UK PhD Centre for Financial Computing Prof. Philip Treleaven Department of Computer Science University College London Gower Street London WC1E 6BT T: 0207679 7288 E: p.treleaven@ucl.ac.uk
Presentation New UK PhD Centre for Financial Computing Good news - major interest and activity in the banks in highperformance computing. Doctoral Training Centre overview Computational Finance Algorithmic Trading High-Performance computing
Doctoral Training Centre University College London (UCL) in collaboration with the London School of Economics (LSE) and the London Business School (LBS) and twenty leading financial institutions have been award 7m ($10m) by the UK Government to establish a national PhD training centre in financial computing. Financial computing is defined as covering financial IT, computational finance and financial engineering: This UK Government funding will allow us to start at least 10 UK/EU PhD students each year for five years, plus additional industry-funded and self-funded foreign PhD students.
DTC Background The Research Councils are targeting Financial Services and Retail as new collaborating industries Three previous initiatives have failed. UCL Computer Science has very strong links with the major banks: Undergraduate course (Goldman Sachs, Deutsche Bank, Reuters) Masters Programme (Goldmans, Credit Suisse, Merrill Lynch, Morgan Stanley) MSc Software Engineering group project with Citigroup Research projects in Algorithmic trading, Risk with various banks Research projects with Hedge Funds Virtual Trading Lab. with Thomson Reuters (Reuters 3000 Xtra) New PhD Centre with 20 financial institutions
DTC Background The Research Councils are targeting Financial Services and Retail as new collaborating industries Three previous initiatives have failed. UCL Computer Science has very strong links with the major banks: Undergraduate course (Goldman Sachs, Deutsche Bank, Reuters) Masters Programme (Goldmans, Credit Suisse, Merrill Lynch, Morgan Stanley) MSc Software Engineering group project with Citigroup Research projects in Algorithmic trading, Risk with various banks Research projects with Hedge Funds Virtual Trading Lab. with Thomson Reuters (Reuters 3000 Xtra) New PhD Centre with 20 financial institutions We would like to facilitate links between UK banks and UK universities We believe the banks offer unique large-scale experimental research environments
PhD Programme The PhD programme is four years with the student following a Masters programme in the first year and then working on an applied research project with one of our financial industry partners during year's 2-4. The Financial Services industry want PhD students who are stellar in three areas: analytics, programming and finance. Therefore each student will have a bespoke set of masters courses drawn from UCL, LSE and LBS. Any department at the three institutions can participate: Computer Science, Mathematics, Statistics, Physics, Economics, Finance Each PhD student will have an Academic Supervisor and an Industry Advisor. Each student will be expected to undertake a significant period at one of our partner financial services companies, working with their Industry Advisor.
FinancialComputing.org
Financial IT Software Engineering Communications and Networks Human Computer Interface Operations Management
Computational Finance No exact definition Financial Modelling - the most general of the related terms, covers computation of finance problems, such as option pricing, with the central aim of modelling valuation under uncertainty. Mathematical Finance - is the branch of applied mathematics concerned with the financial markets. Financial Engineering focuses on financial innovation, which aims to produce new securities, specifically derivative for the options and futures markets. Computational Finance is a cross-disciplinary field that focuses on the financial services industry and relies on mathematical finance, numerical methods and computer simulations to make trading, hedging and investment decisions, as well as facilitating the risk management of those decisions.
Computational Finance II
Computational Statistics and Machine Learning Artificial Intelligence: Supervised Learning - Regression Trees, Discriminant Function Analysis. Unsupervised Learning - Neural Networks, Self-Organising Maps (SOM), Principal Components Analysis. Reinforcement Learning - q-learning, temporal difference learning, Neuro-dynamic programming. Computational Statistics: Probability Density Estimation Parameter Estimation (Statistical Inference) Bayesian inference
Electronic Trading Algorithmic trading is the use of computer programs to automate one or more stages of the trading process: pre-trade analysis, trading signal, trade decision and trade execution. Electronic Trading broadly electronic trading is any method of electronically trading securities (stocks, bonds), foreign exchange (FX) and derivatives (options, futures). Order Management Systems an order management system is a softwarebased platform that facilitates and manages the order execution of securities, typically through the Financial Information exchange (FIX) protocol. Program Trading program trading is informally defined as the use of computers in stock markets to engage in arbitrage and portfolio trading strategies. Automated Trading an automated trading systems (ATS) is a computer trading program that automatically submits trades to an exchange (Instinet).
Algorithmic Trading
High-Performance Computing When signing up members for the DTC, the most common question was are you doing any work on high-performance computing?. Multi-computer systems - multiple computers that work on the same task. Grids several computers to a single problem at the same time. Clusters group of linked computers, working closely together possibly linked by a local-area network. MPPs - massive parallel processing (MPP) refers to a computer system with many independent arithmetic units or entire microprocessors. Multi-core processor (or chip-level multiprocessor, CMP) - combines two or more independent cores (normally a CPU) into a single package composed of a single integrated circuit (IC), called a die, or more dies packaged together. Accelerators - specialized parallel computer architectures are sometimes used alongside traditional processors, for accelerating specific tasks. Parallel Computing multi-core and multi-processor computers having multiple processing elements within a single machine.
HPC and Finance The Financial Services Grid Initiative Computationally intensive financial applications Derivatives pricing Portfolio management Risk managment