Machine learning in trading: theory, models, practice and algo-trading - page 3263
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
I think that's exactly the point.
They must be statistically profitable....
They may not be profitable on the test period, but due to the fact that there are many strategies and they are statistically profitable, the law of large numbers will work and the total equity will be in +.
Why then is the trading systems league not working very well ?
Because the idea is different there and the realisation itself is wildly bushy
There is no selection of quality TS, tests for durability, etc.
There is just a constant selection of TS that worked in the recent past (which does not say anything).
There is a constant carousel of pseudo profitable TSs.
I have proven TSs and they are not changed and not replaced, at least not yet.
Because the idea is different and the implementation itself is wildly artisanal
There is no selection of quality TCs, durability tests, etc.
There is just a constant selection of TCs that worked in the recent past (which does not say anything).
There is a constant merry-go-round of pseudo profitable TCs.
And I have proven TS and they are not changed and not replaced, at least not yet.
All that you write is a banal ideology of portfolio, about which everything is known since Markowitz (1980).
Briefly in your case, when trading a portfolio, the profit/loss is added up, but the drawdown of the portfolio is not more than the drawdown of the most drawdown TS. Usually the drawdown is less up to 50% of the worst TS.
Everything you write is banal portfolio ideology, about which everything is known since Markowitz (1980).
Briefly, in your case, when trading a portfolio, the profit/loss is added up, but the portfolio drawdown is not more than the drawdown of the worst drawdown TS. Usually the drawdown is less up to 50% of the worst TS.
yes
Only there is a portfolio of "buy-and-hold" stocks, and I have a portfolio of strategies....
By the way, you can also rebalance the portfolio of strategies, it will be a step towards similarity with the "league of trading systems" (or whatever it is called).
Just because each individual strategy is aspirational doesn't mean it's sustainable
Nobody said that...
Just as the non-strength of a strategy does not indicate that it is sustainable, but rather that it is over-learning....
Now the question is: Which strategy that works on new data is easier to get? A strategy that is robust but works or a strategy that is not robust and works?
=================
In the proposed approach I change the requirements to the TS (change the target function) , make them softer, simpler, more accessible ...
And why should they break if there is a requirement of stability (in fact, this is the main requirement).
NOT beautiful equity, NOT minimum drawdowns, NOT maximum profit, etc... just a stable, statistically significant result.
And why should they break at the same time if they are different, respectively not correlated, also remember about the requirement of"stability" which must be fulfilled without fail, otherwise everything else is meaningless.
from the above it is clear that this statement is not true.
Exactly!!!
Logic should be practised, not simulated ). Strong models are just built on the principle you are now preaching.
But it does not cancel the fact that it is interesting to try )