Registration for the Real Accounts (Cents) Championship July 2017 . - page 15

 
Server Muradasilov:
Sixteen people already registered, registration is on-going

REGISTRATION IS ONGOING!

WAITING FOR NEW PARTICIPANTS

FRIENDS, PLEASE SIGN UP FOR THE CAR-TRADING CHAMPIONSHIP!

 
Roman Shiredchenko:


Who were you talking to? :-)

Only after P.S. Already figured it out when I finished reading the branch to this page. IMHO, it's complete nonsense for this one.

P.P.S. Everything is stolen before you. See how winners are determined in other contests... for example our regular one starting in September until the end of this year on the exchange...


bullshit is bullshit...

But only this nonsense, i.e. the Sharpe coefficient, is taken to determine the winners in the monthly competition. And that's bullshit squared.

.

Can you answer the first question posed?

Whatdoes the Sharpe Ratio measure?

And so on down the list:

Forum on trading, automated trading systems and trading strategy testing

Registration for the Real Accounts (Cents) Championship July 2017 .

Elena Kukharets, 2017.05.15 01:57


Can you give clear and concise answers to the questions:

What does the Sharpe Ratio measure?

For what purposes is it used?

What surprises can it provide?

At what time interval do its readings become meaningful rather than being noise with no meaning?

Is it even appropriate for a one-month competition?


The formula for calculating the Sharpe Ratio and defining its constituent values is here.


 

I'm even curious as to who pulled this Sharpe ratio into the defining indicators of the monthly competition? This person just doesn't get the point, doesn't understand where, how, when and why to stick this Sharpe ratio.

According to the terms of the contest: there must be at least ten (10) deals. And that's fine for a monthly contest.

But Sharpe's coefficient for its calculation requires defining a standard deviation, and this is from the statistics section, if you don't know. So, to operate with statistical values we need at least 30 - 50 - 100 (thirty - fifty - one hundred) points, and the more the better. If the points are less, the statistical approach is inapplicable, and attempts to apply it in such a case will show nonsense. This is a well-known and generally accepted statistical fact. (and here it is apparently unknown or unacknowledged... But rather a thought about it did not even arise, by virtue of, so to speak, "statistical illiteracy" and thoughtless lumping of anything).

It is the nonsense that will be issued by this Sharpe on contest accounts with 10 - 20 - 30 trades, --- quite a normal number for a monthly contest --- and quite normal, good accounts will be thrown out of consideration by this Sharpe into the dustbin.

But that's not all of Sharpe's wonders either.

For example, even with a sufficiently large number of trades, an account with good large exponential growth and no losing trades by Sharpe will be worse than an account with paltry profits (with profitable and losing trades) and a standard deviation close to zero.

This can all be seen from the formula for calculating this miracle Sharpe ratio. But to see it, one has to understand what is behind the letters that are present in the formula for the calculation (and with this, apparently, it is difficult...).


zy.

I can demonstrate the "surprises" of this miracle Sharpe.

 
Roman Shiredchenko:


Oooh! Glad to read you.

You write intelligently...

Mostly... :-)

Speaking of "mostly" - people here have options for determining the "WINNERS" in the contest...

but not monitoring signals...

Ahhhhhh, I see :-)
 
Олег avtomat:


I can clearly demonstrate the "surprises" of this miracle-Sharp.


As for me, this indicator is easily manipulated.

For example, any averager taking a small, but rigidly fixed amount of profit will have an exorbitant Sharpe figure, an order of magnitude higher than those who trade on the plus side and competently translate trades into b/w (getting quite a big spread)

But in the final formula, it doesn't have too much impact.


Also, the drawdown rate taken into account in the contest is easily manipulated.

I did not manage to show it correctly in May (within rules), but there was an idea. I closed my profit on the first trade after getting a small drawdown of 0.8% and withdrawing equity of +20% of growth and traded with a larger drawdown, but it was not visible, as the profit was not fixed.

 
Igor Volodin:


The way I see it, this indicator is easily manipulated.

For example, any averager taking a small but rigidly fixed amount of profit will have an exorbitant Sharpe figure, an order of magnitude higher than those who trade on the plus side and competently convert trades to b/w (getting quite a big spread)

But in the final formula, it doesn't have too much impact.


Also, the drawdown rate taken into account in the competition, is easily manipulated.

I did not manage to show it correctly in May (within rules), but it was my idea. I closed my profit on the first trade after getting a small drawdown of 0.8% and withdrawing equity with +20% of growth and traded with a larger drawdown, however it was not visible, as the profit was not fixed.


I would make two nominations - and 3 prizes in each, for a total of 6 prizes. The more people get certificates and pennants the better. It will be something to remember when you are old)))) Then the grandchildren will rummage in the drawer - Oh, shit! Grandfather was a trader! Ahahahah))))

The first nomination - Absolute, the calculation is stupidly balance - who has a higher and the winner no matter what the risks and drawdowns.

The second nomination - The best risk-management, the calculation is made by balance minus drawdown (but which to take - the maxi, absolute or relative?)

for example:

the first participant earned $100 at a drawdown of 30%, minus the drawdown of $ 100 - ($ 100 * 30/100), remaining an honest $ 70

The second participant earned $ 80 with a drawdown of 10%, subtract the drawdown of $ 80 - ($ 80 * 10/100), remaining an honest $ 72,


This is the most fair and simple method, especially given the rules of the contest need at least ten deals! Why do you calculate some kind of sharps by sinusoids)))) Who needs it?

Decipher to me please, what are these drawdowns in terminal reports such - why they are 3 kinds at once?

Absolute drawdown 9.88 Maximum drawdown 27.66 (14.23%) Relative drawdown 16.16% (17.37)

 
Олег avtomat:

I'm even curious as to who pulled this Sharpe ratio into the defining indicators of the monthly competition? This person just doesn't get the point, doesn't understand where, how, when and why to stick this Sharpe ratio.

According to the terms of the contest: there must be at least ten (10) deals. And that's fine for a monthly contest.

But Sharpe's coefficient for its calculation requires defining a standard deviation, and this is from the statistics section, if you don't know. So, to operate with statistical values we need at least 30 - 50 - 100 (thirty - fifty - one hundred) points, and the more the better. If the points are less, the statistical approach is inapplicable, and attempts to apply it in such a case will show nonsense. This is a well-known and generally accepted fact in statistics. (and here it is apparently unknown or unacknowledged... But rather a thought about it did not even arise, by virtue of, so to speak, "statistical illiteracy" and thoughtless lumping of anything).

It is the nonsense that will be issued by this Sharpe on contest accounts with 10 - 20 - 30 trades, --- quite a normal number for a monthly contest --- and quite normal, good accounts will be thrown into the trash by this Sharpe from consideration.

But that's not all of Sharpe's wonders either.

For example, even with a sufficiently large number of trades, an account with good large exponential growth and no losing trades by Sharpe will be worse than an account with meager profits (with profitable and losing trades) and standard deviation close to zero.

This can all be seen from the formula for calculating this miracle Sharpe ratio. But to see it, one has to understand what is behind the letters that are present in the formula for the calculation (and with this, apparently, it is difficult...).


zy.

I can visually demonstrate the "surprises" of this miracle-Sharp.

Igor Volodin:


The way I see it, this indicator is easily manipulated.

For example, any averaging operator that takes a small, but rigidly fixed amount of profit will have an exorbitant Sharpe ratio, an order of magnitude higher than those who trade on the plus side and competently translate trades into b/w (getting quite a wide spread).

But in the final formula, it doesn't have too much impact.


Also, the drawdown rate taken into account in the competition, is easily manipulated.

I did not manage to show it correctly in May (within rules), but there was an idea. I closed my profit on the first trade with a small drawdown of 0.8% and fixed equity with +20% of growth and traded with a larger drawdown.

Do you have any examples?

I don't remember who suggested k.sharp as part of the formula, but personally I don't think it's important/necessary. I would replace it with a recovery factor.

 
Andrey Dik:

Any examples?

I don't remember who suggested the Sharpe coefficient as part of the formula, but personally I don't see it as important/necessary. I would replace it with a recovery factor.

hmmm... Isn't one look at the Sharpe formula enough to see all of the above? An "optimiser" should be able to see it without saying too much... or is it?

Want some examples? Sure. Lots of them.

Especially for that I'll open a separate thread where I'll show behavior and absurd results of Sharpe factor on a large number of examples, but again, you need to understand for what purposes it is intended.
(I'll do it over the weekend).

In the meantime, give an answer to the question:Whatdoes the Sharpe Ratio measure?

 

me write mt5

I still don't get it, what's on the shoulder?

500?

Is the account currency dollar?

 
Oleg Tsarkov:

me write mt5

I still don't get it, what's on the shoulder?

500?

Is the account currency dollar?


No, 1:100 for all
Reason: