You Name It... The Models Time It..

T

tradingjournals

In this thread, I want to test some timing methods (direction and entry). You can post the name of an instrument you want the models to use for testing. This journal may introduce an innovation in journaling (the use of crypotography). If you like to guess, what is the message associated with this key:

24b1f4f2562b28ec9e7b07e7f8c89b7bf84b2b86f000ec4171e4384717a3b13550981910eabba9c161a373954d35bba82c7e42e84bc2cb7c610acb73aef1d2a2

If you break it, I would like to know it.
 
take the spx, ndx,djt and dow and add in xlf,goog and appl track them together and lets see what you get..after wards, find a correlation with bonds,the nikkie and the yen and dollar
 
Quote from ammo:

take the spx, ndx,djt and dow and add in xlf,goog and appl track them together and lets see what you get..after wards, find a correlation with bonds,the nikkie and the yen and dollar
More...

Could you give an example from the history of these instruments of what you want as output?
 
there are hundreds of formulated indicators when applied to single instruments they sometimes work, sometimes early, late .sometimes useless


cant read your code and i only guess at what will happen, would like to see you stick these instruments into your performance models,build a collective history and relationship and see how they predict turns, continuations, the spx is 500 stocks , dow 30 , transports 20,oex 100, why isnt there a collective index watcher
 
Quote from tradingjournals:

In this thread, I want to test some timing methods (direction and entry). You can post the name of an instrument you want the models to use for testing. This journal may introduce an innovation in journaling (the use of crypotography). If you like to guess, what is the message associated with this key:

24b1f4f2562b28ec9e7b07e7f8c89b7bf84b2b86f000ec4171e4384717a3b13550981910eabba9c161a373954d35bba82c7e42e84bc2cb7c610acb73aef1d2a2

If you break it, I would like to know it.
More...

The messages number is probably your hexadecimal bit coin wallet number.
 
Quote from tradingjournals:

In this thread, I want to test some timing methods (direction and entry). You can post the name of an instrument you want the models to use for testing. This journal may introduce an innovation in journaling (the use of crypotography). If you like to guess, what is the message associated with this key:

24b1f4f2562b28ec9e7b07e7f8c89b7bf84b2b86f000ec4171e4384717a3b13550981910eabba9c161a373954d35bba82c7e42e84bc2cb7c610acb73aef1d2a2

If you break it, I would like to know it.
More...

228353553953677500000

And no I won't be depositing any money into your wallet.
 
Quote from bwolinsky:

The messages number is probably your hexadecimal bit coin wallet number.
More...

:) . No, it is not. It is the hash(sha) of this message:

I like this thread! I wish I could decode the messages, but at least I tried!
 
Quote from tradingjournals:

SPY @ 181.38. Short signal from model xx.
More...

It is 110 minutes since the signal came. SPY now at 180.91. The reward part is $0.47. Is this too little, too much, just fine? Some analysis would need to be done.

Before I do, I want to write some (random) thoughts:

1. The historical arithmetic return on SP is ~ 12.5% with a standard deviation of ~ 22.5%. The mean/STD ratio is therefore 0.55. I believe this is an important number to keep in mind. If one cannot have a long side trading system that on a yearly basis does not get a return higher than 12.5% with a ratio of 0.55, and if the sole aim of the system is to trade it alone and to beat the market, the merits of it are limited at best. If it is part of other systems, then there might be merits but these could be marginal and not easy to explain to others.

2. What some may not realize is that the short side is a lot tougher than the long side, as one would have to fight the positive mean in the market, obtain a positive mean for the system that justifies its trading, and do so at a ratio that is higher than 0.55 (as a minimum) if one's aim is to only trade on short side and trade this system only. In such conditions, such short system would need to essentially make 25% (because one would have to fight the positive mean of SP) and get the positive results at a normalized standard deviation of 22.5%. Short side systems are however useful even at a reduced mean, because of their negative correlation to the market, so they add return and reduce variance.

3. Current SPY volty is 0.75%. With 390 minutes of trading and a current spy at 181, this translates to 1.81*0.75/SQRT(390) volty per minute, which is around 6.8 cents per minute.

2:46PM: SPY at 180.75 area. This is a point where a system would suggest some profit taking. I did not analyze things in detail, but from a volume point of view, there would be trapped early morning bears at 180.74 who would be waiting to get out. I also think the market is now in a mood of shorting the rallies and I expect sellers at black line and above for rest of day.
 
The 12.5% long term return is about ~ 100*(12.5/252) (non compounded) daily basis points. In absolute terms, with SPY at 181, it corresponds to 8.9cents. The market is therefore rising at rate of 8.9cents per day, with volty of 136 cents per day. A long term investor goes through a daily mean to volty ratio that may be better felt if the inverse is reported instead: 136/8.9 ~ 15.28.

It is as if the investor takes 15 unrealized losses in a day to gain 1 dollar in year. This illustrates the healing part of time, and how steady the daily gains can catches up on the losses from the violent daily volty.

It is also important to note that the volty used in the calculations is the one when things are going well. If market goes down, the volty could double, and the ratio above is then ~ 30.

Please let me know if you see typos (I am sure many), errors in numbers, clarifications to be made, and suggestions for additions and improvements.
 
You must log in or register to reply here.
Back
Top