Forums  > General  > What would you do?  
     
Page 1 of 1
Display using:  

TheGodPill


Total Posts: 6
Joined: Dec 2019
 
Posted: 2019-12-15 05:49
Pretend it was possible that a person without any formal or college education could spend a couple years researching, digesting, and condensing learned information to create the formula that generates infinite money by timing the swings in the indexes and capitalizing off short term profitable trades.

If you were down to your last $40,000 what would you do to guarantee your idea is fully tested vetted and proven false or right, without selling the idea out to potential thieves and or incompetent quants.

I understand you all think it's impossible, so please indulge me for a moment pretend that its not and tell me who when where why I would be best of pursuing your method to validate my strategy.

NeroTulip


Total Posts: 1062
Joined: May 2004
 
Posted: 2019-12-15 09:53
Ok I'll take the bait... At least paper-trade it until you get statistically significant returns that match your backtest. Definitely not sufficient, but at least you can rule out some of the worst errors: overfitting, forward looking bias, unrealistic execution costs and other errors in your backtest.

The real answer is you'll never know for sure until you do it. Even then, you'll never know if you've just been lucky.

"Earth: some bacteria and basic life forms, no sign of intelligent life" (Message from a type III civilization probe sent to the solar system circa 2016)

wquant


Total Posts: 3
Joined: Nov 2019
 
Posted: 2019-12-15 16:57
Maybe something like this would be useful?

https://www.winton.com/research/testing-our-strategy-for-testing-our-strategies

I'm guessing people here might have done research along these lines, does anybody have thoughts about such a process?

TheGodPill


Total Posts: 6
Joined: Dec 2019
 
Posted: 2019-12-15 18:51
I have traded 6 hours a day for over 150 days and lost over $400,000 learning, testing and hypothesizing. My strategy is impossible to test through paper trading as it requires measuring the sp500 by individual components and breaking down each second into 4 categories of data and then creating a pattern recognition formula, of patterns in the data that are only expressed during the movement of the index.

I need a team that we will guarantee me the results of my strategy as I've spent $60,000 on losers that were highly unqualified.

JTDerp


Total Posts: 62
Joined: Nov 2013
 
Posted: 2019-12-15 21:16
Without giving away anything, what factors prevent you from making a *rough* proxy to the patterns/price changes that your order signals are based on, in paper trading? Is there not an acceptable compromise that you could make?

I assume that the usual paper environment factors of (un)realistic slippage, spreads/fills, etc are what's stopping you?

"How dreadful...to be caught up in a game and have no idea of the rules." - C.S.

TheGodPill


Total Posts: 6
Joined: Dec 2019
 
Posted: 2019-12-15 21:21
No, the trades and order fill is the least of my concerns and will little to no effect on making the money. I just literally lack the programming and big data formulating education to organize the data and then the one piece that I lack the actual info behind is the pattern recognition formulas. I could tell what pattern consistencies will produce the results we want, but i couldn't te you how to a write a formula that measures the data in comparison to previous data sets.

But once the data trigger has been recognized and triggered, the logic to buy the most profitable to least profitable available orders by time to expiration vs out of money/ in the money options is the easy part.
Previous Thread :: Next Thread 
Page 1 of 1