MathWorks Newsroom

Modeling and Analysis in the Wake of the Global Financial Crisis: The Financial Services Perspective

This report aims to shed light on some of the challenges facing the buy- and sell-sides, and to identify the key business drivers for popular data analysis and modelling techniques within the financial services industry.

The impact to your business of poor models: the buy-side/sell-side split.

By surveying experts from the buy- and sell-sides and academia we have captured a range of industry insights and opinions to better understand the importance that models and data have in all aspects of the trade lifecycle.

The research comes against the backdrop of the financial crisis of 2008 and the subsequent unprecedented levels of regulation which the industry now faces.

State of the Industry

The second largest outlay for the average financial institution is its data – second only to the cost of staff. Data is used across the organisation, most commonly and potentially profitably in trading strategies, but it also supports every other function, including risk management.

The 'data deluge' has captured the interest of the media worldwide, as industries seek to capitalise on the data that exists in their own organisations and in the public domain. However this is not a new phenomenon for financial services, which has sought techniques to monetise data for decades. Today's challenge is dealing with the 'three Vs of data' – volume, velocity and variety.

Unprecedented levels of regulation are facing the industry, post-2008, including MiFID II, UCITS, Basel III and Dodd-Frank. However it is apparent from our research that while firms are adapting to adhere to regulation, this is not a key driver in model production or model use. In fact, businesses commonly hold the view that instead of dedicating resources to risk models to appease regulators, this would be better allocated to bring alpha generating models into production.

Against this backdrop, MathWorks has sought to obtain industry insights and opinions from within the financial services industry to better understand the importance that models and data have in all aspects of the trade lifecycle.

This in-depth study involved questioning 43 experts from within the buy- and sell-sides, as well as academia, including hedge funds, institutional investors, banks (including proprietary trading desks), broker/dealers and university lecturers. On the sell-side, four of the top 10 investment banks1 participated in the study.

In this report, we have tried to shed light on some of the challenges facing the buy- and sell-sides, and understand the key business drivers for popular data analysis and modelling techniques.

Key Findings at a Glance

Business Outcomes

  • 88% of financial institutions believe they would lose their competitive edge, 79% believe their profits would decrease, and 54% believe risk would be increased, if they were operating poor models – for example flawed or outdated models
  • Slow model development will result in firms lacking the agility to respond to market changes, and ineffective risk management , say 82% and 74% of firms, respectively

Model Integration

  • Cost (65%) and risk (62%) are the biggest concerns of integrating models into business processes
  • It is currently taking 'months' (51%) to integrate models into business processes. However the buy-side would like to cut this down to 'days' (75%); the sell-side ambitiously wants to reduce this time to mere 'hours' (40%)
  • 83% of financial institutions are trying to speed up the process of model development

Trading Strategy

  • Quality of data (67%), smart models (54%) and speed of execution (50%) are the most critical elements of a successful trading strategy
  • 59% of financial institutions are looking to increase levels of automated trading; 67% of sell-side and 46% of buy-side firms are looking to increase levels of automated trading
  • Only buy-side respondents believe automated trading models have had their day, and 31% of buy-side firms are looking to move towards alternative trading models
  • 54% of financial institutions are looking to improve the execution times of models

The Data Deluge

  • The biggest issue associated with the data deluge is data quality, with 68% of respondents citing it as a challenge. Creating effective models (57%) and data variety (38%) were also among the main three issues facing financial services
  • The actual volume of data is not a core challenge for financial institutions, with only 32% citing it as problematic
  • The datasets being dealt with are, in general, not as large as externally perceived: 49% of financial institutions are dealing with datasets in gigabytes; 28% with megabytes
  • The sell-side is grappling with larger datasets than the buy-side, as 19% of sell-side is handling datasets of terabytes

For More Information

1 Financial Times: Investment Banking Review, H1 2012, http://markets.ft.com/investmentBanking/tablesAndTrends.asp