The healthcare sector went on a tear beginning in 2011, thanks in large part to the passage of the Affordable Care Act and its impending implementat
Thursday, 19 January 2017
Last updated 17 hours ago
Aug 26 2009 | 10:04am ET
By the Essvale Corporation -- High-performance computing (HPC) can be described as the use of (parallel) supercomputers and computer clusters; that is, computing systems comprising multiple (usually mass-produced) processors linked together in a single system with commercially available interconnects. This is a branch of computer science that involves research into systems design and programming techniques to extract the best computational performance out of microprocessors. In essence, HPC is about running applications and business processes faster and more reliably than before.
Hedge funds, in a bid to comply with current and future regulations, will have to manage and analyse their data quicker than ever before. The data volumes are expected to grow in size and complexity and the analysis involved will challenge even the most ardent financial analysts. These challenges in data and analytics management make the need for greater computing power within the firm compelling. The solution lies in the deployment of a high-performance computing platform.
Another compelling reason for adopting HPC is the intense competition in the hedge fund sector. From a commercial standpoint, the rewards of quicker time-to-market for a new product or service than that of the competition are huge and whenever faster calculation, improved end-user productivity and operational efficiency provides a competitive edge, then it stands to reason that the use of HPC will be beneficial.
The following are some commercial drivers for the implementation of HPC. Business Knowledge for IT in Hedge Funds
Algorithmic Trading, Speed and Data Volumes
Algorithmic trading activity is on the increase and, as a result, trading volumes are on the increase given that more trades are implemented in smaller transaction amounts across multiple trading venues. Other factors responsible for the growth in trade volumes include changes in exchange quoting methods and regulations such as MiFID and RegNMS in the USA that have led to the search for price transparency and liquidity, causing algorithmic trading products to search across a greater number of trade execution venues.
Trading opportunities based around techniques such as statistical arbitrage, whereby traders search for “alpha” from intraday opportunities, are also leading to increases in trade volumes. Traders are increasingly interested in intraday tick data given that equity markets have matured and there are fewer opportunities for arbitrage trading by using end-of-day data.
Credit Derivatives and Product Complexity
The increasing use of credit derivatives such as credit default swap (CDS) and the creation of more complex credit products require extensive use of Monte Carlo simulation methods given that the pricing techniques in the credit derivatives market are relatively new and less mature than those of other more established and better understood asset classes.
Currently, there is intensive competition among hedge funds and with other financial markets for a slice of the credit derivatives market. To achieve this, a great deal of consideration should be given to scalability across business processes, which may provide a competitive edge. HPC is a veritable technology that is an enabler for scalability and faster pricing of existing and new types of exotic derivatives.
Hedge Fund Valuation and Risk Management
As hedge funds continually use derivatives and complex investment strategies to generate better returns for their investors, they require better risk management and derivatives valuation. HPC can enable effective generation of alpha and well as returns that are less correlated with mainstream markets.
Application of HPC in Hedge Funds
Backtesting, Data Mining and Complexity in Algorithmic Trading
As algorithmic trading practices evolve, strategists in hedge funds will be looking for new ideas to develop and backtest in production-like environments as quickly as possible in order to achieve faster time-to-market. This task could be time-consuming and involve much iteration of rewriting and retesting in an automated trading environment that is production worthy.
In addition, the growth of the automated trading market will necessitate the use of more complex algorithms that combine real-time statistical analysis of real-time intraday and historic data. The degree of complexity is likely to increase considering the new asset classes that might fall into the realms of automated trading, resulting in the requirement for real-time pricing for derivatives and fixed-income pricing within all algorithmic trading solutions. As execution latency makes the difference as to whether an algorithm is profitable or not, and is increasingly viewed as the current key technology for algorithmic trading, processing power will be essential for success in this market.
Given the growing complexity of algorithmic trading, there will be the attendant difficulty in analysing huge volumes of stored real-time data, which will present a challenge to traders.
To overcome this data analysis challenge, hedge funds need scalable, fault tolerant server-side tools that emit processing power which traders can use to try new ideas and strategies quickly and easily. The advantage to be derived is faster time to market for new trading strategies that is enabled by faster parameterisation, faster backtesting, faster repricing of derivatives and faster data analysis.
Hedge fund traders use “what if” scenarios for market risk analysis at both instrument and portfolio level. This is because they need to manage and hedge the position of each instrument, after pricing the instrument and trading with a counterparty, within the context of a wider portfolio of derivatives, underlying securities and hedge instruments.
Faster revaluation of their portfolios and analysis of more complex scenarios in shorter periods of time is the advantage that the processing power of HPC offers traders. They are also able to carry out more vigorous hedging and, in addition, analyse pre-trade opportunities more quickly.
Risk measures such as Value at Risk (VaR) have been gaining popularity among hedge fund risk managers, despite the view taken by industry experts that there is no universally accepted VaR model in the hedge fund industry. This is attributed to the fact that hedge fund returns, unlike traditional asset classes, exhibit rather unstable statistical properties as hedge funds can also migrate from one strategy to another.
Nevertheless, the two main techniques used to estimate VaR are historical simulation and Monte-Carlo simulation. Monte-Carlo simulation is used to value and analyse instruments, portfolios and investments by simulating the various sources of uncertainty affecting their value, and then determining their average value over the range of resultant outcomes. Historical VaR involves running a current portfolio across a set of historical price changes to yield a distribution of changes in portfolio value, and computing a percentile (the VaR).
These two methods involve intensive computations and in order to reduce calculation time, it will be beneficial to spread this large compute load across multiple servers and processors. Furthermore, given that VaR is calculated as part of an end-of-day process, the growth in portfolio will be accompanied by longer processes which could make a mockery of the concept of overnight reporting, as there might not be any allowance for reruns in the event of process failure or anomalous results that may require further investigation.
A number of notable software giants including Sun and Microsoft are already offering HPC platforms as it is evident that as financial institutions face the challenging journey towards real-time pricing, trading and risk management across asset classes, they will need high-performance computing.
The deployment of HPC into the IT infrastructure of hedge funds will be the basis of many projects in these firms in the years to come.
The previous excerpt is from the book, Business Knowledge for IT in Hedge Funds, which is published by the Essvale Corporation. The 12 chapters in the book cover the following topics: an overview of the Hedge Fund Industry; recent trends in Hedge Funds ; the business environment in Hedge Funds; major players in the Retail banking industry; the common systems used in Retail banking; and, the future of IT and business in Hedge Funds. Latest innovations in business and IT in the Hedge Fund industry discussed in the book include Popularity of Energy and Environmental Hedge Funds , Hedge Funds Trading in Weather Derivatives, 130/30 Strategies in Hedge Funds, Continued Evolution of Fund Administration, High Performance Computing, Storage Area Networks, Algorithmic Trading and more. The contributors to this book include: Dr. Mark Mobius of Templeton Asset Management, Dermot Butler of Custom House Services, Mike Harriman of Positive View, Flex Trader, Finanalytica, Advent Software, Sophis, Beauchamp (Linedata) and more.