Clemmo's Trading Odyssey | Page 9 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Day Trading vs. Swing Trading vs. Position Trading 86 replies

Attachments: Clemmo's Trading Odyssey
Exit Attachments

Clemmo's Trading Odyssey

  • #161
  • Edited 11:44pm Aug 25, 2021 10:25pm | Edited 11:44pm
https://www.mql5.com/en/signals

Attached Image (click to enlarge)
Click to Enlarge Name: 5f6995f9d902d138246dd1b981c63a08.png Size: 64 KB

The MQL5 signals service offers the chance for traders who lack confidence, knowledge and skills to presumably follow someone who does and profit when they do. It sounds great. It's a win-win for both the service provider and the client. Why then, isn't everyone making a fortune with signals?

For many of the same reasons everyone isn't making a living trading. For one thing, being a consistently profitable trader is hardly easy.
For another thing, offering a signals service and making it look good is comparatively easy. There are probably hundreds of ways to dress up a dangerous system and make it look good. I won't go into all of those now, but it's a good topic for later.

Let's just look at the top rated systems on the signals service. MQL5 has something called 'reliability' and I don't know how it is calculated, but let's just see what is reliable about these signals.

I'll start with 'Kauser Ahmed' which has produced 205% growth this year. Not bad at all, but such a 'low' percentage compared to the most profitable systems probably doesn't draw in the crowds. Still, being at the front page of the signals service can't hurt.

The first thing to notice is that this has been running for 17 weeks - just over 4 months. Not such a long track record, but recall what Taleb said about track records anyway.
Attached Image (click to enlarge)
Click to Enlarge Name: 6e2f3f771bb26e1c7d29b2995cdc6b2a.png Size: 20 KB

The signal started with $100 and then Ahmed added $317 (likely pissing off his existing customers, as doing this scales all trade volumes up) and he has been able to double that. What's not to love?
  • #162
  • Aug 25, 2021 10:59pm Aug 25, 2021 10:59pm
A little known feature of the signals service is that if you click on 'copy signal' and go to the signals panel there's this little button that says 'visualize on chart'. Click that! It's free.

Attached Image (click to enlarge)
Click to Enlarge Name: 3c04644e5f2685b70b9685f39249af1d.png Size: 67 KB


On MT5 it's in the upper right hand corner. You don't need to be subscribed to the signal to see this. You don't even need to be logged in to MQL5.

Attached Image (click to enlarge)
Click to Enlarge Name: 31167e474240e34691969bf592301c1f.png Size: 19 KB
  • #163
  • Aug 25, 2021 11:14pm Aug 25, 2021 11:14pm
What this does is open charts that have been recently traded by the signal provider. It not only lets you see what trades have been recently completed but it gives you a good idea of what kind of system the trader is using!

Here we get a very good sense of how 'Kauser Ahmed' trades and what the rationale is behind each trade.

Attached Image (click to enlarge)
Click to Enlarge Name: 99afce8b7dfcdf9fbc4b069c9aa7b7af.png Size: 23 KB

You can see he (or the algo) went short early in the session. And they were wrong. So what does it do? It goes short again.
And it's wrong. So what does it do? It goes short again.
And again.
And again. Each time the long trend continues? Does this system change it's mind? Hell no! It goes short again, and at a higher volume each time.

Know what this is called? Doubling down, or commonly, a Martingale strategy. Now, I won't get sidetracked into going over the relative merits and risks of the martingale system, but I'm sure we can all agree, what it is not is associated with RELIABILITY, and this is the first or second most reliable system shown to us under the 'reliability' table of the MQL5 signals service!

Is this system going to blow up? Oh hells, yes. Yes it will. When?? Well, you never can tell, but it will take all 42 of its current unsuspecting subscribers with it.

If I had time (or if there was interest) I could go through all the top systems in MT4 and MT5 and together we'd find nearly all of them are martingaling, or trading a simple grid, or doing some other kind of shady trick that pulls down the big numbers temporarily only to stall or blow up later. When this happens, you just apologize to your subscribers and wash, rinse, repeat.

But, if these highly subscribed and visible systems are no good, does that mean all signals are bad?
  • #164
  • Aug 25, 2021 11:28pm Aug 25, 2021 11:28pm
If I go to the very back pages of the signals service I can find people like Borgesti.

Attached Image (click to enlarge)
Click to Enlarge Name: 8e35b331c4bc222548e2718bd3f5419c.png Size: 56 KB

They've only made 50% in over a year. They aren't exciting. They aren't considered 'reliable' by MT5 for some reason. But when I look at their charts.

Attached Image (click to enlarge)
Click to Enlarge Name: d199d6c7e0d1562d0981ab3f6e6d2cf9.png Size: 56 KB
I don't mind what I see. They have some way of detecting that EU is in a relatively expensive state and they trade short scalps, waiting for retraces before going short again. I don't like everything they do, as I see in other charts they have a tendency to hold on to hopeless positions too long, but this is far better than what the crowds flock to on the most popular signals. If you follow Borgesti you'll make 50% a year, and never blow up. That's a better proposition than you'll get from your bank or WealthSimple!

Borgesti has 3 susbscribers. This is maybe 3 too many. I bet I can find some lowly anonymous signal provider with no subs who is killing it with a safe and profitable strategy but without the fast and flashy numbers that draw in the crowds.

If you need more examples, let me know but that's Clemmo's Conjecture in action.
  • #165
  • Edited Aug 26, 2021 12:00am Aug 25, 2021 11:37pm | Edited Aug 26, 2021 12:00am
The Talosian Trap
Attached Image (click to enlarge)
Click to Enlarge Name: TOS_1x00_011.jpg Size: 59 KB

The second principle I want to talk about I'm now calling the 'Talosian Trap', not to be confused with the Talos Principle which is a puzzle game by Croteam/Devolver (recommended).

The Talosians are aliens from the original Star Trek TV series who are so damned smart they don't even need to talk, as they can communicate telepathically. When they do, their heads pulsate and we hear their voices in our heads like someone talking in a large cavern. They are so smart they can even beam illusions directly into the minds of their intended victims.

The Talosians are not half-smart, they are intellectual giants. They are so smart they are actually more evolved than human beings. We cannot hope to match them on any IQ measurement scale. And yet...and yet, somehow Captain Kirk and Spock are able to rescue Captain Pike from the Talosian prison and save the day at the end of 'The Cage'. Pure science fiction fantasy nonsense right?

And yet, somehow some version of this plays out in trading, I suspect everywhere. The most spookily intelligent and technically competent traders are often woefully unsuccessful, where non (or even anti-intellectuals) thrive on the chaos of markets.

What's my proof? Well, again, this is conjectural, but let's start with something I read today from the MQL website again.
  • #166
  • Aug 26, 2021 12:11am Aug 26, 2021 12:11am
Quoting clemmo17
Doubling down, or commonly, a Martingale strategy. Now, I won't get sidetracked into going over the relative merits and risks of the martingale system, but I'm sure we can all agree, what it is not is associated with RELIABILITY, and this is the first or second most reliable system shown to us under the 'reliability' table of the MQL5 signals service! Is this system going to blow up? Oh hells, yes. Yes it will.
Very true. Good to see you pointed out all this!

Here is place for a new website where analysis is done and true reliability ratings applied. But again... that website wont be in google 1st page search results and nobody will notice.

How about darwinex? They provide various metrics for signals. But trades are not exposed. Still its posible to import history and then see metrics.
  • #167
  • Edited 1:44am Aug 26, 2021 12:27am | Edited 1:44am
https://www.mql5.com/en/articles/8274

The article is by Evgeniy Ilin and is called BASIC MATH BEHIND FOREX TRADING, but there is in fact, nothing basic about it. First of all, if you read the articles on the MQL5 site you know that most of them are written by quants (most of them Russian) and nearly all of them involve advanced algorithmic approaches to trading, either using some abstruse computer science concept or learning how to code classes in some library, or using machine learning, neural nets, that kind of thing. It is not for the fainthearted or the glibly technical. It's for serious math nerds, programmers, if you don't have a PhD in comp sci you are going to be tested.

Let's see what Mr. Ilin considers 'basic math'.

He starts by describing a basic trend/ranging dichotomy. He then outlines a model for pricing and market forces with market and limit orders. It reminds me of the market structure description that we just read in 'The Physics of Wall Street'. The essential market mover is the triggering of limit orders. Clusters of limit orders can cause the market to stall. Ilin then constructs a mathematical model for the 't argument' where t is time. It is a discrete function of ticks, that make up bars, in patterns of various scales and durations. Like me, Ilin believes the market is boundedly random. He examines events in terms of expected payoffs. He produces some equations to model the random payoff with a spread of zero.

He's only warming up though as now he creates a new equation to add closing by stop levels, and closing by signals.
M=P3*M3+P4*M4=P3*M3+(1-P3)*M4, where M3=P1*tp-(1- P1)*sl, while M4=Sum(P0[i]*pr[i]) - Sum(P01[j]*ls[j]); Sum( P0[i] )+ Sum( P01[j] ) =1
Got that?

There are two antithetic events whose outcomes form another set of independent event spaces. And then things just get crazier. He incorporates a profit factor equation with combinatorics and probability distribution densities. Ilin produces a diagram depicting nested events (just, you know, to make things clearer) and builds a random value distribution probability density graph for a discretized price at a fixed step in points.

Attached Image (click to enlarge)
Click to Enlarge Name: a524b48782f0081b01cbf8f61dff80db.png Size: 86 KB
A diagram notable for its aesthetic qualities, not its explanatory powers.

He goes on to outline a combinatoric theory of cells and attempts to prove his theory using MathCad. This is all before he starts getting into the code.

The point of all this is to say, Mr. Ilin is no intellectual slouch. He has technical credibility, at least I think he does. I don't have the skills or training to find the flaws in any of this. So such a skilled quantitative coder and mathematician must be a trading genius, yes?
  • #168
  • Aug 26, 2021 12:32am Aug 26, 2021 12:32am
Quoting joyny
{quote} Very true. Good to see you pointed out all this! Here is place for a new website where analysis is done and true reliability ratings applied. But again... that website wont be in google 1st page search results and nobody will notice. How about darwinex? They provide various metrics for signals. But trades are not exposed. Still its posible to import history and then see metrics.
One of my pet projects is to watch the Darwinex leaders and see how many of them stick around over the long term. There is actually one that seems to stick out from the crowd. https://www.darwinex.com/darwin/SYO
I have a lot more work to do to really understand the Darwinex ecosystem. Now that Darwinex will take Canadians I might even try my hand at my own Darwin, but I've been dragging my feet for a while.
  • #169
  • Aug 26, 2021 1:27am Aug 26, 2021 1:27am
Mr. Illin is not only brilliant but admirably honest, as most brilliant folks are.
He says, "With all my knowledge and experience, I currently don't have a profitable stable account. I have EAs showing good test runs but everything is not as easy as it seems."

But how can this be? If I had 1/10th of Mr. Ilin's technical ability I think I could generate millions in a few years. In fact, one of my ongoing quests is to partner with a competent coder to test out some more of my ideas, of which I have no shortage. However I'm neither persuasive or outgoing so I must take the first steps on my own. Still, how I envy people like Mr. Ilin. Why can't they make money trading when a relative ignoramus like myself can do it?

The Talosian Trap states that highly technical and mathematically skilled quants often get bogged down in the details of their analyses and fail to make the necessary compromises in their models in order to make pragmatic choices in trading. Hmm, I might need to try and tighten that up a little bit.

Mr. Ilin goes on to talk about martingales and grids, but he doesn't seem to be curious about the many other possible approaches to trading, and I have to wonder if the creativity that he shows in his math isn't being applied to his big picture thinking.

Says Ilin, "If you have no programming skills, then you have practically no chance at all. I've seen a lot of pseudo traders bragging about some results after having traded 20-30 deals. In my case, after I develop a decent EA, it may work one or two years but then it inevitably fails...In many cases, it does not work from the start." Well, who said that success means getting to the end of time with one solution? This isn't a river boat race, it's a war game. Technologies and strategies evolve. You have to evolve with them!

As for needing to be able to code to profit, I think it can only help, but to say it's necessary denies the existence of people like Rambo Moulton, and the many successes in books like 'Market Wizards' and heck yes, even some people on ForexFactory (many long gone) like Copernicus, Parisboy, 919Gilead, mr brown, hayseed, and hilmy83 who still has the highest public live trading result on the factory, as far as I know. Of course these could all be lucky outliers, but I somehow doubt it. Also Ronald Raygun, xux99, abokwaik, (though each of these can code).

It seems, instead like an attempt to explain how so many authors of articles like Mr. Ilin, can produce brilliant experiments and still manage to miss the forest for the trees. I think of people who can produce autocorrelation heat maps, and catboosted machine learning algorithms, but when they fail still think the answer must lie in neural nets. The market, being somewhat random, doesn't mean that it is a complex nut to crack that requires a correspondingly complex solution to solve. I know there's a term for this but I can't find it just now. It's almost the 'nirvana fallacy'. It means the same thing as thinking that a complex problem must need a complex solution. If anything, trading punishes overthinking. This article sheds some light on that when it describes how complicated problems are soluble with systems and processes but complex problems are trickier. The thing is I suspect trading success is neither a very complicated or complex problem, but intelligent people tend to see it that way.

On FF I think particularly of people like PipMeUp, and Mingary who complain about the random nature of forex without seeming to be able to do much about it. Both of them are mathematically clever, at least. The Talosian Trap doesn't try to explain why this happens as much as it warns that it can happen. Neural nets, advanced algo analysis, extensive modelling, none of it is necessary or sufficient for success in trading, though it doesn't preclude it by any means either.

I hope I haven't slandered anyone, and if I got some facts wrong, please accept my apologies.
  • #170
  • Aug 26, 2021 1:39am Aug 26, 2021 1:39am
Quoting clemmo17
If you have no programming skills, then you have practically no chance at all. I've seen a lot of pseudo traders bragging about some results after having traded 20-30 deals. In my case, after I develop a decent EA, it may work one or two years but then it inevitably fails...In many cases, it does not work from the start.
I was in the same trap for several years... I focused on 1 holly grail EA and never could find that - all EAs when optimized/baktested then emulated forwardtests seemed to me unstable. There might be even several years when some EA "not working" and then "again working"...

But this is a trap... if you operate in randomness and use statistics then the correct way is make portfolios of various EAs. You never know which particular EA will succeed or lose but the entire portfolio will make decent profits quite stable. There exists theories how to make portfolios for stocks. Someting similar should be applied to EAs.. for example in portfolio can be super risky asset but with high profit % and then very safe but with low profit. So you combine and make balanced portfolio.

This requires several years.. I am only at the very beginning with that idea of portfolious. Here can see my results so far. There was 50% dd - just learning, my first steps.. there were some manual trades too... not a good example but at least something that gives optimism for idea of "n EAs, portfolio, diversification".
  • #171
  • Aug 26, 2021 2:04am Aug 26, 2021 2:04am
Quoting joyny
{quote} I was in the same trap for several years... I focused on 1 holly grail EA and never could find that - all EAs when optimized/baktested then emulated forwardtests seemed to me unstable. There might be even several years when some EA "not working" and then "again working"... But this is a trap... if you operate in randomness and use statistics then the correct way is make portfolios of various EAs. You never know which particular EA will succeed or lose but the entire portfolio will make decent profits quite stable. There exists theories how...
This seems like a very sensible approach. There is no limit to the permuations you can arrange with this system too. Imagine an 'overlord' EA that monitors each EA in your stable and distributes a higher percentage of margin to the EAs that are performing best while scaling back or pausing the EAs that are doing poorly? Just the other day I tested an EA that removed itself from one chart, but not the others when it started to lose. This is such a simple idea and yet how many expensive EAs in the market bother to add such failsafes? We are very much at the start of the journey in algo trading, not the end. There are a lot of untested ideas. A type of CAPM for EA portfolios probably does exist but it would be a closely guarded secret for now.
  • #172
  • Sep 3, 2021 2:01pm Sep 3, 2021 2:01pm
In my quest to go from an incompetent coder to a merely mediocre one I've started writing out sections of the MQL5.com documentation, in my own words, longhand, the way I used to study things at school. I use a dollar store composition book and I'm on my third volume.

Attached Image (click to enlarge)
Click to Enlarge Name: IMG_8433.JPG Size: 1.3 MB
I'm not sure if it will help me memorize useful facts, but I don't think it can hurt. It forces me to take note of small details and not move too quickly through things that I think I've learned already. Barbara Oakley talks about a common failing in students who think they have learned something after a cursory reading, even when they haven't actually formed all the necessary connections. I think she calls it the 'illusion of learning'? They get to the exam and panic because when they go to recall what they think they know it's just not there.

I have learned this the hard way a few times because sometimes I think I know exactly how to write a piece of code and find myself pausing, failing to add the most elementary sections. It's very humbling. Of course, being an old fart doesn't help. Learn from my mistakes and learn how to code early in life.

Did anyone learn how to code easily and surely within a year or two without serious effort? How old were you when you started?
  • #173
  • Sep 3, 2021 10:49pm Sep 3, 2021 10:49pm
Quoting clemmo17
Did anyone learn how to code easily and surely within a year or two without serious effort? How old were you when you started?
I started coding at age 10. My first working piece of code painted house with lines. Remember how proud i was. That coding language was named BASIC. Proper naming.

It does not matter how old you are. You can learn coding much faster now as it was for me. Google and find a lot good learning stuff. But think you should start not with mql... Start with general concepts for coding which applies to any language. Use udemy.com maybe.. I have an employee who re-qualified in 2 years from non-IT to IT specialist. Doable.
  • #174
  • Sep 5, 2021 1:53am Sep 5, 2021 1:53am
Quoting joyny
{quote} I started coding at age 10. My first working piece of code painted house with lines. Remember how proud i was. That coding language was named BASIC. Proper naming. It does not matter how old you are. You can learn coding much faster now as it was for me. Google and find a lot good learning stuff. But think you should start not with mql... Start with general concepts for coding which applies to any language. Use udemy.com maybe.. I have an employee who re-qualified in 2 years from non-IT to IT specialist. Doable.
Yes I started with C/C++ and python. I took many beginner classes on Udemy/Coursera as well as some Codecademy and runestone.academy. I always get good grades in the classes I take but when it comes time to actually make something I feel lost at sea. I don't know if I will ever get completely comfortable with it. I had a chance to learn to program around 11-13 but I did not know how to learn on my own at that age. Oh well, we do what we can with the time we have left. Once I'm finished with the MQL4 book I'm going to try finishing the Ryan Young book and then go back to see if I can build my dream EA. I know it would be faster and easier to simply pay someone but I didn't start this journey to be dependent on others.
  • #175
  • Sep 5, 2021 1:58am Sep 5, 2021 1:58am
For my wife's birthday I'm going to pay off her student loans. She did most of the work herself but it's time to finish them off.
I didn't think I'd ever be able to do something like this.

Attached Image (click to enlarge)
Click to Enlarge Name: 241438975_10165602966620249_8183179913998816661_n.jpg Size: 39 KB
  • #176
  • Sep 6, 2021 3:19pm Sep 6, 2021 3:19pm
Today I tried making an illustration, something I haven't done in a long time. I used to do illustration for a living. I stopped years ago when I started a family and decided I needed to do something more stable. It's ironic that I settled on trading.

I didn't expect my drawing to turn out, since I'm so out of practice but it ended up being even worse than I expected. I also showed it to someone I respect who kindly spared me their opinion. That combination of disappointment in seeing a skill degrade, and social embarrassment made me feel bad, but my rational brain knew there was no good reason to feel bad.

What's interesting is that I noticed how I felt right afterward, and it was like a physically painful feeling, right behind or in the middle of my head (but not really a headache). I could actually feel my disappointment. It was like my body was trying to tell me, "never do that again!" Which is silly, because I knew the right thing to do was simply to try again and make a better drawing, and keep doing that until I make something I don't hate as much.

What's also interesting is that this feeling is the same one I get when I make a bad trade; a really bad one. It's like my body is trying to protect me from making bad trades - never do that again! I don't know if it's helpful in that case either.

After Googling this phenomenon I came across this article that might explain it.

Anyway, after thinking about it, I decided to just redo the drawing anyway, and it's still not great, but it's better than the first draft. I wish my brain would stop being so protective.
Here's a link to my drawing (the second draft) if you're curious.
  • #177
  • Sep 7, 2021 3:59am Sep 7, 2021 3:59am
With no disrespect to the invisible army of traders who work their socks off and post very little... I have to say you're one of the most diligent and thorough guys around here and for that alone I hope you end up storming it and discovering whatever you're looking for.

On another note, I met a daytrader that had permanent stomach issues from the stress of trading. He was on the one hand fairly successful but it seemed a heavy price to pay for financial success. Interesting you feel it in the head - I always thought everyone felt these things in the gut but I guess it must be varied.

And quality illustration btw I'd be showing that to everyone
  • #178
  • Sep 9, 2021 5:41am Sep 9, 2021 5:41am
Quoting triphop
With no disrespect to the invisible army of traders who work their socks off and post very little... I have to say you're one of the most diligent and thorough guys around here and for that alone I hope you end up storming it and discovering whatever you're looking for. On another note, I met a daytrader that had permanent stomach issues from the stress of trading. He was on the one hand fairly successful but it seemed a heavy price to pay for financial success. Interesting you feel it in the head - I always thought everyone felt these things in...
This made my day triphop, thank you!
The thing is FF won't let you edit posts after a certain period and some day I may decide I don't want to show the world that picture! lol!
  • #179
  • Sep 9, 2021 5:57am Sep 9, 2021 5:57am
Attached Image (click to enlarge)
Click to Enlarge Name: 394149b01760c79af34fd4f574040f78.png Size: 72 KB

Today the 'turnTable' system hit 10%, that's exactly three months plus a day since I began. So it is on track to hit 40% for the year. The relative and max dd remain the same, under 4%. It's good to know that it's possible to make money this safely in a smaller (<$10k) account.
  • #180
  • Sep 15, 2021 7:25pm Sep 15, 2021 7:25pm
Today I learned a lesson about swap rollovers.

At least on one of my brokers, when you trade a CFD based on an underlying instrument like oil, or US indices, and most commodities, the future that the CFD is based upon will 'rollover' at a certain date in the future. I don't trade futures and don't know much about futures trading but a rollover is when you hold a position in the instrument and the front-month contract that expires is switched to another contract that is further out in time.

Why does this matter? Well, it seems many brokers charge a special type of swap charge during this rollover period. I had never noticed before as my accounts were fairly small, but now that they're enlarging these fees are significant and it's just one more thing traders have to pay attention to. This will never be remarked upon by your broker, as it's all part of the way the house retains its advantage.

I had two positions open, and closed one, unluckily, the wrong one, not for the price change which was minimal but for the swap fee which was huge - about 10x a triple swap day fee. At first I thought my broker was up to some dirty tricks (and surely misunderstandings like this are about 25-30% of all negative broker reviews).

If I had kept both positions open I would simply have broken even as the swap was 97 points (not forex points, but futures points or are they equivalent?) in either direction, and if I had closed the other position I would have made a tidy sum. In fact the sum is so significant that you could probably strategize by opening a position before rollover and closing it shortly afterward, though usually they deduct a certain number of points in the new contract to offset this. For some reason, my broker's offset is about the same as the swap rollover in some securities but tiny in others - seems like a money-making opportunity. I can't believe it's an oversight, so if anyone knows more about how this works please enlighten me. I doubt I'm the only one who doesn't know all the details.

Attached Image (click to enlarge)
Click to Enlarge Name: d798f0a0a4c1af02a44651c3185f44a4.png Size: 39 KB

Look at the size of the Nasdaq swap rollover just for holding a long position! This seems like money in the bank. Although I'm likely overlooking something.
Attached Image

And this is the corresponding 'price fix' (what is this actually called) for the rollover. This means that if nothing occurs between today's closing and tomorrow's opening, the open price for OIL.WTI, US.100, US.30, US2000, and US500 should be lower. However for the Nasdaq and Russell 2000, as well as S&P500 the offset is miniscule compared to the swap rollover offered.

This is arbitrage income as there is practically no way price is going to move against you by so much that you won't earn an offsetting income - or does it move precisely because this is a rollover date? More research required. The only disadvantage to this strategy is you can only do it once every season.
0 traders viewing now