Talk by Dieter Hendricks on Adaptive Learning for Trading

Before QuERILab (2004)

RESEARCH PROBLEM STATEMENT (2004)

It is common to consider financial markets as made up of stocks which can be grouped or clustered according to similar characteristics, for example some stocks offer low returns with respect to improvement in their value but pay high dividends, while others offer small dividends but increase in value relatively quickly. The understanding of grouping characteristics is important for understanding the overall dynamics of the market in which such clustering takes place. The appearance of grouping or clustering behaviour may be due to random effects (noise) since there are bound to be overlapping properties when the number of stocks is high. Hence it is necessary to investigate methods for measuring possible noise contributions to clustering. Furthermore, grouping may change over time so that it is necessary to identify the time horizons for which clusters are stable. Is it possible to identify sectors, groups of stocks which display similar behaviour with respect to returns, and states, time periods for which the market behaves similarly, in SA financial data, by purely quantitative methods under the constraint that noise and temporal stability are understood? 

Thuthuka grant:  TTK2004072200035 (Funded from 2004)

TEAM in 2004

• DW = Diane Wilcox (applicant); 
• TG = Tim Gebbie (co-investigator);
• 1 Msc student [Project :Fourier Method for the Measurement of Univariate and Multivariate Volatility in the presence of High Frequency Data (graduated 2006)]
• 1 Internal project proposal reviewer (UCT Dept Mathematics & Applied Mathematics)
• 10 External reviewers (appointed by the NRF)

Other outputs by DW since 2004 are documented on http://www.dianewilcox.net/currresearch.htm

High-speed detection of emergent market clustering via an unsupervised parallel genetic algorithm

High-speed detection of emergent market clustering via an unsupervised parallel genetic algorithm

Type: Research Article
Authors: Dieter Hendricks  (contact author) Tim Gebbie Diane Wilcox
Issue: January/February 2016
Number of pages: 9
DOI: http://dx.doi.org/10.17159/sajs.2016/20140340
Published: 01 February 2016

Abstract:
We implement a master-slave parallel genetic algorithm with a bespoke log-likelihood fitness function to identify emergent clusters within price evolutions. We use graphics processing units (GPUs) to implement a parallel genetic algorithm and visualise the results using disjoint minimal spanning trees. We demonstrate that our GPU parallel genetic algorithm, implemented on a commercially available general purpose GPU, is able to recover stock clusters in sub-second speed, based on a subset of stocks in the South African market. This approach represents a pragmatic choice for low-cost, scalable parallel computing and is significantly faster than a prototype serial implementation in an optimised C-based fourth-generation programming language, although the results are not directly comparable because of compiler differences. Combined with fast online intraday correlation matrix estimation from high frequency data for cluster identification, the proposed implementation offers cost-effective, near-real-time risk assessment for financial practitioners.

Keywords:
unsupervised clustering; genetic algorithms; parallel algorithms; financial data processing; maximum likelihood clustering

QuERI Lab Graduate students at QMF2015

The Quantitative Methods in Finance 2015 Conference (QMF2015) : 15-18 December, Sydney. Australia

QMF 2015 Program Abstracts

QuERI Lab contributed talks:

1. Detecting Temporal Financial Market States Using Clustering  

      

Dieter Hendricks, University of the Witwatersrand, South Africa       

ABSTRACT:    We propose the application of a high‐speed maximum likelihood clustering algorithm to detect temporal  states in the financial market, using estimated correlation matrices from intraday market microstructure  features. We first determine the ex‐ante intraday temporal cluster configurations to identify financial  states. Next, temporal state features are studied to extract characteristic feature vectors. The latter serve  as low‐dimensional state descriptors which can be used efficiently in learning algorithms, enabling online  state detection for optimal planning in the high‐frequency trading domain.       

Authors: Dieter Hendricks,  Tim Gebbie,  Diane Wilcox    

When: Wednesday, 16 December 2015, 14h20 Room 3

2. Reconciling Order Book Resiliency and Price Impact          

Michael Harvey, University of the Witwatersrand, South Africa        

ABSTRACT:    Understanding and quantifying the impact and persistence of trade events on limit order book dynamics is  of critical importance for trading decisions. Specifically, the trading trajectory needs to be chosen to cause  least  impact  with  a  reasonable  guarantee  of  execution.  In  this  paper,  empirical  point  processes  are  extracted from intraday tick data which represent key liquidity demand and resiliency events. Using these  point processes, trades and quotes are modelled as a mutually‐exciting four‐variate Hawkes point process  with a sum‐of‐exponentials kernel. The calibrated model allows us to quantify order book resiliency in  terms of expected time frame and magnitude of quote replenishment in response to a trade event. We  conjecture that certain anomalous shape characteristics of empirical price impact curves can be explained  by measuring quote replenishment following trades which move the mid‐quote price. By examining a  period of increasing trade velocity on the Johannesburg Stock Exchange, we show that the empirically  observed  increase  in  low‐volume  price  impact  can  be  explained  by  a  lack  of  commensurate  quote  replenishment following low‐volume, price‐moving trades.       

Authors: Michael Harvey  Dieter Hendricks

When: Wednesday, 16 December 2015, 16h10, Room 3

Fintech in the future

Periklis Thivaios of IE Business School (and formerly part the AMF team at Wits) speaking at Risk Minds International 2015 in Amsterdam on how banks can manage the risks from the Fintech industry. He will be hosting a boardroom risk discussion at RiskMinds International on Thursday 10th December at 10.20 posing the question, ‘how afraid should we be of Google, Apple and Facebook and what can banks do about it?

http://www.riskmindslive.com/periklis-thivaois-on-if-banks-should-fear-google/

AMF 2015 Project Presentations 23 and 24 November

Congratulations to the AMF BSc(hons) class of 2015 for a wonderful 2 days of market-microstructure and computational finance final project presentations!

SAMS 2015

SAMS 2015 Conference at the University of the Witwatersrand 4-6 November

Special Session: Computer Science and Finance
Special Session: Computational and Mathematical Finance