The Killing Floor | Page 9 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Oanda Killing CPU 3 replies

making a killing!! in this market 48 replies

making a killing!! in this market prt 2 71 replies

Off quotes killing my EA 0 replies

Spam filters are killing my chart alerts 0 replies

Attachments: The Killing Floor
Exit Attachments

The Killing Floor

  • #161
  • Jul 2, 2018 6:07pm Jul 2, 2018 6:07pm
Quoting John.Doe
Is there any reason why you use TF 5M and MAs with periods 200, 50, 10 instead of TF 1M and MAs with periods 1000, 250, 50?
I've found 200/50/10 to be great for the 5 minute chart - maybe your suggestions of 1000/250/50 could be good for other time settings.

Yesterday's trades - still need to work on those entries to get in and out quickly. I think I will look at trading first two hours of London only and maybe first hour of NY open. Other times I find prices just moving sideways and not doing much.
Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot1.png Size: 59 KB
  • #162
  • Jul 2, 2018 10:05pm Jul 2, 2018 10:05pm
My style short term trading since 2010 , i know the risk and money management.. I never blame other about any technique error , but mistake teach me be better . I do have same similar technique like this, i call hit n run on GPB/JPY .. That time can make 300pip perday 2011 , i like to try something new and learn more .. As lone wolf always know the risk .. Hehehe .. Thank for advice No stress
  • #163
  • Jul 3, 2018 4:48am Jul 3, 2018 4:48am
Quoting Sis.yphus
{quote} 1:1 leverage per order is about 0.3% risk for each order that takes a 36 pip loss. So 3 full stop losses in a row is 1%. If you stack properly, you can make up the lost pips on the next 2-3 moves easy. The fluctuations are supposed to be small so you can trade often, because it's the CLT and Law of Large Numbers playing out over 1000s of trades that is actually going to make the difference, not the result of any one trade.
Can you provide some statistics?
How many trades are you keeping open in one time? There is a consolitated risk trough pairs, currencies and open positions.
What is the number of trades per week and what pairs are you trading?
With such low leverage and TP you need to make 1000 trades per month to get some perfomance..

SP
  • #164
  • Jul 3, 2018 8:39am Jul 3, 2018 8:39am
Quoting John.Doe
Is there any reason why you use TF 5M and MAs with periods 200, 50, 10 instead of TF 1M and MAs with periods 1000, 250, 50?
There is nothing special in the periods of the MA. The point was to use the most popular settings, but instead we use them in a scientific way based on data. The cycle numbers will look similar for any periods remotely close to the ones mentioned in the first post.

Quoting SuperPip
{quote} With such low leverage and TP you need to make 1000 trades per month to get some perfomance.. SP
As I mentioned in my last post, that is literally the point of trading so small, so we can trade extremely often. This is extremely important if you understand the Probability of Profit in A Series post I put up.

With this system, every 400 trades you place in a series, 5% of the time, you will be in the red after those 400 trades... This is because of the law of large numbers and the central limit theorem.

In order to really see the edge play out you need to make 1000s of trades. Ideally if you could place 30 trades a day, or 150 a week, you have an 89% chance of ending the week in the money.

If you place 30 trades a week, one out of every 10 week trading periods you participate in, you will lose money after the 10 weeks, and that is statistically normal.

If you place 3 trades a week, you can trade for the entire year with a winning system, and 1 out of every 10 years you will lose money on the year. The big advantage HFTs have is they can do what most traders do in 2-5 years, in 1 day. So their statically normal losing periods only last hours relative to the amount of trades they make a day, instead of a year relative to the amount of trades another guy makes in a decade.
  • #165
  • Jul 3, 2018 10:32am Jul 3, 2018 10:32am
Quoting Sis.yphus
{quote} In order to really see the edge play out you need to make 1000s of trades.
Hi Sis!
Thank you for the answer! I really appreciate what are you doing here!
How many simultaneous trades is allowed to have?
One open trade for one pair?
Is it OK to have trades open in same time for UJ, GJ, EJ (risk for JPY is tripled)?
What pairs are suggested to trade, all 28?

SP
  • #166
  • Jul 3, 2018 10:59am Jul 3, 2018 10:59am
Quoting SuperPip
{quote} Hi Sis! Thank you for the answer! I really appreciate what are you doing here! How many simultaneous trades is allowed to have? One open trade for one pair? Is it OK to have trades open in same time for UJ, GJ, EJ (risk for JPY is tripled)? What pairs are suggested to trade, all 28? SP
I would only trade one pair at any given time. either EU or UJ since they will have the lowest spreads. How much you stack is up to you.
  • #167
  • Edited 11:29am Jul 3, 2018 11:19am | Edited 11:29am
Quoting Sis.yphus
{quote} With this system, every 400 trades you place in a series, 5% of the time, you will be in the red after those 400 trades.
With win rate 95% and 4:36 RR you can risk up to 50% per position by the Kelly criterion.
Monte Carlo simulation is attached.
Attached File(s)
File Type: xlsx Kelly MonteCarlo.xlsx   182 KB | 518 downloads


SP
  • #168
  • Jul 3, 2018 1:13pm Jul 3, 2018 1:13pm
Quoting Sis.yphus
{quote} I would only trade one pair at any given time. either EU or UJ since they will have the lowest spreads. How much you stack is up to you.
Hey Sis,

I would love to read your perspective on why you are against automation (or think automation would yield in different results...assuming you meant poor performance). Doesn't have to be specific to this strategy.

Looking at your background, almost anything manual can be algorithmically automated. I like automated because of the consistancy in application of an approach without any emotions so just curious if there is anything else more to it.
Staying in my lane...
  • #169
  • Jul 3, 2018 1:59pm Jul 3, 2018 1:59pm
Quoting VEEFX
{quote} Hey Sis, I would love to read your perspective on why you are against automation (or think automation would yield in different results...assuming you meant poor performance). Doesn't have to be specific to this strategy. Looking at your background, almost anything manual can be algorithmically automated. I like automated because of the consistancy in application of an approach without any emotions so just curious if there is anything else more to it.
I have waved back and forth on this issue as my knowledge and experience in trading has grown. I agree with you regarding the emotional aspect and what you mentioned in regards to consistent application of the system.

As I first started getting into the statistics/data a few years ago I was 100% gunho on the fully automatic strategy that you turn on and go sit at the beach for the rest of the year. Now I am still very much for automation, and breaking down a system to the point it could be automated by completely understanding the 100%-mechanical entries and exits as this gives you a really good idea of how price behaves within certain parameters, I just think you can benefit from having a layer of discretion on top of it.

I think you can have great success with a fully automatic strategy, however, I believe truly exceptional trading has a degree of discretion associated with it.

The algorithms/automated trading are great for most day-to-day activity in the markets where most of the moves are typical noise of the market. However there are occasions where you will know you need to get out of a trade before your algo does, allowing you to cut some costs, and there are times where you will know to size up using house money to catch a crazy wave and milk it for all it's worth.

That is just my general opinion on the issue.
  • #170
  • Jul 3, 2018 2:24pm Jul 3, 2018 2:24pm
Boosting Win Rate Through Management
***WARNING: TRADE MANAGEMENT CAN BE DANGERIOUS IF YOU DON'T KNOW WHAT YOU ARE DOING***
***DO NOT ATTEMPT ANY CRAZY/SIGNIFICANT MANAGEMENT STRATEGIES UNTIL YOU ARE COMFORTABLE WITH THE SYSTEM***
***BELOW IS FOOTAGE OF THE LAST GIRL AGGRESIVELY MANAGING HER TRADES WHEN SHE SHOULDN'T HAVE
Attached Image


Something I have talked a lot about on this thread is how the risk/money management is a huge aspect of really taking the most advantage of this system, and any system for that matter. I have been able to gather that many are thinking: How this is true? How can it really make that much of a difference? This guy is full of shit!

This post will be short because the point is rather easy to make. In order to solve this mystery, every one will need to go and buy a magic conch from Amazon, and wait for further instruction upon the arrival of your conch...

Once your conch arrives, come back to this thread and check out the data below.

In this post we are going to use a 1:1 risk:reward ratio and a edge over the spread that equates to 0.1% - very tiny! Not as bigly as we'd like to see. But it'll do!

Without management, just taking the trades as they come, what do you think the probability is that we will see a profit after a series of 100 trades?

If you guessed around 50%, you are pretty close, it's actually only 46.759%.

So that means with this system, we can expect to lose money every 100 trades, more than half the time. That sounds like a nightmare to trade, I know.

Now let's see what a little bit of management can do for us in this scenario. I am not going to go into the details on how the simulation is managing the trades, it's nothing fancy though, essentially stacking in one form or another.

So how much of a boost do you think we will see? 5%? 10%? 20%? I know some of you greedy bastards reading this are looking for 30%+.

Welllllllll if you guessed 98.78% you are correct!

That's right. With a proper management strategy, you can boost your probability of profit in a series of 100 trades from 46.759% to 98.78%! Pretty flipping cool if you ask me!

So is trade management important? Hell yeah it is. It's so important that it can make the tiniest edge ever very profitable.
  • #171
  • Jul 3, 2018 4:08pm Jul 3, 2018 4:08pm
Hahh teach us that management strategy you probabilitywizard! lol
Necessity knows no law except to conquer
  • #172
  • Jul 3, 2018 6:21pm Jul 3, 2018 6:21pm
Quoting Sis.yphus
If you guessed around 50%, you are pretty close, it's actually only 46.759%.
Wouldn't that number be more close to 50.05 %? 0.1 % advantage would mean to have a probability of winning in 50 % + 0.1/2% = 50.05 %. Now you're just sampling from the population and thus (on average) the expectation for samples should be equal the mean of the population.
  • #173
  • Edited 4:38am Jul 4, 2018 3:36am | Edited 4:38am
Couple people DM me on discord asking about entries. Sorry for my poor drawing skills, but here is an example;

1. 200MA down
2. 50MA down
3. 10 MA down

Now wait for a small pullback to the 10MA, and enter. Follow post #1 - TP 4 pips, manually close at 15pips and put in SL for 36 pips in case of emergency (internet cut or what not).
Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot1.png Size: 20 KB


EDIT - couple of minutes later, out with profit
Attached Image
  • #174
  • Jul 4, 2018 3:52am Jul 4, 2018 3:52am
Quoting Sis.yphus
{quote} I have waved back and forth on this issue as my knowledge and experience in trading has grown. I agree with you regarding the emotional aspect and what you mentioned in regards to consistent application of the system. As I first started getting into the statistics/data a few years ago I was 100% gunho on the fully automatic strategy that you turn on and go sit at the beach for the rest of the year. Now I am still very much for automation, and breaking down a system to the point it could be automated by completely understanding the 100%-mechanical...
Thanks. I would agree with having some discretion with exits or around trade management but prefer entries to be 100% automated (in a deceptive way that is to make it harder for anyone to figure out). I think it is necessary to have a uniform setup "framework" in order to consistently apply trading stats analysis for future trades... I hate moving parts, I hate unknowns, I hate thinks that I haven't thought about or accounted for. With my discretionary trading, I never bothered with trading stats analysis in detail because how can one statistically measure our own actions ? It just did not make sense (besides, looking back, one of my main weakness was with execution... no matter how hard I tried, I just could not work and think like a robot!). Now, with my automation, I do "allow" discretionary actions but only within the context of my automation.
Staying in my lane...
  • #175
  • Jul 4, 2018 4:09am Jul 4, 2018 4:09am
Quoting Sis.yphus
Boosting Win Rate Through Management ***WARNING: TRADE MANAGEMENT CAN BE DANGERIOUS IF YOU DON'T KNOW WHAT YOU ARE DOING*** ***DO NOT ATTEMPT ANY CRAZY/SIGNIFICANT MANAGEMENT STRATEGIES UNTIL YOU ARE COMFORTABLE WITH THE SYSTEM*** ***BELOW IS FOOTAGE OF THE LAST GIRL AGGRESIVELY MANAGING HER TRADES WHEN SHE SHOULDN'T HAVE {image} Something I have talked a lot about on this thread is how the risk/money management is a huge aspect of really taking the most advantage of this system, and any system for that matter. I have been able to gather that many...
Good post. Do you think an edge-less (i.e. 50:50 win/loss negative-sum system before spreads) could be converted into a profitable with proper trade or risk or money management? In my findings also, most presumably profitable systems I tested (systematically) resulted in a 48%-52% range leading me to believe the following:

  1. For every profitable trade, there is almost always going to be a losing trade
  2. For every stop loss hit, there is also always going to be a corresponding take profit
  3. For every statistically proven profitable entry, there is almost always going to be a losing exit (and vice versa)
  4. For every successful bounce or breakout or fakeout or ___________, there is a unsuccessful/failed bounce or breakout or fakeout or ________.

I am not being a debbie downer but this is what I was seeing over a significant large data set which really made me focus more on volatility-induced fat tails when real profits are made. Anyways, I feel I am repeating what I have said before in my posts.

Keep the gems coming bro. You are doing a great service in breaking down the complexities for most of us to understand.

Staying in my lane...
  • #176
  • Jul 4, 2018 4:17am Jul 4, 2018 4:17am
Quoting Ivanyo
Couple people DM me on discord asking about entries. Sorry for my poor drawing skills, but here is an example; 1. 200MA down 2. 50MA down 3. 10 MA down Now wait for a small pullback to the 10MA, and enter. Follow post #1 - TP 4 pips, manually close at 15pips and put in SL for 36 pips in case of emergency (internet cut or what not). {image} EDIT - couple of minutes later, out with profit {image}
How many retests will you take before you like to say a larger pullback to the 50 or 200ma?
  • #177
  • Jul 4, 2018 4:40am Jul 4, 2018 4:40am
Quoting Hazza88
{quote} How many retests will you take before you like to say a larger pullback to the 50 or 200ma?
Usually the best entry is when price has moved away from the 10 MA, then pulled back 2-3 candles. Those are the best entries.

If it goes back to 50 or 200 MA, I would cut the trade as less likely to turn out the be a winner.

For example, yellow circle designates what would be deemed as a good entry area to go short.

Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot1.png Size: 30 KB
  • #178
  • Jul 4, 2018 5:30am Jul 4, 2018 5:30am
Quoting Ivanyo
{quote} Usually the best entry is when price has moved away from the 10 MA, then pulled back 2-3 candles. Those are the best entries. If it goes back to 50 or 200 MA, I would cut the trade as less likely to turn out the be a winner. For example, yellow circle designates what would be deemed as a good entry area to go short. {image}
Hi Ivanyo. Your chart example above (while good) gives a clue as to why you are sitting in trades so long.

If you look at the sequence of your MAs as you enter, you have the 50 over the 200 over the 10.

Now, if you look just slightly farther on, you will see that we get the 200 over the 50 over the 10.

Notice how here, the candles don't go back up and close over the 10 but close below and we then get a much stronger drop away?

Now, of course, you will likely get a lower entry (in fact you will get a lower entry) most of the time but the move will be stronger.

Balance that against your more aggressive entry (earlier) that will keep you in a holding pattern for a while (often quite a long while) and could possibly be a turn against that little bit more patience rewarded with a stronger move which you can then stack like a madman as it drops.

However, in the end, it's horses for courses - but perhaps just something to bare in mind.
  • #179
  • Jul 4, 2018 7:08am Jul 4, 2018 7:08am
ivanyo, nice trading.
it reminds me very much to choros thread here: https://www.forexfactory.com/showthread.php?t=236210
trading is like dancing, don't think and just listen to the music!
  • #180
  • Jul 4, 2018 8:00am Jul 4, 2018 8:00am
Quoting AntiCre
{quote} Wouldn't that number be more close to 50.05 %? 0.1 % advantage would mean to have a probability of winning in 50 % + 0.1/2% = 50.05 %. Now you're just sampling from the population and thus (on average) the expectation for samples should be equal the mean of the population.
i know it seems like that would be the case, but because the count of wins and losses will folllow a geometric Brownian motion themselves(along with price) it is extreme rare that after 100 trades the losses and wins will equal each other. That is because in these sample sizes, we will see an arcsine distribution instead of a Gaussian one.

if this wasn’t the case we could easy just find over supply of losers within a series and bet it will even out before the 100 trades are up.

probabilties are some of the strangest behaving things and that is what makes them so fun/challenging to work with.
0 traders viewing now