the money to invest ie like retirement investing/401k. Scaling is not a superior tactic when trading.

 

[murray t turtle]

Averaging down or up is a very good strategy as long as one follows this rule-- Never overextend. For example if I have a position go against my portfolio by .25% and I decide to add to a "Loser" , then that may well be the prudent thing to do. However, if the position is against me for 5% and I decide to average in, that's where it is a bad strategy. Just to say "Losers average Losers" is too much of a blanket statement and taken out of context means nothing. First one must define what a losing trade is. For me a trade only becomes a loser when I touch 2% of my account size. Until then the trade still has possibilities, especially when I am using decent directional analysis to enter the trade in the first place. Traders who average into trades that initially go against them by a bit, are the real winners in any market, not trend followers. We do follow the trend, but we have been in from much earlier on and better prices than trend followers.

As for scaling out---Never ever scale out.

%%
Good points, not that the Paul Tudor Jones quote is too much of a blanket statement.[ Losers average losers.]As you implied , use discretion, with any quote.I agree with much of your post, not that you would want to average down , @ the inflection ,when a bear changes in to a bull market, Buy1Sell 2.

My comments did not apply to any single stock, but large size mutual fund managers who get paid on the gross , average down all the time .LOL