How to survive as a quant AND a trader as Big Data takes over
One of the most impactful elements of the big-data era, especially for traders, is the rise of predictive analytics - using statistics, data analysis and modelling to identify patterns and forecast future performance.
This is already starting to influence what financial services firms are looking for. What do traders, data scientists and tech professionals need to know to get hired, avoid getting fired and get promoted?
1. Traders don't need to die just yet
The new thing among financial services organisations is to use 'semantic technologies' in their analysis of data. This means that the vast amount of data available to financial services organisations is presented in a way that is easier for both machines and humans to understand. Practically speaking, data is no longer siloed and can be analysed in a unified platform in real-time. But is also means that the 'old' skills required from traders are no longer as valid.
“We joke around the office that the old style of trading is dead, but all the old traders aren’t dead yet,” said Howard Getson, CEO, Capitalogix Trading, which develops hedge fund technology with artificial intelligence (AI) trading systems that evaluate global markets in the cloud at the Trading Show Chicago 2016. “It must be scary to be somebody like that – you’ve got a buggy whip, which is a horse-drawn carriage, and even if you had the best driver on the planet, he’d be thinking how to make the buggy whip better, rather than inventing a car.
“We now have AI that enables us to create trading systems that are better than a human is ever going to do it, with millions of algorithms, so I don’t hire traders to be good traders – they have to be skilled, but I care more about their contacts, human capital, judgement, energy, hard work and their ability to make our firm better,” he said.
Getson said that predictive analytics gives traders a much better understanding of data and therefore allows them to deploy capital where they expect it to make money at all times, taking much less risk to generate better returns.
“Technology can act as a sensor that does market surveillance to a greater extent than ever before,” Getson said. “Everybody has the same data, and most people are looking at market price and market time, but there are other ways to look at it. To be a successful trader, you have to find a way to do things differently from other people to get a different result. Predictive analytics is making the invisible visible.”
2. Quants with people skills and traders who can code
It can be difficult for hedge funds and quantitative trading shops to find candidates with the right blend of skills. Computer scientists are coders, or mathematicians, and they often don’t really understand the trading models, said Samuel Chen, vice president and quantitative researcher at Seymour Capital Management.
“Using Python, it’s easy for someone on the research side to back-test [a trading model] and see if it works,” Chen said. ”You can implement a trading system in Python, so in hiring people, I’m looking at trading experience and someone who can generate alpha but also knows how to code.
“I will hire C coders if necessary, someone who can throw in models and figure out which ones are useless,” he said. “If the performance is not great based on a certain model, it’s difficult to tell if the model is bad or the timing is off.”
Partners and salespeople bring quants to the fundraising meetings now, because investors want to talk to them.
“We hire for a unique role, someone can talk to people but also knows their stuff,” Chen said. “For us language is a problem with people coming from various fields – it takes time to get to a point where we can have a conversation on the same topic.
“Money is made by finding pockets of concentration, using an algorithm to find it on a risk-adjusted basis, so you hedge it,” he said. “We’re in futures because we believe that’s where money is made, but we hedge that with equities.”
3. Data scientists need to understand trading
Noticing a pattern here? Data science might be the new must-have skill-set, but having this alone is not enough. Those who can code are being hired on Wall Street, but an understanding of the markets and working well with other people is essential as well.
“Managers say, ‘I need someone to develop for me’ – most are hiring candidates not just based on their skills like Python and MATLAB but also have they been executing their own strategies,” said Sri Krishnamurthy, the founder of QuantUniversity.com. “They don’t want someone with all these ideas but doesn’t have the ability to execute.
“They want implementation experience, not just tech knowledge, because they will be expected to scale these [trading] models,” he said. “You’re seeing the demarcations between departments [at financial services firms] disappearing – technology people need people skills and an understanding of what it takes to test and deploy those algorithms.
“When building teams, bringing in complementary skills sets is key – computer scientists are good at coding but don’t know what’s going on in the financial side, which is where traders or quants come it. Just because you have the tools [like predictive analytics] doesn’t make you effective. Ask yourself: ‘Do we have the right skill sets to leverage these technologies and apply them to our business case?’”
Photo credit: shutter_m/iStock/Thinkstock