<style>.blog-template-wrapper{width:100%;max-width:1040px!important;padding:0!important;margin:auto}.blog-template-wrapper .w-richtext{width:100%!important;max-width:none!important;padding:0!important;margin:0!important}.blog-template-wrapper h2{font-size:34px!important;line-height:1.25!important;margin-top:48px}.blog-template-wrapper h3{font-size:24px!important;line-height:1.35!important;margin-top:32px}.blog-template-wrapper p{font-size:19px!important;line-height:1.7!important;margin-bottom:24px!important}.div-block-223{max-width:1040px;margin-left:auto;margin-right:auto}.div-block-223 h2{font-size:34px!important;line-height:1.3!important}@media(max-width:767px){.section.articles .base-container{padding-left:24px!important;padding-right:24px!important}.blog-template-wrapper h2,.div-block-223 h2{font-size:27px!important}.blog-template-wrapper h3{font-size:21px!important}.blog-template-wrapper p{font-size:17px!important}}</style><div class="w-richtext"><p>On September 29th, the Amsterdam Quant Society welcomed Taco Portengen, Global Head of the Quant team at ING, as a guest speaker. Over roughly thirty years, Portengen has seen most of the industry, working in multiple positions across London, Hong Kong, Singapore and the Netherlands, whilst living through every major financial crisis since the late nineties. Today, he leads a team of over 100 quants at ING.</p><p>Portengen studied Applied Physics at TU Delft, completed a PhD in physics at UC San Diego and a postdoc at Oxford, until an advert in his lab caught his eye. A bank was looking for people who could solve drift-diffusion equations, with no economics knowledge required and an attractive salary. As he later added, it was no coincidence that the advert ended up in a physics lab, as the Black-Scholes equation for derivative pricing closely resembles the equation used to describe fluid flow in physics. His break into finance came at Credit Suisse Financial Products in London in 1998, after which he navigated different banks and desks before ending up back in the Netherlands.</p><p>Portengen’s presentation offered a glimpse into the quant team at ING, which is not Goldman Sachs or a hedge fund seeking a 40% annual return, but rather a commercial bank whose core business is lending, and which prioritises safe returns over risky endeavours. It shows how a quant team works in a more conservative environment and removes the illusion that being a quant is necessarily an 80-hour-a-week job. This article offers a lens into the structure and daily life of ING’s quant team, and into the qualities and experience recruiters look for in junior candidates.</p><h2><strong>The Team and Its Roles</strong></h2><p>The structure of ING's quant team is not so different from any other quant team. As Portengen put it, there are many flavours of quant, so here are all the positions at ING:</p><p><strong>Front-office quants</strong> develop the models that are used for pricing, hedging and trading decisions. They sit close to the salespeople and traders, and are the ones who get asked why a price is what it is and how it will move when the market moves.</p><p><strong>Risk quants</strong> build the models that calculate capital, counterparty credit risk and stress tests. In other words, they look at how much ING could lose if things go wrong.</p><p><strong>Model validation quants</strong> check whether what the other quants have done is actually correct and complies with the rules and regulations. A validator typically gets two to three months per model, partially rebuilds it, and then moves on to the next one.</p><p><strong>Quant developers</strong> are more IT-oriented. Their job is not so much designing the model or doing the mathematical derivations, but implementing it in a live trading system. That system has to be robust, because if it goes down, the traders cannot see their positions. </p><p><strong>Algo quants</strong> use time-series analysis to predict what the next-day move of a certain underlying will be to improve how it is hedged and achieve the most efficient way of managing the risk. </p><p>Not all of these desks are present in Amsterdam, for instance, ING’s Algo Quants are active in London and do not sit within Portengen’s team, however, there is still a line of communication between them.</p><h2>A Day in the Life</h2><p>For those who are looking into applying for junior or intern quant roles at ING, Portengen laid out what a day in the quant team looks like. The morning starts at the coffee machine, which Portengen acknowledged to be one of the most useful parts of the day, because "besides exchanging gossip, people also exchange ideas." Next, the objectives for the day begin to form: what are you going to do today, and is anything blocking you? Once that is clear, people sit down and code, with the standard increasingly being Python. For the faster models, however, C++ and CUDA are still the languages of choice.</p><p>Later in the morning, it is time to talk to stakeholders, because "you don't sit there on an island." A salesperson might come by to ask why the price is 100 and what needs to change to make it 90, whilst a trader wants to know how that price changes when the market moves, meaning that having strong social skills and the ability to break down complex information into a few simple sentences are essential.</p><p>The afternoon is for testing and documentation. It is not everybody's favourite part of the job, Portengen admitted, but a model must work all the time, so you test it and test it and test it again until everything works smoothly. Towards the end of the day, a senior quant looks at your results with you, provides feedback and helps you fit your work into the bigger picture.</p><p>And then it is time to wrap up. Unlike the rumours about the devastating quant hours, the ING quant team has a healthy work-life balance. Once the clock hits 17, you are free to leave. Of course, sometimes, additional work needs to be done and employees are expected to be team players, but that is rare. The benefits of ING lie exactly in the flexibility and balance that they offer. </p><h2>Getting In as a Junior</h2><p>Getting into ING’s quant team as a junior is not easy, a role typically draws between 10-20 potential candidates. As he talked, an image of the perfect candidate emerged.</p><p>His opening statement – “I assume most people here study econometrics, but are there any physics, maths or engineering people in the room?” – is indicative of the fact that he is looking for people from highly quantitative backgrounds. “If you hate coding, then look somewhere else.” Portengen stressed the fact that working within this industry is programming-heavy, with Python increasingly becoming the standard. C++ is a bonus, but it is something that can be learned at ING. You also need the basic theoretical knowledge, for instance, how a Monte Carlo simulation works. No work experience or Dutch required.</p><p>So how do you stand out? As mentioned before, candidates who can present themselves and form concise insights from complex information will be more likely to be hired. For instance, when a trader comes down to ask a question regarding pricing, you should be able to explain it in under a minute, so it helps if talking to people comes naturally. Beyond that, experience is always looked at favorably, especially a 6-month internship, though it is not required. Your course choices matter too, as taking more programming and math-oriented classes is always beneficial. Finally, he wants people who are committed to the role and will stay, as a lot of time and effort goes into training people, and the last thing they want is for them to leave after a year or two.</p><h2>Why ING</h2><p>An ING quant career comes with stability and balance, which is quite rare within this industry. Portengen’s own career path is perhaps the best argument. Two of the three banks Portengen worked for, Fortis and Credit Suisse, saw their share prices go to zero. Meanwhile at ING, where he has been since 2008, the rewards need to be quite high for the bank to accept the risk and as a result "things are going okay."</p><p>His last piece of advice was about luck. Portengen only managed to get into a US university because he secured the last spot to take the GRE exam. According to him, if that hadn’t been the case, his life and career would have looked very different. When the dot-com bubble burst, he lost his job at Fortis, which was the biggest hit his career has taken. It was his former boss from Credit Suisse who hired him and sent him to Hong Kong, where he met his current wife. The lesson is that you cannot control everything, but that is fine. </p><blockquote>“Don’t feel panicked if what you planned didn’t quite turn out the way you wanted it, because there’s always another opportunity.” - Taco Portengen</blockquote><h2>Questions from the Audience</h2><h3><em>How many years of work experience do I need?</em></h3><p>“It depends on the role. Sometimes we have vacancies for experienced people, because we need someone that can start right away. But we also have vacancies for juniors, and then clearly no prior experience is required. It always helps: if you have someone that has already done an internship somewhere for half a year, that’s always beneficial, but it’s not a requirement.”</p><h3><em>What kind of education path would you recommend?</em></h3><p>“I don’t really think it makes that much difference, as long as it’s mathematically quite rigorous. We’re definitely looking for people with a hard science background, but I would count econometrics as part of that as well. But if you have the choice of taking maybe some more math-oriented programming, numerical methods type courses, then I would definitely recommend that.”</p><h3><em>Does the job require knowing the theoretical foundation in measure theory?</em></h3><p>“No, it’s applied. If you know how to derive Itô’s lemma and you know how to do measure changes and this sort of thing, that’s enough.”</p><h3><em>What are juniors good at, and what takes them time to adjust to?</em></h3><p>“I think what juniors are good at is they bring a fresh perspective. For example, machine learning for people of my generation is quite a foreign concept. And if you get somebody new in that has actually had a training in data science, which, when I was your age, did not exist, you get a completely different perspective. [...] What I’ve noticed is that people are no longer very patient anymore. In the old days, I guess, people understood that in order for you to really learn the tricks of the trade, it takes time. [...] We invest a lot of time and effort into training people, and the last thing we want is that after one year or two years they leave.”</p><h3><em>How complicated are the models you use?</em></h3><p>“The fewer the better. We tend to use, as much as possible, market standard models. Some of the more complex models we have are for so-called XVA adjustments, where you take into account counterparty credit risk. [...] Maybe with JP Morgan we’ll be trading equity derivatives, commodity derivatives, interest rate swaps, whatever it is, and we need to simulate all that together to look at what is our exposure to JP Morgan next year.”</p><h3><em>Which instruments does the team focus on the most?</em></h3><p>“ING is a commercial bank, meaning that the core of our business is lending. That’s what ING is good at. And so the products we offer are in the service of lending. [...] Many corporate customers want to lock in a particular interest rate. Instead of paying a floating rate, they want a fixed rate. So we do interest rate swaps. It’s not a very complex product, but that’s by far the biggest.”</p><h3><em>How is performance measured?</em></h3><p>“We do clearly set targets for everybody at the start of the year, and we try to define as clearly as possible how we will measure things at the end of the year. It is not always that easy to really say, okay, you need to write this many lines of code, because that doesn’t really make any sense. [...] We will go and talk to the head of the rates desk and ask them, what’s your experience of working with this person in our team?”</p><h3><em><strong>Do you prefer candidates with a PhD?</strong></em></h3><p>"About 20% of the team have one, I would say. [...] I think people who have done a PhD tend to be a bit more independent, in the sense that they're able to manage quite complex projects, which can last for quite a while on their own. [...] So what we find is that this extra four or five years that people spend doing that definitely carries a premium. But then on the other hand, look at my own experience: you spend probably 10 years doing all kinds of things in physics that I probably never used again. So maybe it's actually better to have someone that has a bit of work experience. [...] If you want to do a PhD, do it for the right reasons, because you're genuinely interested in the subject, without thinking too much about what your career path will be like afterwards. If you want to start working as soon as possible, then maybe not doing that is a better option."</p></div>
