Retour au numéro
Vue 96 fois
janvier 2007

Careers in Investment Banking Technology

Publié par Richard MEYER (P87) Dir., Global Equities Risk & Analytics à Citigroup (New York) | N° 424 - Industrie et Contrefaçon / Les ingénieurs et le secteur financier


The financial markets, Wall Street, the City of London, HK, Tokyo and emerging markets offer a vast array of potential careers for bright, motivated people. Clearly a number of more business-minded graduates, maybe with additional finance or business degrees, will pick careers as traders - those with a good maths background may work in derivatives. But for engineers with a strong technical and mathematical bias, information technolo-gy services offer an extremely wide range of career opportunities and fascinating technical challenges, these are what I intend to focus on in this article.
It may seem an odd choice: why would an engineer trained in hard core maths, physics and material sciences (not to mention geology!) choose a career in the financial services industry? In my case, only a frac¬tion of my economics and financial courses had stuck in my hard-core technical mind, and computer science research appeared to be the only avenue offering conti-nuous excitement and challenges (my PhD advisor had warned me that I’d simply end up writing airline booking systems). But as I interviewed with Salomon Brothers in London in 1994, two key factors attracted me: the calibre of the people I would be working with, similar to those I had encountered in my research work, and the fact that my passion for computers and technology would find a challenging outlet. Ultimately, this was a well-paid job that would allow me to build the most sophisticated software possible, tackling complex and varied pro-blems, and seeing the applications I built immediately applied to the needs of a very dynamic business.
The Need for Technology
Investment Banks are huge consumers of technology, investing 10 to 15% of their considerable revenues into software development and the hardware required to meet their insatiable technology needs. Every business manager will tell you that his is a “Technology Business” - one of the key differentiators between com-peting banks. This is driven on two fronts: the functio-nal sophistication of the services they can provide to their customers, and the requirement to reduce opera-ting costs, i.e.: the number of people (traders, sales, support, middle-office etc.) required to service their business. This is further compounded by the ever-redu-cing margins in most high-volume and traditional busi-nesses, and the simultaneous drive to provide evermo-re rich & complex services to their customers.
Software and technology intervene at every stage of life-cycle of any trade: from the initial reception of an order, pricing the contract, executing the deal, confir-ming the transaction, managing payments, market and credit risks to the firm throughout the life of the tran¬saction . We therefore aim to offer the same consistent services, products and reporting globally, and to deve-lop global interconnected solutions. long with these various tasks come the critical and necessary require-ments for proper accounting, risk reporting, audit trails, data retention, business continuity, software upgrades,information security, customer confidentiality, local and interna¬tional regulations.
The result is that large firms, such as Citigroup, run internal Investment Banking development teams of 20,000+ software pro-fessionals spread across 88 countries.
Many Career Paths
All the investment banks compete to recruit and retain the best staff possible in all locations: the world’s large financial centres as well as many near- and off-shore locations.
A young engineer starting out in the financial technology indus-try will be presented with the following high-level career opportu-nities:
· Technical: developing software to cover the huge variety of investment banking requirements.
· Architectural: designing systems and providing guidance to teams, ensuring that the bank’s over-arching business needs are met in a coherent plan.
· Quantitative: developing financial models, automated trading algorithms, data analysis and prediction
· Managerial: covering Project and Programme management, from small to huge teams of developers, business analysts, qua-lity assurance and production support teams.
And of course, combinations of any of the above: the “generalist” education approach provided by the various Ecoles des Mines real-ly pays off in this environment since there are many facets and trades to be mastered.
From a business or service perspective, the types of roles would typically break into the following specializations:
· Analytics: called “Quantitative Analysts” or “quants” for short, most of the developers in these areas will require PhD-level maths or physics in order to develop the cutting-edge financial models to price the latest exotic deals. A number of universities offer specific training in financial analytics. These roles are inva-riably based on the trading floor itself and usually make up small teams of up to 10 people. In many cases the financial analytics are considered a “solved” problem, however the technical impli¬cations still lead to considerable challenges as most major exo-tics desks require the continuous use of 1000s of CPUs to pro-vide the necessary calculation power for near real-time risk and pricing calculations.
· Trading Systems: these are the heavy artillery of the trading floor, servicing the sales and trading desks. These systems have to be extremely flexible, to encapsulate a robust trading work-flow, to support all the features of all the exchanges, to provide extensible frameworks for automated trading, and ultimately to cope with high volumes of trades, (order of millions per day with millisecond transaction response times). In many cases, firms will be seeking to reduce the number of traders or sales-people to conduct their vanilla business, automating this as much as possible. An extreme case is the automated market-making of options where a single trader may be continuously quoting 100,000 options across 100 underlyings with the sys-tems automatically responding to every market price adjustment within milliseconds whilst simultaneously automatically optimi-sing the underlying hedge portfolio and feeding this information back into the option price generation systems. Large-scale tra¬ding systems frequently operate on globally distributed server-farms of 100+ machines.
· Connectivity: This is where the trading systems are hooked to the outside world, pulling in market data, sending out orders to electronic exchanges and broker networks, receiving executions. The investment bank’s external customers can connect directly to the firm’s order-routing networks, with different latencies depending on the level of automated or manual service reques-ted. The FIX Protocol, initially developed by Salomon Brothers (now Citigroup) in 1992 has become the financial world’s order-routing standard. The vanilla options1 pricing feeds in the USA have exceeded the 100Mbit bandwidth and are set to quadruple over the next 18 months. This alone is a major challenge affec-ting network & routing designs from front to back, right up to the trading floor & workstations, but also directly impacting the algorithmic trading systems and front-end client applications as software and CPUs need to scale, to support the exponential growth in data volumes.
· Risk and Pricing: closely allied to the desk-based quantitative analytics teams, these systems are mostly required for the various derivatives businesses and provide the entire framework for pricing, structuring and entering deals, calculating intra-day market risks, profit and loss, analysing scenarios, capturing contractual and official observation levels, calculating payments and the entire life-cycle of derivative trades - many of which are 3 to 10-year contracts. The major challenges in this area are the complexity of the data requirements, the calculations perfor-med, the flexibility required, as well as the need of providing solutions that will scale to 2000+ CPUs. Grid solutions are beco-ming the norm in the financial industry as evidenced by the many top investment banks now listed in the super-computing sites, purely to support their derivatives calcu-lation requirements.
· Middle-Office: the systems in this area will deal with everything that happens to a trade post-execution. This entails officially reporting the trade back to the customer and custodian entities, settling the deal, arranging and executing payments and also includes added-value services such as allocating a single trade across the multiple fund portfolios a customer may be mana-ging. Latency is less of an issue here, but high volumes and 100% robustness of the processes and data quality pose their own technical challenges.
· Production Support: this is a critical area of the financial tech-nology services, possibly more so than in other industries, as all the complex interlinked systems above need to provide 100% availability. Production Support teams, usually based on the tra¬ding floors, need to understand both the technology and the business sufficiently well, to either guide users to use the sys-tems effectively, or actually drive the resolution of potentially major problems. Not everybody is cut out for this sort of role: excellent communication skills and the ability to stay calm in crisis and marshal various teams to resolve major issues are key attributes here.
· Technology Infrastructure: clearly the counterpart to all the software and systems, detailed above, are the hardware and networks which they run on. My focus in this article will remain on the software side, but needless to say, all investment banks run major data centres and large engineering teams continually adjusting to provide and administer the huge networks, data feeds, data centres, hardware and security infrastructure requi-red to keep all areas of the investment bank up and running 24x7.
Most software development is conducted in the typical “cubicle farm” type environment - as popular among most of the software industry. However, investment banks will also have small pockets of developers working in front office, typically on the trading floor itself and in close proximity to traders or production support teams. I would recommend these areas as an ideal learning and proving ground to slightly more experienced engineers with 2 or 3 years of practical software development, but be warned that not everybody is cut out to work in this sort of noisy and demanding environment. Requirements change quickly, often at the whim of the head trader, and demands may be unrealistic. It is probably the most challenging environment to work in and a great way to learn a lot very quickly, but simultaneously it is not the best envi-ronment to architect and build large and complex systems. Note that quantitative development teams are frequently trade-floor based and provide a lot of 1st line support to traders and deal-structurers.
The key to all of these roles is allying business knowledge with core technical skills. Success, especially early in one’s career, comes from being able to build a relationship with your business partners, talking their language, translating their business require-ments into real applications, anticipating needs and ultimately providing them with a smarter, faster, better service than is avai-lable to the competition.
Technical Challenges
As I highlighted earlier on, technology is my passion, and I hope that the preceding paragraphs will have highlighted the type of complex technical challenges that engineers will face within Investment Banking:
· High transaction volumes and throughput: multi-million trades per day in real-time.
· Low latency: typical latencies for providing Direct Market Access to customers are measured in milliseconds.
· High Availability: extreme reliability is required so that custo-mers do not experience and service outage. This requires consi-derable use of high-availability systems design, redundant sys-tems and networks, guaranteed messaging, database persisten-ce and replication - all the way to entire “disaster recovery” data-centres available to take-over at any point.
· Data Complexity: the demand for flexible data structure and large data sets pose considerable challenges as systems need to pull in vast amounts of complex information from basic listed securities information to the detailed financial market analysis required to price exotic contracts. Additionally most regulatory and control requirements place heavy demands on data reten-tion and audit-trails.
· Scalability: distributed systems are now the norm as single ser-vers are no longer able to address the financial industry’s capa-city, latency and high-availability needs. In practice, expanding systems capacity to match business needs is a continuous exer-cise that has to be baked into the software design, development and maintenance process.
· Flexibility: the need to respond quickly (sometimes within hours) to customers’ requests is pushing us more and more to provide “dynamic” software solutions leveraging recent techno¬logies such as just-in-time compilation.
· Cost Optimised: solutions must be right-sized to the revenues of the businesses that finance them and make best use of hard¬ware and infrastructure.
Citigroup’s technology standard catalogue lists over 10,000 ven-dor products used to develop the 6,500 largely internal applica¬tions developed by its investment bank - so it is by definition impossible to list all the products and technologies used to over-come the aforementioned technical challenges! Key components however usually rely on Java, C# or C++ on Windows, Solaris or Linux, with Sybase, Oracle or MS-SQL providing the relational database support. Many technical and business vendor products will be brought into the software mix. The engineering challenge becomes knowing how to leverage this combination of software tools and vendor applications to create a suitably powerful, flexible and integrated framework for the business you are sup-porting.
This is a highly challenging technical environment continuously pushing the envelope. But as the market becomes ever more sophisticated, only the larger financial institutions will have the financial clout to build the highly complex systems required. I would therefore recommend that young engineers start out with a large firm, rather than a smaller venture that is unlikely to be able to really finance the complex solutions these businesses require.
It is possible to lead a very successful technical career evolving from pure software development to architect, senior architect and senior technical specialist roles. It is however likely that the suc-cessful software engineer will be faced with the career option of taking on positions of increasing management scope.
Beyond the usual technical, distributed people and global project management challenges that one might normally encounter in the software industry, the financial markets impose their own idio¬syncrasies:
· Time to Market: is often deemed highly important by the busi-ness stakeholders. In competitive markets, being the first to be able to offer a given product or service is often critical in ensu-ring high returns before the competition catch on, everybody joins the fray, and margins erode correspondingly. For this rea-son, Citigroup is encouraging “Agile” and “Scrum” type metho-dologies to accelerate delivery on critical projects where these approaches are suitable and called for. Note also that it is not uncommon to see delivery and release cycles of days or weeks, particularly in front-office applications, which may be unusual in most other software industries where the end-users are not quite so close at hand.
· Controls and Compliance: the investment banking and retail banking industries are highly regulated with internal audit teams preparing the ground ahead of external audits by regulatory bodies. Many of the business audits reach directly into how applications are built and designed, what functions they per-form, how they enforce regulatory trading checks, how they detect and warn of any violations. The development, testing and releasing procedures themselves will also be audited regu-larly. Both of these are of critical importance to the firm as in many cases a “failed audit” would then constrain the business to remedial action, potentially under extreme regulatory pressu¬re, prior to being able to focus on business development.
· Operational Support: all the systems must be maintained fully operational even whilst incremental software improvements are being carried out. This by itself provides additional challenges as to: when to deploy? - Around which “green zones” and anti-

cipated “business events” - and how to rollback in case of any systemic or hardware failure?
· Global Businesses: in many cases, even if the trading busi-nesses themselves are not managed globally or consistently, it makes sense to have global applications providing a wide array of services with local adaptations to cope with regional diffe-rences in function, product base, volumes and regulatory rules. Clear communication and a fair level of diplomacy are required to manage the ensuing budget allocation and project prioritisa-tion meetings!
Overall the Project Manager’s role is a challenging career opportu-nity as it calls for a complex mix of skills and multi-disciplinary approach. Project managers constantly juggle between business priorities, technical challenges, testing and deployment calendars, the operational aspects of keeping systems up and running, bud¬gets, audits, recruitment, business and financial reporting and all the while handling standard people and project management issues and maintaining a good relationship with their business partners and clients.
I hope that the above article has given the reader an overview of the multiple career paths and challenges open to a well qualified software engineer. These are fascinating and demanding roles that
will stretch anybody’s skills to their limits and provide continuous challenges and great rewards.
In some ways the students from the various Ecoles des Mines are ideally suited for these roles as their training allows them to span many disciplines beyond the purely technical, including their inva-luable mathematical background. Simultaneously it is worth noting that the investment banking mentality is very much one of a meritocracy. Performance is measured on what you achieve, not on academic reputation or whom you know. Citigroup places a lot of importance on being an equal opportunities employer and this is strikingly demonstrated by its international staff and manage¬ment teams at all levels. Therefore, as a note to any prospective candidates, remember that you are up against the best candidates from the top universities worldwide and being interviewed by people who are themselves experts in their various domains. Possibly unlike some interviews for French companies, the name of your university will not carry you through (you’ll be lucky if the interviewer has ever heard of it!) and you will be measured on what you can demonstrably bring to the table - including excel¬lent levels of both written and spoken English!■


Richard MEYER (P87) Dir., Global Equities Risk & Analytics à Citigroup (New York)

Articles du numéro



Vous devez être connecté pour laisser un commentaire. Connectez-vous.