Math/Stats theories sharing

I see under location you have KCMO ?

I assume that is Kansas City currently with a temp of 23 degrees and a humidity of 73%.


You are either having a late night or an early morning ?

Rock on Kansas

Having looked at many systems without much success I feel some answers may lie in the area of Bayes.
 
Quote from Humpy:

I see under location you have KCMO ?

I assume that is Kansas City currently with a temp of 23 degrees and a humidity of 73%.


You are either having a late night or an early morning ?

Rock on Kansas

Having looked at many systems without much success I feel some answers may lie in the area of Bayes.
More...

Here is dilemma I am dealing with.

Take look crude oil WTI (CL) daily ATR 5 period. Should be 1.62 using July contract. As percent of current underlying price about 1.7 percent

Take Nat Gas (NG) daily ATR 5 period. Should be .104 using July as well. As percent of current underlying price about 2.5 percent.


Method I use works well in CL not in NG any longer since nat gas came down in price below say 4.00- is ATR any measure which to consider why not working ? Absolute price changed method went bad, however volatility fron 2010- back to 05' was real high in NG

What I see really is one contract CL per lot equals 90k in leverage, and other NG equals rough 40k in commodity, does the actual value of the contact need to be consideration because one is double the other in leverage, margins are rough intraday 1k for CL and 500 for NG
 
Quote from bwolinsky:

You wouldn't trade a losing system, doesn't this debate assume that it is profitable in terms of producing at least somewhat above 60% monthly win percentages?
More...

winning% can be not an issue. You can have a 35% winning chance system but as long as you win big lose small, you can be rich. You can even have an option system that some of yours total lose while just 1/10 of your option wins big then you can be rich too.
 
Quote from j2ee:

winning% can be not an issue. You can have a 35% winning chance system but as long as you win big lose small, you can be rich. You can even have an option system that some of yours total lose while just 1/10 of your option wins big then you can be rich too.
More...

I'm just trying to make the criterion as broad as possible, around 60% which could be any type of system capable of those kind of monthly win percentages so it wasn't like I'd chosen any particular style just with that sort of end term result of some significance.
 
Quote from trilogic:

Here is dilemma I am dealing with.

Take look crude oil WTI (CL) daily ATR 5 period. Should be 1.62 using July contract. As percent of current underlying price about 1.7 percent

Take Nat Gas (NG) daily ATR 5 period. Should be .104 using July as well. As percent of current underlying price about 2.5 percent.


Method I use works well in CL not in NG any longer since nat gas came down in price below say 4.00- is ATR any measure which to consider why not working ? Absolute price changed method went bad, however volatility fron 2010- back to 05' was real high in NG

What I see really is one contract CL per lot equals 90k in leverage, and other NG equals rough 40k in commodity, does the actual value of the contact need to be consideration because one is double the other in leverage, margins are rough intraday 1k for CL and 500 for NG
More...

Trading size is not so granular unless you're speaking of many tens of contracts.

The method I would say I've found best at researching trading algorithms long term has been to use a fixed position size model that increases according to the desired leverage factor and annualized rate after resetting each account to the exact fixed position size you want to have each time you add "units" on top of a "base unit."

So I only have to worry about the model's base unit trades and all the other "units" are based on decisions I don't have any input on unless we're talking about how many to allocate to my own proprietary trading.
 
Quote from bwolinsky:

Trading size is not so granular unless you're speaking of many tens of contracts.

The method I would say I've found best at researching trading algorithms long term has been to use a fixed position size model that increases according to the desired leverage factor and annualized rate after resetting each account to the exact fixed position size you want to have each time you add "units" on top of a "base unit."

So I only have to worry about the model's base unit trades and all the other "units" are based on decisions I don't have any input on unless we're talking about how many to allocate to my own proprietary trading.
More...

I am was using just one original lot and not talking about "pressing" position with size.

the original size began to fail in NG relative to CL- the real quetion is has NG become less volatile than CL over the ladt wo years, I think yes but I want to prove this
 
You must log in or register to reply here.
Back
Top