fbpx

Quant analytics: What is the best causal filter? Wavelets? Mark Jurik’s JMA? …?

(Last Updated On: April 12, 2011)

What is the best causal filter? Wavelets? Mark Jurik’s JMA? …?

Hi, I’ve been doing some research on wavelets and they seem to show promise to filter intra-day fx time series because they can deal with non-periodicity (contrary to Fourier-based techniques). What is your experience with them? Are there better causal filters available?

 

I’m glad to see that you are open-minded. First, Cronus is an astrophysical measure and not astrological. Astrology assigns particular relationships for aspects, such as with the Bradley measure. This is one reason why the Bradley measure fails so badly. Also, Cronus includes multi-galactic information. I care less what people think about astrophysical relationships because their ignorance is what creates the information asymmetry for them to be the counterparty of my winning position. Through discussions with various folks, I have learned that the degree of open-mindedness is a trait that separates those with 7 figure and 9 figure personal net worths. Usually, folks with net worths close to a billion ask different type of questions than those with only a few million. Billionaires ask questions about what is to occur 10-30 years into the future, while someone with only a few million tend to ask questions about the next 1-5 years. I am a simple person, and I have been blessed to be around generous people that have been patient with my limitations, especially my teachers.
I am not here to convince folks of the validity and efficacy of astrophysical methods. I let my ACE letters do my talking. The ACE letters are private, noncommercial, and uncirculated letters that I write to validate my bold statements, such as “Cronus allows me to see the future of any market and much more”. For example, I wrote in the 5/31/10 ACE letter, “The end of this weekly down trend will be realized by the end of June 2010 and no later than the first week of July 2010. Also, the Master Time Factor principal of alternation is signaling a Low near the end of June 2010.” In the 7/11/10 update, I stated, “The MTF principal of alternation is an idea that is applicable in ascertaining the DIRECTION of a short-term trend that will occur years into the future with a probability at least 85%. The MTF is also applicable in determining extremal events too.” When I say years, I am talking about a decade. In the 5/31/11 letter, I stated, “Cronus8 crosses below the centerline on 6/3/10. Cronus6 forms a bottom below the centerline (without going above the centerline from the previous bottom below the centerline). These combined characteristics next week will signal the strengthening of
the weekly down trend.” Also, I discussed an interesting timeframe in the 5/31/10 letter, “Cronus6 forms a bottom below the centerline near the end of the month, going from strictly decreasing to increasing on 6/18/10. Cronus8 forms a bottom below the centerline on 6/18/10. DlyWklyCronus forms a bottom below the centerline on 6/19/10.” In the 7/11/10 letter, I wrote, “The Spx printed a high of 1121.01 on 6/18/10 and printed a higher high of 1131.23 on the following trading day (6/21/10). After the 6/21/10 high, the Spx printed a low of 1010.91 on 7/1/10.” Also, this timeframe corresponded with an intense X-Ray burst reaching Earth from a supernova that occurred over 5 billion years ago. How did I know that an event that occurred over 5 billion years ago would reach Earth at this time? Cronus. This is the deterministic part of this universe, which makes seeing over 5 billion years into the future as easy as seeing over 5 billion years into the past. Mike, if you are a member of Mark’s Yahoo Group, you can see the Cronus postings, although my intuition is telling me that you have already seen these statements. Anyways, Mark Jurik is patient and very generous with his time. For the disbelievers, I say, “Thank you for your donations”.

.Part2…… Why does Cronus work you ask? Energy is the answer. Cronus is an energy measure. Here is a simple prediction. I claim that very few folks will stand chest deep in the waters of Lake Michigan for 5 hours straight during January 2012. Also, I claim that outdoor plant growth in Michigan, USA is significantly less (1% significance level) during January than July. Clearly, we know that the sun’s rays hit the Earth at a more obtuse angle in the Northern hemisphere during January than July; thus, less energy reaches the Northern hemisphere. Third, I claim that plant growth is significantly more during when the large, yellow ball is in the sky versus when the small, white ball is in the sky. Heck, I can even predict when the large, yellow ball will be visible years into the future. The same idea applies to Cronus. These are more examples of the deterministic part of this universe. Basically, I am addressing the relationships between energy changes and the response in question. The how is intellectual property. I needed 7 years to go from an intraday measure to a weekly measure. Since I have no interest in a Nobel Economics Prize, I did not disclose the Cronus logic at GT, although I am interested in a Nobel Peace Prize. Hopefully, I have not bored you too much, and you can learn more about Cronus and system building at Mark’s Yahoo Group; otherwise, just give me a yell. Thank you

as I am not an expert with years of experience, I just can’t judge the idea and value of the system You described. However, for now, I’d agree with Michael Holder and I’d state that this idea covers “hard to define area

I am guessing the limits of an interest from Quant finance members in astrophysical methods, might hinge upon whether this approach can capture wealth from a direct application to financial time series, but beyond that – whether there is an ability to contexualize or classify the method within some known stream of mathematical endeavor that is applicable to finance. Additionally, important questions would include whether the process is or can be made”algorithmic”, or can be fully automated – based on algorithms – towards the systematic production of a consistent risk-adjusted profit. If instead the mathematical basis of the method is unclassified (or anecdotal) and simply provides an intuitive or discretionary inspiration for human forecasts that are then manually executed in financial markets – then I suspect appropriateness as part of quant finances group is far less. That doesn’t mean the process isn’t interesting to some and profitable to a (very) few.
I apologize if my original queston re Astro methods took the “filter” discussion way off subject – I hope my comments fall into the category of “nine figure” questions – although I used to be only 6 figures using Jurik type methods, and these days I am zero figures trying to “figure” out what still works in these complex markets. go figure.

I don’t think you should be using causal filters. Wrong approach. You strategy should be based on equations which model the dynamics of market participants

I would agree that putting casual filters in to the model won’t give me the correct outcome for market predicting or tracking. I am wondering if the theory of digital filters (FIR) could be applied to deal with nonlinearities in markets. The adaptive filters could be interesting idea for selecting the correct period and filtering the noise from tick data too

My approach would be to develop an underlying system dynamic model. Observations (quotes) would then be used to update the model. FIR filters may serve to reduce uncertainty associated with combining a sequence of observations provide an assessment of market trends but you don’t know that until you have a model.

Let me put it this way. Suppose you have entered a long position and sometime after that the price of the stock starts to drop for a few bars. I think you would be hard pressed to explain a decision to exit the position or not based on what is known about FIR filters. The only way you can justify your decision is with a model. Create a model first and then figure out what you need to update the model in discrete time using observations.

 

From a Linked In group discussion

 

NOTE I now post my TRADING ALERTS into my personal FACEBOOK ACCOUNT and TWITTER. Don't worry as I don't post stupid cat videos or what I eat!

Subscribe For Latest Updates

Sign up to best of business news, informed analysis and opinions on what matters to you.
Invalid email address
We promise not to spam you. You can unsubscribe at any time.
Scroll to Top