sharpe ratio

[alistera]

how quantum computing helps? Any has quantum computing?

That statement is not troubling in any way, quantum is in theory but the people I know say it's already deployed, there are trade patterns that cannot be processed by supercomputers let alone normal people.

You cannot set up a fund with a Sharpe less than 3 without the fund managers going through the usual 3-5yr, 7-10yr, 15-20yr incubator duration, which is why the architecture I have access to has a Sharpe above 3, it never ceases to amaze me how the human race can take the simple and make it beyond complex.

 

[Businessman]

That statement is not troubling in any way, quantum is in theory but the people I know say it's already deployed, there are trade patterns that cannot be processed by supercomputers let alone normal people.

lol, quantum computers. Do they go into the future and get the price, before coming back to the present time.

 

[alistera]

lol, quantum computers. Do they go into the future and get the price, before coming back to the present time.

Actually you are pretty close, the public and investors are so far behind the curve that quantum computing, and quantum based methodologies, allow data to be processed in realtime via qubits with multiple states at any one time, meaning trades placed at that moment are forward projections of where retail and investors will be in minutes, hours, day, weeks, months, years ahead, that is what high Sharpe ratios provide.

 

[trend2009]

That statement is not troubling in any way, quantum is in theory but the people I know say it's already deployed, there are trade patterns that cannot be processed by supercomputers let alone normal people.

You cannot set up a fund with a Sharpe less than 3 without the fund managers going through the usual 3-5yr, 7-10yr, 15-20yr incubator duration, which is why the architecture I have access to has a Sharpe above 3, it never ceases to amaze me how the human race can take the simple and make it beyond complex.

a multiple sharpe less than 3 can combine to produce above 3

 

[JonLivingston]

especially automated trading (by us retail traders), is never that easy.

Boy if this is true.

Since I started to build/program my own automated trading system from scratch I'm wondering how long it will take me to make it fly!

 

[JonLivingston]

the markets have context built in to them to disintegrate algos indirectly and retail directly

Can't remember where or who did it but this was mentioned elsewhere on this forum...

What do you mean by that? What's "context" in this context (pun intended) and how the "disintegration mechanism actually work?

That's something it's not yet computing in my head. I'm missing pieces of the puzzle here.

 

[JonLivingston]

The other market wizards books I'd stay away from if you want a reasonable understanding of what's actually possible

Interested to know the reason behind this statement. Why do you feel that way if you don't mind me asking?

I found useful tidbits in all of them. But yes the HF one was the best.

 

[Businessman]

Boy if this is true.
Since I started to build/program my own automated trading system from scratch I'm wondering how long it will take me to make it fly!

Can take many years, you will probably go through many iterations and refinements. Im probably on version 10.

Funny thing for me is version 1,2 and 3 made a lot of money by dumb luck over the first five years i traded with automated systems. But none had good long term sharpe ratios. You can trade with a system that has a lower in reality sharpe ratio than you think it has, and get lucky and be on the right side of the bell curve for the first few years. And avoid the nasty drawdowns for a long time.

Even a system with a long term sharpie ratio of below zero: yearly returns like -30%, -5%, +20%, +5%. If you luck out and start trading it at the right time you can make a lot of money. You have been fooled by randomness but you can still make money for a while.

Back then i only used to backtested with about 3 to 5 years of data, ignorance was bliss. There was less back data for some electronic markets back then, but now every electronic market has been around for 20+ years.

So there are lots of traps and illusions involved. The most obvious one being curve fitting.

 

[alistera]

Can't remember where or who did it but this was mentioned elsewhere on this forum...

What do you mean by that? What's "context" in this context (pun intended) and how the "disintegration mechanism actually work?

That's something it's not yet computing in my head. I'm missing pieces of the puzzle here.

https://hbr.org/2014/09/contextual-intelligence

Charts are technical, context can be built in to markets, in to algorithms, indirectly, it is why algos once designed need to be updated in days, weeks, months, rarely longer as the context changes which is why no one buys algos because they will stop working, you need programmers constantly updating them but if you didn't design them you will probably miss the initial context.

The holy grail is not in a technical algo or methodology, it is in understanding the most basic context of life and the markets, once you achieve that you can build on top of it, the entire world including everyone here will try to interpret to match their own expectations and experience, only the microscopic of fraction can be transferred to someone else hence 100s post with one word or sentence that may be useful to some people and deterimental to others, the more people write the more they are trying to interpret their own understanding confusing most other senseless along the way, a pyramid from the bottom up.

The only exception is for anyone who understands the underlying context, then you can write endlessly, an inverted pyramid from the bottom up or a pyramid from the top down, the world is your oyster.

 

[JonLivingston]

Beautiful answers @Businessman and @alistera - thanks for that.

Let me reply on a few points..

Im probably on version 10.

How long from v0 to v10 if you don't mind me asking?
But yes makes sense. Especially so if one has a full time job, family, a life, etc. 24h a day are not becoming 48h any time soon..

Back then i only used to backtested with about 3 to 5 years of data, ignorance was bliss. There was less back data for some electronic markets back then, but now every electronic market has been around for 20+ years.

Again, when was "back then" if you don't mind me asking?
How long back do you go now? I've often heard conflicting opinions about this. The "test only the most recent years" vs "test since the big ben!" - that kind of debate.