anti-martingale by Lauriston Livermore

as i win fewer than 3 out of 10 trades,i need to have a bigger position when i win than when i lose or i am never going to make any money

the anti-martingale by Lauriston Livermore

Lets us suppose,for example,that i am buying some stock.
ill buy two thousand shares at 110.If the stock goes up to 111 after i buy it i am,at least temporaily,right in my operation,because it is a point higher
it showsme a profit. Well,because i am right i go in and buy another two thousand share.If the market is still rising i buy a third lot of two thousand share.

Anything Anti-Martingale immediately grabs my attention.

If I hear you correctly ... we should increase lots when we are winning .. and not when loosing???? (aka Martingale)

We are ALL EARS ...

ECFOREX

Say the price goes to 114.i think it is enough for the time being.I now have a trading basis to work from.i am long six thousand shares at an average of 111 and 3/4,and the stock is selling at 114.I wont buy any more just then.I wait and see.I figure at some stage of the rise there is going to be a reaction

seems straight forward,invest the profits from the original position,to treble it,wait for the pullback and double up,

your position is now six times as big,

You will get killed on reversals.

Do you mean position building?

Yes it is true that reversals will kill you alot faster but with proper management you can build big-huge pyramid.

Position building within a trend.. look into it

With anti-gale, you will increase your profits when markets trend, and (as Finally points out) increase your losses when the markets move sideways. So you'd need to conduct testing to ascertain which market condition outweighs the other, within the context of your entry/exit strategy.

MM alone can not provide an edge.

I think Hanover mentioned before that the best way is to risk a fixed percentage of your bankroll. This way you ride out your winning steaks and minimize your losses when you hit a rough patch. Even that to me is still a little risky and can lead to big swings in capital.

In Technical Analysis Fallacy, fti's approach to position sizing, what they call position snowballing or skews is worth a read.

That's not true, depends how you pyramid the add ons and when you close them out

Phorex Phreak

Adding positions are vital.

My advice, please read into 'strat long term stress free trading' thread, just page 4.

Building a huge equity train within a established trend. I believe this is what you are looking for

Is this a serious post? Increasing or decreasing your stake based on whether the last trade won or lost won't help you. Pyramiding won't help if your trades aren't already profitable.

Scaling in or out in itself can't provide an edge. That's pretty obvious if we think in terms of the component trades, individually; each component ultimately contributes to the eventual bottom line, in its own right. Mathematically, if any one of the components is -EV, then it will reduce overall P/L.

To benefit from pyramiding (adding component positions as a trend emerges), it's necessary to prove that EV has increased significantly enough to justify adding the component at that point. Otherwise there's no basis for not including the component as part of the original position, i.e. to exploit the 'better' price. Similar mathematical logic to that posted in Craig's link here is required to prove (or disprove) such an assumption.

yes, but what do you mean by negative component? all your doing is adding components or units to the position as it goes in your favour at different intervals, i bet you most of the systems on this forum have statistacally relatively the same win percentage but the results are radically different because of the position sizing not including psychological factors of course....

i also bet you if i used my position sizing on any of the half decent systems on this forum in the end i would still make money, my current system is only 45% win accurate but its my position sizing( and psychology) that lets me stay in the game during my losing streaks and its position sizing that allows me to win big when i catch those trends, my wins far outweigh my little losses at the end of the month, and i believe thats what trading is all about, so in reality, i have more losers than winners at the end of the month but my winners are far greater than the losers...the difference in the way you use position sizing has a radical effect on your results....

starting off with full positions is krazy in my opinion, you should add to these positions once the trade has proven itself and as soon as a loser is identified, chop it off immediately, and of course the loser will be only a quarter position so your really not losing that much so therefore you can take many losses and not really feel it...the faster traders realise this the faster they will start seeing success in there trading....

Phorex Phreak

For Tiago-

Your position sizing has nothing to do with your edge.

Flat bet it, taking what becomes your full position during the times that you scale in, and obviously adjust the price to the average price of the scaled position... see if you do better or worse.

thats where you and many traders are wrong, the position sizing is the edge....flat bet it? whats the point in that? of course the results would be different....what we are talking about here is or you talking about here is a small piece of the pie my friend, which is the so-called edge of a system...a systems performance or edge or whatever you want to call it has a few components to it and position sizing along with psychology is by far about 90% of the pie my friend...there are many systems on this forum that have negative edges but because some traders know how to position size they still make money

Phorex Phreak

I was hoping that you would say that your profits, from my proposal of flat betting, would turn out to be the same.

I respect your view, though I don't agree. That's what makes our community great. Hopefully, I will gain some insight from this thread.

Tiago,

The following refers to the diagram below.

If price takes either path "1" or path "2", Trader "X" will be 20 pips better off than Trader "Y", and 2x20 pips better off than Trader "Z". Provided that all 3 traders ultimately exit at the same point, then that will always be the case, regardless of where the exit point is (in profit or in loss). In both of these scenarios, and everything else being equal, X (who entered his entire position at A) will outperform Y (who scaled in).

But everything is not necessarily equal; we must take the probabilities into account. If upon reaching level B, the probability of outcome "1" is significantly greater than outcome "2", than it was at A, then "Z", who has weighted his position more heavily at "B" will fare best, and "X" worst. No matter what the outcome, Y's performance will always lie somewhere between X's and Z's, because he has hedged his bets. Scaling in is a form of diversification.

However, if price takes path "3", then it depends on the exit point. If the exit point is higher than A, then X will outperform Y, and both will outperform Z, who never managed to enter. However, if the exit point is below A, then Z comes off best (no loss), while Y's loss is only half of X's. And so on.

Of course we can't calculate the probabilities like we could with, for example, Poker or Roulette. But if we tested the idea over a very large number of trades, we'd get a rough idea. That's what I meant by "prove or disprove". Hence I'm not saying that pyramiding can't work, just that you'd need to run appropriate tests (within the context of your entry/exit rules). If you've already proved that to yourself, then great! Nobody can argue against profit and success.

Of course the scenario is an oversimplification. We could have additional entry points (C, D, …), and apply different methods of exit (e.g. trailing SL). But no matter how complex the scenario, the same principles apply in much the same way.

About position sizing. If you put higher bets on higher probability outcomes, you will win more. But simply putting higher bets on without taking probability into account will result in larger wins, and larger losses, in the same proportion. If the probability of trend continuation (outcome "1") is sufficiently greater at B than it is at A, then you'd be justified in buying more lots at B than at A. But that's only because the probabilities justify it; the position sizing by itself doesn't create an edge...... but, used judiciously, it can magnify (or diminish) the dollar value of an existing edge.

Probabilities would likely vary, depending on the location of A and B, i.e. if they were at key price levels (S/R, pivots, Fibo retracements, round numbers, etc). But again, that's related to entry methodology, rather than position size.

Likewise, psychology by itself can't create an edge. An EA will trade with perfect discipline, but simply running an EA doesn't guarantee that you will be profitable. In fact the opposite is true; I posted on this topic here. If simply pyramiding into every emerging trend created an obvious edge, then everybody would use such an EA, and become a forex millionaire. But apparently this doesn't happen.

Although there are times, where adding to a position which is in profit already could be advantageous, like when really nice powerful trend moves form then you can add another position and set a SL at over all break even 1/2 way between the 2 ofcourse if you want to.

most of the time, if you buy 20pips higher it'll just chop back against you, hitting your BE Stop ( If you took that route ), 10 pips lower or drop into a over all negative position.

I prefer averaging down, the Trend is up, you've had a pull back so add a Long and if it pulls back further but the trend is still up add another, exiting when the trend is no longer up.

Obviously the case #1 above where the market doesn't stop going is the downside, but there rare if you consider every up / down move on the chart.


Both can have the downside of over leveraging your account and a sudden move causing huge losses.

Thanks for that, that's just great. That's sorted that one out then.

hanover,

i respect you bro, but i dont agree with you, i wish i can show you the way i do things and what i've come up with but im not at liberty to share it here on a public forum....if your interested pop me a pm and i can go over it with you....cheers

Phorex Phreak