R has come a long way since its inception in the early 1990s. Although data scientists have been singing its praises for years now, myths about R and its use in financial services and big data analytics still persist.
In this article, we’ll address six of the most common misconceptions about using R in financial services and show you how R can be used as a powerful tool for business growth, big data analytics and more.
1. ‘R is not as powerful as other software packages.’
This is inaccurate. In fact, the opposite is true: R outperforms other statistical software packages in terms of processing power, capability and flexibility.
Let’s break it down:
- R can do more than most software packages. Robert Muenchen, author of R for SAS and SPSS Users, wrote that ‘there are very few things that SAS or SPSS will do that R cannot, while R can do a wide range of things that others cannot.’
- R has a strong, cutting-edge package ecosystem. It’s an open-source language, meaning users can contribute their own packages and libraries to the community. The result? R users get the latest techniques before anyone else – including novel algorithms and experimental techniques. If there’s an analytic technique out there, there’s probably an R package for it.
- R is extendable. You can build your own tools and methods for data analysis with R – which is something you can’t do as easily with other software packages.
2. ‘We need to visualise our data and create graphics. R can’t do that.’
Think again. According to data scientist Matt Adams, R’s graphics and charting capabilities are ‘unmatched.’
R has many packages for data manipulation and plotting, which make it easy to generate charts, build graphics and visualise complex datasets. Miguel Ríos at Twitter visualised every geotagged tweet since 2009 using R code and the ggmap package, and the results were stunning. The fact that newspapers such as the Financial Times use R as part of their data journalism toolkit is further testament to R’s agility and powerful graphic capabilities.
3. ‘Everyone is using SAS, so we’d be the only ones trying to use R.’
Not true. R is the fifth most popular programming language according to IEEE Spectrum. R is following its momentum from previous years, as part of a positive trend in general for modern big-data languages. This list compares against other general-purpose programming languages, which shows just how widely used it is.
Here are a few examples of financial services firms using R:
- Credit Suisse has been using R on the trading desk for several years now. In a 2012 interview with The New York Times, Ryan Sheftel (then head of the bank’s automated Treasury bond trading) explained that the best traders were making use of languages like R to automate investment decisions. Five years on, it looks like R is still the language of choice at Credit Suisse.
- Lloyd’s of London uses R for everything from performance management and reporting to exposure analysis and Monte Carlo simulation. For example, the visualisation tools Lloyd’s Statistics, a statistical guide to the firm’s insurance market, were built using R.
- ANZ Bank, one of the largest financial institutions in Australia, uses R for credit risk analysis. In particular, the firm used R to create a custom model that estimates the probability of default for individual loans.
So, if you’re thinking of migrating to R, you’ll be in good company. And you’ll have no problems attracting experienced, qualified talent.
4. ‘R is a language for programmers and engineers, not analysts.’
False. R is the lingua franca of statistics and data science, and has been for a long time. It’s widely taught in universities, and it’s the default choice for undergraduates studying econometrics, business and quantitative analysis.
But aside from that, the number one reason why R is the language of statistics is simple: it was developed by statisticians, for statisticians. It’s syntax and structure are in tune with the way statisticians think, and the core language has elements like data frames built-in.
5. ‘It’s free, which means it’s not supported and there’s no documentation!’
R is free because it’s open source, that’s true. But to say that R isn’t reliable or supported is simply wrong. ‘Think free as in free speech, not free beer,’ as Richard Stallman puts it.
The R community is huge, and it’s active. Packages are released at an astonishing pace and the number of blogs, forums and websites devoted to R have proliferated. At the time of writing there are 168,289 questions about R on Stack Overflow alone, and companies like Microsoft are producing resources for R Server as well as offering the support and backing of a multinational software company for their implementations.
That’s a huge amount of support and documentation, all produced by a community of statisticians, developers and companies that provide commercial support for R (like Mango). For Glenn Meyers, Vice President and Chief Actuary AT ISO Innovative Analytics, ‘the most powerful reason for using R is the community.’
6. ‘R doesn’t work on our platform or integrate with our existing tools.’
You absolutely can use R with tools you’re already using, and on the platforms of your choice. It’s possible to develop and run R models on:
- Windows, with R Server for Windows
- Linux
- Hadoop/Apache Spark
- Teradata database
Plus, you can integrate R with SPSS Modeler, Tableau and whatever else you’re using!
It’s not too late to get on the R train
They say that data is the key to business growth, but your data is worthless without the ability to rigorously analyse, manipulate and visualise it.
This is what makes R so powerful. With R, you can analyse complex datasets using the latest techniques – and you can visualise your insights in compelling graphics and charts. It integrates with other tools, runs on a variety of platforms and is supported by an active community of users who are committed to the development of the language itself.
And if you’re still not convinced, read more about R, come to one of our events or talk to an expert. It’s never too late to get on board.
