Rethinking Scalping Concepts

[Kiwi 16]

daedelus,

 

I think the reasons for scaling out are mainly psychological:

- early because of the pain of a winner turning into a loser

- long holds to reduce the pain of missing out.

 

Mechanical traders rarely scale out because testing usually shows that one set of choices are good and the others are weak. The DIBs method (buying trend inside bar breaks) over at ff is a good example of this where the originator suggested 1/2 off at 1:1 and long hold on the other half - which gives people the sense of a free trade and thus ease of holding. But it turns out that the 1:1 with that setup has a negative expectancy so its main/real purpose was just to make the hold on part 2 easier.

 

So, I guess my point is: if the expectancy of one part is much higher then you should do it that way but be prepared for the psychological downside of not scaling and armour yourself against it. Its also possible though that your strategy for the second part is suboptimal - have you looked at some sort of trailing strategy that would really capture big trends when they come?

[SIUYA 11]

that DIBS method is an interesting one - you are right Kiwi, the take profit on one is more the mental crutch that helps as some people tested to show it is net unprofitable.

Reminds me of the Joe Ross - buy three, sell, one, sell one, run one.... then add buy three, repeat.....

Problem then always becomes of when to take profits....how long to run things, how to trail the stop?

Is one of the reasons scalping appeals is that this avoids these issues?

The never ending trade off!

[daedalus 10]

Well after more testing (well actually only on 4 of the 6 pairs) it became obvious it was a flawed idea. The initial results looked awesome but that's only because i'm an idiot and had the spreadsheet doubling the amount of ticks won. Once I fixed that error reality hit home.

 

Now I think its fairly obvious you need a VERY high win rate (>90%) to justify this kind of thing, the thing was, my win rate on the entry I was using was only around 60%. Nowhere near good enough.

 

Here were the combined results at various targets using this entry:

 

What this shows is that with a hit rate at best of 60% you're best results still stay at trying to reach 1:1 on the trade.

 

Now is this to say that its a invalid concept? No. But it does mean that my trade setup wasn't as successful as it needs to be for this idea to work.

 

As far as the scaling out many of you have been discussing I too have never been a fan. It always seems like a psychological crutch for "reducing risk" (or in my view reducing profit) and having massive amounts of risk when the inevitable straight out loss comes along. I never got how the TTM guys got away with promoting the philosophy when its pretty hard to get the numbers to justify that kind of management.

 

Back to the drawing boards. :crap:

[bakrob99 11]

  daedalus said:

However, this would negate the moves that simply blast through those levels and turn them into pars rather than gains. But its something worth considering all the same.

 

If you're going to scalp and not take the bigger move, then missing out on the bigger move is not a problem.

[Kiwi 16]

  bakrob99 said:

If you're going to scalp and not take the bigger move, then missing out on the bigger move is not a problem.

 

Hi bakrob,

 

I seem to recall you describing and impulse retrace method of trading that would be perfectly suited to high probability low win/loss "scalps." Are you still trading this style? And, thus, as perhaps the only "scalper" in this thread would you comment on the merits of different probability/winsize ratios from your experience?

 

kiwi

[MightyMouse 12]

  daedalus said:

Well after more testing (well actually only on 4 of the 6 pairs) it became obvious it was a flawed idea. The initial results looked awesome but that's only because i'm an idiot and had the spreadsheet doubling the amount of ticks won. Once I fixed that error reality hit home.

 

Now I think its fairly obvious you need a VERY high win rate (>90%) to justify this kind of thing, the thing was, my win rate on the entry I was using was only around 60%. Nowhere near good enough.

 

Here were the combined results at various targets using this entry:

 

What this shows is that with a hit rate at best of 60% you're best results still stay at trying to reach 1:1 on the trade.

 

Now is this to say that its a invalid concept? No. But it does mean that my trade setup wasn't as successful as it needs to be for this idea to work.

 

As far as the scaling out many of you have been discussing I too have never been a fan. It always seems like a psychological crutch for "reducing risk" (or in my view reducing profit) and having massive amounts of risk when the inevitable straight out loss comes along. I never got how the TTM guys got away with promoting the philosophy when its pretty hard to get the numbers to justify that kind of management.

 

Back to the drawing boards. :crap:

 

Daedalus,

 

You might want to try to use stops and targets that have substantially different volatilities.

 

4 and 8 tick, say, volatility during the high volume periods can be 1-2 minutes apart or the rate at which volatility changes for such small ranges can be so fast that when you enter a trade, order flow is such that the 2 minute ATR increases from 4 ticks to 8 ticks or higher in no time. So in spite of the fact that your stop is twice as far as from your stop, surges in volatility make you a dead duck with minor changes from normal market activity.

 

If instead you choose something more like 50 ticks as a stop and try to scalp 10-12 ticks, from a purely mechanical perspective, I suspect you will have much better success. To trade this, you need to be good at identifying failure early and escaping before your stop gets hit. With a wider stop, you have the luxury of time on your side to let the normal market rotations bring price back to your entry or better to let you exit BE or at a gain less than you had hoped.

 

I hope this makes sense.

 

MM

[Jeremytang 10]

Deadalus,

 

What criterion are you using for entry? Is it mechanical? If so I may be able to assist in backtesting a scalping strategy over a very long history over the 6A and perhaps others.

 

Also considering that a scalping strategy is basically trying to steal profits from the general noise of the market, ever consider dynamically basing profit and loss stops off some multiple of recent atr?

[bakrob99 11]

  Kiwi said:

Hi bakrob,

 

I seem to recall you describing and impulse retrace method of trading that would be perfectly suited to high probability low win/loss "scalps." Are you still trading this style? And, thus, as perhaps the only "scalper" in this thread would you comment on the merits of different probability/winsize ratios from your experience?

 

kiwi

 

HI Kiwi - Basically I realized after 5 years that I was more comfortable putting some money in my pocket rather than sit and watch the tape (and my trade's profit) go up and down.

 

So I take a profit and have come up with a reliable method for re-entering in the trend (trade).

 

I do exit my first position at 4-7 ticks (ES) and my 2nd half rarely exceeds 12 ticks. I will re-enter up to 3 times on any additional move in the "trend" I am following.

 

In one move up or down I may have as many as 4-5 entries ... but the caveat is that each entry price must be LOWER in a down trend than prior entry (higher for an uptrend)

 

I used to have difficulty thinking this way because I would always be getting a worse price - but then I realized that if I werent getting a "worse" for my setup then I was actually entering when the trend had either stopped or reversed.

 

If I have consolidation (defined by me as 4 overlapping price bars, I won't take an entry until and break and trend continuation (either way) occurs.

 

Bottom Line: If you are a scalper - get a good technique for getting back in the trend.

[Jeremytang 10]

I'm curious bakrob, have you ever compared tightening stops as opposed to taking the trade off and on (in this case tightening it 3-4 ticks once say 7-8 ticks of profit has been reached)

[daedalus 10]

I appreciate all of the input here folks. I'm personally at the point where i'm going to stop pursing this specific concept.

 

And continue a different discussion here: http://www.traderslaboratory.com/forums/f229/rethinking-every-concept-8332.html#post101997

 

Cheers!

[bakrob99 11]

  Jeremytang said:

I'm curious bakrob, have you ever compared tightening stops as opposed to taking the trade off and on (in this case tightening it 3-4 ticks once say 7-8 ticks of profit has been reached)

 

Sure ... if I see big tape movement ... I will try to get into a runner ... but it is very rare. I am actually better off exiting ... then re - entering. FYI, my re-entry criteria has a stop which is quite a bit smaller than my original entry criteria.... so it either works or I'm out and waiting for the next setup. I may trade 1/2 size for the re-entry... and keep doing it until it fails.

[friva 10]

Hi Daedalus,

 

no, I do not think it's BS, expectancy is key.

 

In my view it is important to make all single trades you take as insignificant as possible from an emotional point of view.

 

One way to achieve that is to trade as frequently as you can using a positive expectancy strategy.

 

Expecting "too much" from a trade is one way to make a trade emotionally "too significant", which also means it will be more difficult to close it when necessary.

 

You can make astonishing returns on your margin even missing most of the market moves.

Of course if you catch most of the market move your returns will be no less than "monstruous" but I think that "astonishing" is good enough. Why be greedy ?

 

All the best

 

FR

[klatchers 10]

1) Frankly, I think the entire premise of the thread was flawed. I have seen enough of the person you referenced in your original post to know he is a screaming idiot.

 

2) I don't care if Mahatmi Gandhi came up with this idea. Without a multi-year backtest, and a lengthy walk forward test to prove the concept, and an independently audited track record (including realistic commission/slippage/spread costs), it is an idea, not an edge or something I would put real money on.

 

3) I have tested all sorts of profit targets, stop losses, entry and exit methods. As have some researchers and publications. It does not give you an outperformance edge, it just lets you balance potential profit/ potential losses. The whole basis for this seemed flaky and quackish to begin with. Your edge comes from uncovering some kind of quirk in market behavior, not money management. Yes the casinos have an edge, that comes from the fact the games FAVOR them. Playing around with money management has none of that. The fact is, you are still the gambler, trying to overcome commission/slippage/spread costs.

 

Now the guy who posted that initial quote of his results and compounding i'm pretty sure is legit... Another guy who uses a similar philosophy is the loved and hated Avery also known as TheRumpledOne and his Never Lose Again threads that lets be honest, have been posted on every trading forum under the sun and threads have typically gone to hell quickly due to a combination of stated success, thread title, the typical bantering about risk: reward, and avery's very combative personality.

 

But all of that aside - I think hes legit and is making serious money. I mean the guy got his group of disciples to meet in Monaco for the F1 grand prix this year.... I don't care which way you slice it that city is one of the most expensive in the world and the most expensive time of the year there is for the F1 race. Somehow I don't think he footed the bill from his donational indicator payments.

[pathfinder62 10]

Thanx Daed & M Knight. This is a great post you started and I hope to catch up in the next day or two and hopefully add something beneficial for consideration.

 

In the meantime, thanx again for sharing.

 

Tom M.

[maxr 10]

  MidKnight said:

Let me start by saying I think the whole concept of risk and reward is total BS....When we are entering trades, most of us wait and observe the market before we decide to jump in, the same should be done for exits rather than make up some big number to give a 10:1 reward so one can feel good about the trade prospect and put the trade on.

MK

 

I know this area is tangential to the topic here, but I'm very happy to hear an experienced trader say that MidKnight, and I'd appreciate some opinions on the following if you'll forgive the diversion. As a rookie trader in receipt of a good deal of trading education from the conventional to the whacky, the way R/R is handled by many trading coaches worried me from the start - typically something like: "OK guys, we got a pattern forming up here, there's your technical profit target, here's your stop, we got a R/R of, let's see, 6:1 - wow, way to go...". Problem is, the only quantifiable part of that statement is the stop (providing you don't move it!). This kind of 'R/R' feels like conjecture based on assumptions as to whether a particular chart pattern achieves a pleasing geometrical symmetry (and the observed probability of that happening is rarely part of these R/R calculations).

 

One day, though, a coach said "OK, what are you actually worried about? The answer is simple - how much can I lose on this trade? So, decide the maximum % of your trading capital you're prepared to lose on each trade (1%, 2% or whatever). If it looks like a potentially good trade on the chart, decide on a good technical stop, work out your position size from that and your maximum loss per trade, and go trade it. You've fixed your risk, now follow your exit rules as the trade progresses." That made sense to me - he was saying, 'set the risk, because the reward is unknowable in advance, and if your method has an edge and you follow it, you'll get there on probabilities'.

 

That's great - except.... Say you enter long on a 15 min chart after a classic trend break and fade. You work out your position size from the technical stop and maximum loss (let's say 2 %) you've decided on - comes out to say 2 lots. It goes your way, and you set the stop to breakeven. Risk is now 0 (apart from maybe some slippage on your stops). Up it goes, and another pullback entry presents itself. You're happy trading at max 2% loss, so you enter with another 2 lots (which gives you total 1% risk now), and you're in luck again - this is becoming a trend, so you can move the stop up to breakeven on the last 2 lots and the same level on the first two. You're now trading 4 lots at breakeven or better with guaranteed profit on 50% of the position. Another leg up, and another entry signal occurs after a third pullback, so you do it again - 6 lots now, with guaranteed profit on 2, and better than breakeven on 2. We'll leave the trade there in case Elliot wave enthusiasts get overexcited, but here's the question...

 

By scaling up in these circumstances, would I be actually following the 2% max risk level I set (and indeed reducing the total risk below 2% with each additional entry), or would that figure as generally understood include the assumption that a stop woul be set to breakeven as soon as practicable and the position size would only be held or reduced, not increased? I know the answer is as long as a piece of string, but I'd appreciate your views on whether this would be a logical course of action or not.

 

Thanks, Max

[SIUYA 11]

  maxr said:

You work out your position size from the technical stop and maximum loss (let's say 2 %) you've decided on - comes out to say 2 lots. It goes your way, and you set the stop to breakeven. Risk is now 0 (apart from maybe some slippage on your stops). Up it goes, and another pullback entry presents itself. You're happy trading at max 2% loss, so you enter with another 2 lots (which gives you total 1% risk now), and you're in luck again - this is becoming a trend, so you can move the stop up to breakeven on the last 2 lots and the same level on the first two. You're now trading 4 lots at breakeven or better with guaranteed profit on 50% of the position. Another leg up, and another entry signal occurs after a third pullback, so you do it again - 6 lots now, with guaranteed profit on 2, and better than breakeven on 2. We'll leave the trade there in case Elliot wave enthusiasts get overexcited, but here's the question...

 

By scaling up in these circumstances, would I be actually following the 2% max risk level I set (and indeed reducing the total risk below 2% with each additional entry), or would that figure as generally understood include the assumption that a stop woul be set to breakeven as soon as practicable and the position size would only be held or reduced, not increased?

 

- it is as long as a piece of string....with many questions of how to pyramd in, pyramid out, is it better to buy things all at once, exit all at once, how to control the overall risk. Will you put a limit on things....

 

Most will depend on your personality and risk tolerance.

I think you really need to have in your head as well exactly what you are doing, how you view it and why (proove to yourself ) that is makes sense mathematically, sensibly and that it works well in the market.

Understand there is a big big difference in viewpoints of risk...... eg; you may think you have zero risk, but what then is your gap risk, what is your actual exposure..... dont forget most futures contracts are actually highly leveraged.

Your comment "so you enter with another 2 lots (which gives you total 1% risk now)" I would say you have some confusion.....

You determined that you would risk 2 contracts based on 2% risk, just because you had some unrealised PL does not change this.

Are you going to base everything off you original stake going forward, or are you going to trade the equity as you have more or less PL? How soon do you enter after the original trade, an ATR level, a PL level, on a time basis?

I understand what you are getting at but more info, thought and scenario analysis needs to be done for the whole process.

 

But lets just say you have entered really quickly, you have now bought 3 units (6 contracts) and you have all the stops at break even. What happens if there is a gap through the stops?

Also "guaranteed profit on 50% of the position" - 50% of what? the original stake, from where? There are also no guarantees in trading!

 

Best bet is to take something like Ninja trader or something that can easily show you the difference between what happens when you pyramid..... use a very simple model eg, a MAtrigger and then a continual pryamiding as it goes your way.... dont worry about what happens to the PL - look at the series of trades...... can you live with that?

[neutral 10]

I cant look at a clean message anymore.

There is an add covering part of the tekst always ???????????????????????????????????

This is very annoying!!!!!!!!!!!!!! and absurd.

 

 

http://www.forexyard.com/en/landingpage.tpl?zone_id=8187&banner_id=790&lp=2.5perc

 

 

Owner of TL, please aline the need to generate income more with consideration for the the members experience.

Edited August 14, 2010 by neutral

[snowbird 10]

  bakrob99 said:

If I have consolidation (defined by me as 4 overlapping price bars, I won't take an entry until and break and trend continuation (either way) occurs.

 

bakrob,

 

appears you have a very simple method of defining consolidations.

 

just curious as to your definition of 4 overlapping price bars.

 

Do you mean 4 bars (3 of which are inside bars as compared to the 1st?

 

or do you just look at the last bar printed, and if any part of it overlaps the previous 3 bars you deem a consolidation and trade only breakouts of the high or low of this consolidation?

 

snowbird

[maxr 10]

  SIUYA said:

 

....But lets just say you have entered really quickly, you have now bought 3 units (6 contracts) and you have all the stops at break even. What happens if there is a gap through the stops?...

 

Thanks very much for your comments SIUYA. I forgot to mention that I only day trade, and watch the chart most of the time I'm in a position, but I take the very good points you made, thanks. The problem for a new trader is that few trading coaches I've met appear to rarely think through these things thoroughly (hey, they have to find something they can teach easily), and most full time traders I've spoken to have found something that works 'well enough' for them, so don't continue to analyse it beyond that - maybe that's not such a bad idea:doh:

 

Max

[windsurfer 10]

I haven't seen this point before, so I'll mention it a see if anyone finds it interesting.

 

I traded for a while with a number of former pit traders. They all make their living by scalping, although each has a different strategy. They all have a basic strategy of taking a trend (e.g. bullish) and scalping in only 1 direction (e.g. on dips). They are of 3 types- changing the trend direction during the day; taking a trend direction for the day and only trading in that direction; and 2 traders who only trade on the short side. Surprisingly to me, the short only scalpers are the most successful. They usually lose on bullish days, although often they are even profitable then from shorting on overbought ticks, but they have always profitable and some even great days when the market falls.

[SIUYA 11]

MaxR - I think windsurfer makes some good points about doing one thing and one thing well.

Especially the shorting - I also know a couple of equity traders that only every short..... perpetual bears, they make money quick and fast and usually still make money in bull markets.

 

Dont get me wrong on the previous post, scaling in/pyramiding/stacking orders works and works well, but you really need to understand the various scenarios you can get into trouble, and really understand why you make or lose money.

I have seen a lot of styles and its often scary the variation between the risk assumed and the actual risks in a disaster.

If you are pyramiding while day trading, you effectively are just looking for trend days. Again I would look to test and see if your setup is good at picking these, or you have some other measure to help minimise the losing days when there is no trend.

 

Its actually something I have been working on, but more longer term....thanks for a couple of recent threads (thank you Daedalus ) they are topical. like every trading style there is usually a trade off with regards drawdowns, required capital and how far to run things.

Do a test....look at the various scenarios.... really understand what suits you.

[bredin 10]

"I tried that and it doesnt work"

 

Most annoying phrase on a forum.

Or in life.

 

"I tried" is a phrase used by losers in order to make themselves feel better about losing, a way to divert the cause of failure onto something other than themselves.

 

What they mean to say is "I made a few trades and got it wrong, so I quit."

 

Scalping does work, there are a myriad of methods (some better than others) that produce long term positive results with practice.

 

The method taught to me to scalp on M5 also works on H4, which I am now trading exactly the same way except with less of my actual money on the table at any given point, a slightly tighter stop, and less intense chart time.

To clarify, if my M5 risk was 10 pips, on H4 I have a smaller lotssize and so only risk 3-5 pips for gains of 7-10 pips. I may not be entirely clear, but guys that can calculate risk will be able to work it out

Think of it as scalping in slo-mo

 

G.

[TheRumpledOne 14]

  klatchers said:

1)Now the guy who posted that initial quote of his results and compounding i'm pretty sure is legit... Another guy who uses a similar philosophy is the loved and hated Avery also known as TheRumpledOne and his Never Lose Again threads that lets be honest, have been posted on every trading forum under the sun and threads have typically gone to hell quickly due to a combination of stated success, thread title, the typical bantering about risk: reward, and avery's very combative personality.

 

But all of that aside - I think hes legit and is making serious money. I mean the guy got his group of disciples to meet in Monaco for the F1 grand prix this year.... I don't care which way you slice it that city is one of the most expensive in the world and the most expensive time of the year there is for the F1 race. Somehow I don't think he footed the bill from his donational indicator payments.

 

Combative? Me? I never hit first. But I don't turn the other cheek and that's what gets me banned. Usually, the instigators are part of the in clique and don't get banned for violating forum rules when they attack me.

 

When I look back over the years I have been posting, I have to chuckle at all those who have hurled insults my way. Why anyone would take issue with what I do is beyond comprehension. A forum is like a buffet - take what you want and leave the rest. If you don't like a certain dish at a buffet, you don't spit in it, do you? So why trash someone's thread just because you don't like it? :helloooo:

 

BTW: Monaco was nice to visit but don't think I want to live there. It is too noisy. Next year, I am thinking having the pirate gathering at the 24 Hours of Le Mans.

[SIUYA 11]

TRO - just out of interest, how many forums have you been banned from, and is the stint here at TL possibly going to head toward a record?

 

I for one, appreciate the simplicity of the style/manner you use....even though it did take me a little while to follow (but thats just cause I am slow)

[thalestrader 14]

  SIUYA said:

...I for one, appreciate the simplicity of the style/manner you use....

 

I agree. TRO may never hit first, but the fact that he hits back at all has caused many to overlook the potential usefulness of much of his work. Of course, many also have a problem with his "donational" package of indicators for which he does require a payment. So what? It seems he gives much more for much less than most who make a business of selling their programming creations to traders. Some find it disagreeable that TRO is an IB for a forex broker. Again, so what? If the guy gives something of value, and if he provides support to those who are using his indicators, why should he not make a return on the time he invests in providing that value?

 

 

 

 

Best Wishes,

 

Thales