AFAM: Hello, Sacha, and thank you for participating in this interview series. Please tell us a few words about yourself. What are you doing today at Kaiko and what is this company about?

 

Sacha: Thank you for your invitation and for making this series!  During my time at Arts et Métiers, I did an “S8” semester (which I think may not exist anymore) in Spain at UPV, then worked on wind-tunnel experiments for a space shuttle at the von Karman Institute in Belgium, moved to Paris to complete a double degree with a Research Master in Fluid Dynamics and Aeroacoustics during which I worked at Safran on the future of commercial aviation propulsion systems, and then moved to London to pursue a PhD sponsored by Innovate UK and Airbus at Imperial College on the physics of turbulence for drag reduction.  After this, I embarked on a journey to entrepreneurship that started at an AI venture founded by a friend from Imperial, then went on to study business and finance at MIT before a failed attempt to start my own company in the space of financial inclusion.  Right at that time, Kaiko was closing a seed round of financing and looking to expand to the US market.  Kaiko is a market data provider that focuses on building reliable data infrastructure for crypto currencies, we sit at the intersection of finance, crypto, and tech, and operate both in the centralized and decentralized ecosystems.  My role has been to open our US office and work with our engineering and business teams to improve our product as well as initiate our research efforts, in order to showcase our data and help clients make valuable use of it.

 

AFAM: From your LinkedIn profile we learned that prior to joining the crypto space, you were running computer simulations for the Aerospace industry. How can you explain this change? What was your motivation to switch to the financial sector? And how did you get where are you now?

Sacha: The funny thing is that I was technically running crypto calculations slightly before I got to get into Aerospace simulations. I first heard about a new electronic cash protocol on the amazing podcast Security Now (to which I still listen regularly, I highly recommend it!), it then took me some time to get to it and it was already too late to mine Bitcoin on my cheap personal laptop when I eventually did during my time at Arts et Métiers (my laptop had about 30min of battery life running CATIA V5), still, I did manage to mine some Dogecoin!  It did not take long though until I decided to dedicate this CPU usage to more scientifically productive calculations like studying the flow around a cylinder using a Boltzmann solver.

The desire to move to the financial sector arose slowly over time.  I guess the 2008 crisis occurring when I was in high school did not act as a catalyst when I was making long-term plans of what my career would look like after starting my undergrad.  Later, during my Research Master, I joined a team of engineers at Safran Aircraft Engines in the Paris area to work on aerodynamic performance of propellers.  I really loved this time spent at Safran where the team was outstanding both on management and engineering sides.  At the same time, I started wanting to understand how corporate decisions were made, for example: “Should we improve our existing engine, or should we build a new one with a new technology?”  Of course, real-life questions are not as simple and binary as this, especially given the long development cycles seen in aerospace, but it does illustrate when you try to come up with an answer that engineering is only one piece of the puzzle.  As frustrating as it may have been at the time for me to realize, most of this answer may come from a subtle combination of politics (public and internal), regulatory constraints, competitive dynamics, and (surprisingly for me back then) long-term predictions of oil prices.

I took a break from those considerations during my PhD at Imperial College London and shifted my focus to bringing a solid contribution to science.  Although it was a challenging time, just as it should, I count those years as some of the best of my life.  Little economic considerations, no politics (apart from a sudden exit from the European Union…), two incredible PhD advisors, life-long friends amongst the PhD cohort, the only things that mattered were intellectual curiosity and understanding the world we live in.  “The world we live in,” or perhaps more realistically a pixel of its infinite picture.  I was running numerical experiments on turbulence; it would take days of computing on tens of thousands of CPUs but then you would get a simulation so precise that the statistics would almost perfectly match a corresponding real-life experiment.  I felt I had a distilled a little bit of nature and I was running it from my computer.  I could visualize it in real time, pause it, then restart it, kill it, nurture it, all this from the tip of my fingers.  Sometimes the turbulence would die in the simulation so you would have to crank it up a little, and then I would check at night that the vortices looked healthy in the simulation, making sure the flow made its “transition” to a turbulent state so that I would have some statistics to aggregate the following morning.

Speaking of transitions, how did I make mine to finance?  After a rather unexpected series of events, my interest in the venture world led me to go from Imperial College, the birthplace of the most popular turbulence model (k-epsilon) to MIT, the birthplace of the most popular financial model (Black–Scholes–Merton).  At the end of the day, investing is about understanding how the world and our society work.  Understanding finance is essential if you wish to lead impactful projects, understand how decisions are made, and the cherry on top is that you get to work with passionate and smart people.  I believe that blockchain and crypto currencies make for a compelling narrative that acts as a catalyst to apply cryptographic technologies at unprecedented scales.  As an aside, I have not closed the door to aerospace, and I am confident that much of what I am learning now will be transferable if I have the opportunity to get back to it one day.

 

AFAM: What do you think is the best way or one of the ways to start in financial engineering especially for our students at Arts et Metiers with a major focus on mechanical and industrial engineering?

Sacha: There are many paths to financial engineering and many skills learnt in mechanical and industrial engineering are transferable to finance. There are a lot of data science techniques that are very much transverse between engineering and finance, just beware that the signal-to-noise ratio can be frustratingly lower in finance.  If you know how to perform a statistical test to ensure that your production is meeting your quality requirements, you can apply that to finance.  Having a strong understanding of statistics will open many doors in financial engineering.  In addition to the buzz words like machine learning and AI, numerical methods also come handy: The methods for solving partial differential equations are the same in engineering and finance, just as you are solving a constrained optimization problem for solving the production planning of your supply chain, you can do the same to optimize a portfolio of stocks.  However, bear in mind that there is more to finance than financial engineering and there are other areas of finance that may appeal to you.

AFAM: Andrew Sheng, a Hong-Kong based Malaysian Chinese banker, named by “Time” in 2013 “One of the 100 most influential people in the world” in an interview for the 2010 financial industry documentary “Inside Job” said: "Why should a financial engineer be paid four to a hundred times more than a real engineer? A real engineer builds bridges. A financial engineer builds dreams and, when those dreams turn out to be nightmares, other people pay for it."  What do you think about it?

Sacha: My first reaction to this would be to think of scalability.  When you build a bridge, a lot of the work you have to do is bespoke, less scalable, and the risks you bear are much more visible and understood in general.  Engineers have spent centuries understanding bridge failures, yet we don’t understand everything: the Tacoma bridge collapsed in the 1940s due to aeroelastic instabilities.  Since then advances have been made and the revolution in computer simulation is for sure contributing to the improvement of the models being used to prevent this from happening again.  However, faulty engineering does also come with consequences (material, social, and financial), and other people are paying for them.  Don’t get me wrong though: I am not implying that engineers should pay for it!

On the other end, financial engineers have their own models to describe the arcane financial system, they build bridges between risk pools so that projects can happen.  When a crisis strikes, some of those bridges collapse and, in part under the pressure of regulators, financial engineers update their models to prevent similar collapses from happening again.  Scalability means you can easily replicate those bridges in other places, however it also means that they can all blow up at the same time when a vulnerability brought to light.

I would argue that engineering has been catching up, it is scaling and starting to face some similar issues.  We see this with artificial intelligence and computer security.  Disparities consequently arise in salaries: take a PhD in Computer Vision and compare their salary to a mechanical engineer working on bridge design, I suspect they are not of the same order of magnitude.  Pushing this scalability one step further will take you back to the previous series of interviews you did on People and Robots: add robots into the mix and you introduce scalability for building and servicing the real world too!

Humans work and operate in networks; modern technology brought us from advancing at the speed of evolution and generation, to advancing at the speed of thought.  In a sense this is good for Progress (I try to remain optimistic on this), but sadly, network effects increase disparities and a simple thought experiment suffices to illustrate this point: imagine you get valued by the number of people you talk to.  Now just think of the difference between an Instagram influencer with millions of followers and someone living a happy life in the countryside mostly interacting with their close family and friends, you will find many orders of magnitude between the two.  Going back to the physics of turbulence, I would say that reducing friction increases the separation between scales, but I won’t bore you with this any further.

AFAM: financial sector is very popular among our students today. What would be your advice to those who will be looking for internships in a financial sector (including in the USA in 2021 as soon as the situation with visas is resolved)? 

Sacha: The first thing is to have a general understanding of the many facets of finance.  For this, I encourage students to reach out to alumni working in the space to learn how they got where they are and what they are working on (do not forget to contact Shasta program first!).  There are many ways to get exposure to finance especially quantitative finance, you can look at open projects, many platforms allow you to test algorithm without putting any real money behind it.  Having skin in the game is key for a financial practitioner but there is a lot you can learn before getting to this point.  Also have a look at Numerai, and if you are interested in the quant side, there are now many other resources available online.  Two other projects worth mentioning are Quantopian and Quandl.

Finance is as broad as its applications.  Engineering skills are particularly valued in quantitative finance but there are other areas of finance like Venture Capital that can provide an interesting blend of finance and engineering where your technical expertise in a field can be an edge.

Finally, I’d say find something you are passionate about and start building your professional network early with an open mind.  Be ready to adapt, the rate of change is accelerating.  Last but not least, in whatever you engage in, always remember that building for a better world is far more rewarding in the long term!

AFAM: Thank you for your time, Sacha!

Photo: courtesy of Sacha and Kaiko

More about Kaiko:

https://www.kaiko.com

Other resources:

Security Now podcast

Quantopian

Quandl