際際滷

際際滷Share a Scribd company logo

Quantitative Modeling and Trading

Download as PPT, PDF
V
Valapet Badri

This document outlines objectives and critical success factors for a high frequency quantitative modeling and trading project in US stocks. The goals are to achieve optimal time frame modeling to generate revenue. Key factors are availability of tools, liquid stock tick data, and software. The approach is to establish an environment, explore models combining industry and academic approaches, and generate revenue. Examples include Engle models, Bayesian inference, particle filtering, and microstructure applications. Next steps are to set up the environment, explore promising models, and use Engle and other approaches.

1 of 7
Download to read offline
- Valapet Badri 1
Objectives High frequency quantitative modeling and trading in US stocks(NYSE and NASDAQ Markets) Achieve this in an optimal time frame and generate revenue stream for the firm 2
Critical Success Factors Availability of tools and information in proper format The most liquid stocks tick information would be available initially as a snapshot Sufficient infrastructure and Statistical software (S- plus /Matlab) are available to perform quant. Analysis. My goal initially would be to focus on establishing a basic environment Minimal amount of time to be spent on pure IT allied endeavors although they are important for overall success of the project 3
Facts & AdvantagesThere are other Hi-frequency hedge funds who have been in operation for 8+ years and are ahead in terms of experience and models for quant analysis and trading strategies There are no standard textbooks for this areas that provide a guide to success How to take advantage and address gaps? Develop High Frequency model that utilizes recent developments in stock market Existing time tested quant principles and recent academic developments Get the right mix of industry models and academic models and generate revenue within optimal time frame 4
Example approaches Models under the framework of Engle (2000) such as Exponential ACD model, UHF-GARCH and more. Bayesian Inference via Filtering for Model Continuous-time Likelihood, Posterior, and Bayes Factors Filtering Equations and Evolution Equations for Bayes Factors Two Computational Approaches and their Consistency The Markov Chain Approx. Method and nearly likelihood etc. Particle Filtering (or Sequential Monte Carlo) A Micro movement Model Random-arrival-time State Space Model More Financial and Market Microstructure Applications (Frequentist approaches interpolation, averaging, Splines) 5
Next steps Establishing a basic environment and acquire necessary tools and infrastructure(details will be provided) Explore promising approaches suggested to come up with a quantitative model and transformation of the tick level information. Follow current industry trend and use a mix of Engle(Rob Engle- Nobel laureate) type of time series analysis and other approaches to generate revenue 6
Next steps Establishing a basic environment and acquire necessary tools and infrastructure(details will be provided) Explore promising approaches suggested to come up with a quantitative model and transformation of the tick level information. Follow current industry trend and use a mix of Engle(Rob Engle- Nobel laureate) type of time series analysis and other approaches to generate revenue 6

More Related Content

Quantitative Modeling and Trading

  • 2. Objectives High frequency quantitative modeling and trading in US stocks(NYSE and NASDAQ Markets) Achieve this in an optimal time frame and generate revenue stream for the firm 2
  • 3. Critical Success Factors Availability of tools and information in proper format The most liquid stocks tick information would be available initially as a snapshot Sufficient infrastructure and Statistical software (S- plus /Matlab) are available to perform quant. Analysis. My goal initially would be to focus on establishing a basic environment Minimal amount of time to be spent on pure IT allied endeavors although they are important for overall success of the project 3
  • 4. Facts & AdvantagesThere are other Hi-frequency hedge funds who have been in operation for 8+ years and are ahead in terms of experience and models for quant analysis and trading strategies There are no standard textbooks for this areas that provide a guide to success How to take advantage and address gaps? Develop High Frequency model that utilizes recent developments in stock market Existing time tested quant principles and recent academic developments Get the right mix of industry models and academic models and generate revenue within optimal time frame 4
  • 5. Example approaches Models under the framework of Engle (2000) such as Exponential ACD model, UHF-GARCH and more. Bayesian Inference via Filtering for Model Continuous-time Likelihood, Posterior, and Bayes Factors Filtering Equations and Evolution Equations for Bayes Factors Two Computational Approaches and their Consistency The Markov Chain Approx. Method and nearly likelihood etc. Particle Filtering (or Sequential Monte Carlo) A Micro movement Model Random-arrival-time State Space Model More Financial and Market Microstructure Applications (Frequentist approaches interpolation, averaging, Splines) 5
  • 6. Next steps Establishing a basic environment and acquire necessary tools and infrastructure(details will be provided) Explore promising approaches suggested to come up with a quantitative model and transformation of the tick level information. Follow current industry trend and use a mix of Engle(Rob Engle- Nobel laureate) type of time series analysis and other approaches to generate revenue 6
  • 7. Next steps Establishing a basic environment and acquire necessary tools and infrastructure(details will be provided) Explore promising approaches suggested to come up with a quantitative model and transformation of the tick level information. Follow current industry trend and use a mix of Engle(Rob Engle- Nobel laureate) type of time series analysis and other approaches to generate revenue 6