Investment banks are technology firms. So says Lloyd Blankfein at Goldman Sachs and Marianne Lake at J.P. Morgan. Both banks have thousands of technologists and have prioritised IT hiring over most other divisions. Across the street, IT specialists with deep knowledge of programming languages are in demand.
“The main demand has always been server-side Java programmers, but we’re seeing a real push for Python, as this is being used more in quantitative programming, and certain banks like Bank of America and J.P. Morgan, probably the biggest firms making the most hires on the street, are building new trading and risk platforms using in-house proprietary languages based on Python, such as Athena and Quartz,” said Nick Vermeire, management consultant and the head of the development team, Americas, at the Palm Mason Group, a recruitment firm.
We conducted an informal survey with banking technology recruiters to see which programming languages their clients are asking for most. Here’s what they said.
The clear runaway winner has been the most sought after programming language for years on Wall Street.
“Java developers are needed for anything from low latency execution and order management systems to in-house risk and valuation platforms,” said Jared Butler, head of financial technology recruitment for North America at Selby Jennings. Java is also really well-suited for data simulations and modeling, added John Reed, senior executive director for Robert Half Technology.
The reason for the all the fervor around Java is two-fold. There’s a high demand for the skillset along with a dearth of qualified candidates. Late last year, our resume database contained only seven candidates for every job that required the skill, the lowest ratio among all major programming languages.
Reed said that Java developers can demand up to a 10% premium on salary compared to others in the market.
“Java is used extensively on the server, less for UI [user interface],” Christian Glover Wilson, vice president of technology and strategy at Tigerspike, a digital agency.
Finishing in a virtual tie were C++ and C#.
“C++ continues to be the go-to language for high volume/high frequency trading, simply because it’s the most efficient tool to build an extensively optimized backtester and execution system in order to process the high volumes of data,” said Butler.
Schiller agreed, adding that C++ is also often used for building applications running on many banks’ legacy systems. “Due to the high cost of moving to new technologies there will continue to be significant demand for those who can program in languages compatible with the legacy environment,” she said.
Like Java, C# can be used in a variety of projects, particularly data simulations and modeling. It had the second lowest ratio of candidates to jobs in our database, giving qualified job seekers plenty of leverage when it comes to pay. Reed said that C# developers can also demand a 9% to 10% bump in pay over colleagues with expertise in other languages. SQL is the third language that tends to offer terrific bargaining power, he said.
“C# is still in use but now pretty much only for quanty, low-latency things,” Glover Wilson said.
Following the big three is an up-and-coming programming language that wasn’t as hot a few years ago. Python is great for creating analytic tools and quant models – critical tools that contribute to investment banks’ and hedge funds’ trading strategies, according to Schiller.
In addition, Python is becoming better known for being easier to use and faster to program than the traditional languages, said Butler, who offered up a number of reasons why it could replace the aforementioned languages in popularity, particularly in investment banking.
“Firstly, programmers are able to do as much with 10 lines of Python code as they are with 20 lines of C++ , and with a much lower margin for error,” he said. “Given the increase in regulations/ best practices, you can see the appeal in using it from this perspective. Moreover, given it’s performance and usability, Python is fast-becoming the language of choice for ‘red hot modelers’ in comparison to C++.”
Plus, as technologists crave greater exposure to the business side of banking, Python has become more popular. It enables programmers to better collaborate on projects with quants, researchers and analysts, Butler said.
“Python is the scripting language of choice, and it’s in use more and more,” Glover Wilson said. “Especially as big data grows in importance, Python alongside Scala, for example, is key.”
However, unlike Java, C++ and C#, along with SQL, Python doesn’t provide much at the negotiating table, simply due to the ease of learning the language.
“From a hiring perspective, most hiring managers are willing to forfeit Python proficiency on the basis that most skilled programmers should find it easy to pick up,” Butler said.
Other languages that got some votes include SQL, PHP and ETL.
“We are seeing strategic hires from associate to executive level for candidates carrying an array of skills, from older ETL technologies such as Informatica to more modern big data-related tools like Hadoop tech stack, HBase, HDFS, MapReduce, Pig, Hive, Impala, Flume and Cloud,” Butler said. “ETL technologies continues to be important for successful data warehousing and the fat-cutting financial data used by investment bank trading arms globally.”
On the other hand, demand for Microsoft’s Windows Presentation Foundation (WPF) is waning, while HTML5 is on the upswing, as are Hadoop, Cassandra and Scala.
“There is a lot demand for big data and data-processing technologies such as Hadoop, Cassandra and Scala, and we’re seeing more and more banks adopt this,” Vermeire said. “C++ and C# tend to be staple development languages, but there has been perhaps less appetite for front-end Microsoft development – WPF – and more demand for HTML5 skills.”
Quant and high-frequency trading languages including Q and R are in narrow but embedded use in shops that do high-frequency/low-latency trading, Glover Wilson said.
“Web stack is huge – it’s continuing to win ground from .net UIs,” Glover Wilson said. “This means that HTML5 and JS are common skill sets.”
Most of the JS is UI-oriented such as Bootstrap, but AngularJS is also very common, he said.
“Node.JS is growing and ReactJS’s growth is slower but it will become important,” Glover Wilson said. “Web developers are expected to know ASP.net (C#) and SQL/DB management, as well as front-end languages.”