084: Yves Hilpisch – Quantitative Finance and Programming Trading Strategies w/ The Python Quants

Chat with Traders is brought to you by Trade the Pool. Did you know that every decade the market reinvents itself? Online brokers opened the doors. Mobile apps made trading seamless. and commission free trading erased barriers. Now a new era has begun. Meet, trade the pool, limited risk trading. And now you also have unlimited time to reach the profit target. From now on, your trading risk is capped. and your trading opportunities are limitless.

Trade the pool funds home-based stock traders with up to$200,000 in buying power. That means you can trade larger positions and scale your strategies without risking your own savings. It's time to trade with more capital, making it truly worth your time and effort. Ready to trade the pool? Click the link in the description and join the stock trading revolution today.

Are you ready to get serious about trading? Then join Tasty Trade, Investopedia's best platform for options trading in twenty twenty six. Options, futures, and more. Tasty Trade has everything you trade all in one platform. Get low commissions, including zero commissions on stocks. So you can keep more of what you earn. Trade smarter with advanced charting tools, a pre-built strategy selector, risk analysis tools, and more features. Visit Tastytrade.com slash.

Chat for more information. Tasty Trade Inc. is a registered broker dealer and member of FINRA, NFA, and SIPC. This is the first one. Podcast. What is going on? you here thanks a lot for tuning in welcome to another episode. Before we get into this episode I have a small favor to ask from you. Okay, so here it goes. I would love it if you could please take just a few minutes to complete a very brief, totally anonymous survey. Now I'll tell you why I'm running this survey.

As you know, the podcast is free to listen to any time on demand. The way the podcast is supported financially is thanks to sponsors. And as you can imagine, sponsors want to have a pretty good idea on who their message is going out to. And of course, I just want to get to know you a little better as well. So if you can spare just a couple minutes.

Please go to chatwithraders.com forward slash survey and complete, I think it's about 15 questions. They're all multi-choice answers. It's very easy and like I said, it'll take you no more than a couple minutes. So that's at chatwithraders.com forward slash survey. And just because I really do appreciate you taking the time to do this and ultimately show your support.

I'm going to give you a free gift. Um, I'm not gonna tell you what it is right now, but if you go to chatwithraders.com forward slash survey, uh, everything is there. So if you can support the podcast. By completing the survey, please do that. I'd be hugely grateful.

Okay, so let me tell you a little bit about this week's episode. Uh it was actually recorded back in March. I'm not sure why it's taken me so long to release this one, but anyway, here it is, and it's with Dr. Eve Hilpish, the founder of the Python Quant. The Python quants do a lot of good for those involved in quantitative finance.

As such, they frequently host meetups and workshops, have developed platforms and analytics libraries, and often contract to exchanges, banks, and hedge funds for custom Python development. Eve is also a three-time published author, with his most notable title probably being Python for Finance, which was released through O'Reilly. He regularly gives presentations and speaks at events on the subject of quant finance. and lectures at universities too.

Needless to say, he is a busy man, but I was thrilled to lock him down for the podcast. So over the next 60 minutes you'll hear us unpack many subjects related to being a quant and why programming in Python can be a useful skill to have in your toolbox. Now I will mention Some of the discussion in the later part may be a little heavy for non-programmers.

So if there is something that doesn't make sense or you'd like more context around, please just write in the comments at chatwithraders.com forward slash eighty four and I'll either do my best to answer this for you or point you in the right direction with a link, or I may even ask Eve if he can respond himself. Anyway, I hope you can take something away from this. I'm Aaron Firefield. Here is Eve Hillpitch.

Hey Eve, welcome to the Chat with Traders podcast. Thank you so much for being here. How's it going? Good. Well, very well. Thank you for having me. It's a great pleasure. No trouble whatsoever. Now, I know you're not exactly a trader, but you are heavily involved in quantitative finance. And you're also a huge advocate for the Python programming language. So I mean guys listening

Know that I've been, you know, learning Python myself over the last kind of six to eight months, give or take. So maybe the reason your hair is slightly selfish, but nevertheless I think you'll be able to share some awesome insight for listeners, especially from a quant perspective as well. So I'd like to start by talking about how you originally got into finance. So take us back, where did the interest come from?

What is kind of interesting, when I was uh pretty young I wanted to become a medical doctor, uh actually, so but my uh girlfriend back then and wife right now, actually she was interested in in business administration. So I I started to get interested as well. uh reading stuff she brought back home, so to say, about business and finance and things like that. This was actually the first time I was like seventeen or something.

uh when I actually heard for the first time about that topic that you can study that and so forth. For me, my family was all about medical uh discipline more or less. Um and later on actually when I started out really wasn't finance as we understand it as of today, but uh I started selling insurance policies back then which had kind of uh an investment component so I started like uh following

uh financial markets, it was really intrigued by asset management, by mutual funds and savings products and so forth. And this was actually the first point. Uh this led me to later on during my studies uh to focus on financial market theory, banking, uh uh actually every everything I did during my my diploma uh with uh a major of business uh was about finance, financial markets and so forth. So started out quite early and it led me to yeah, focusing on this uh on this area actually.

Interesting, interesting. Okay. So just so we're clear, you graduated with a diploma in what was it, finance? Business business administration, but I had a strong focus on on economics, financial economics as it was called. Um, I later on moved on and did my PhD in uh mathematical finance. So I then specialized even more obviously, started doing uh option pricing series, dynamic hatching strategies and stuff like that. So

It was kind of a a continuum of what I did. So Right, okay. So when you did your PhD, what were your intentions um of studying for that? Like w how did you plan on using this after you graduated? Actually I hadn't any concrete plans for or to use the stuff that I did during my studies was really that I really enjoyed doing uh this uh PhD work and uh fiercely wanted to do it so there was no question of not doing it.

I simply did it. Um and during this time I did it kind of in external uh mode, uh which is something you can do in Germany actually, uh where you do part time at the university and then you can do other work or whatsoever. I started working during that time So roughly a total of across total of four years I started working as a management consultant uh for financial companies.

Uh so it wasn't really in this hardcore computational quantitative finance area, uh but still it was in the financial industry. Um and only later on I I merged the two areas, if you like, my professional background of consulting, management consulting work and quant finance and obviously these days technology to to what we do uh today. But this took quite a while, a couple of years.

Okay, sure. So tell us about that first job you had, like what was your role when you did finally, you know, merge those two together and you were starting to put your quantitative knowledge to use? It was actually for the the first three of my first three years of my uh professional working life, if you like. uh where as I said in the management consulting area and I did this for for quite a while afterwards when we founded our own company, the first one in two thousand one already.

uh I got on my own with the two partners and we we still did the same things but I already started back then uh bringing more and more quantitative things into what we did. Uh and now it's kind of uh yeah, it was a credible uh transition but now I'm on the uh more or less completely quantitative uh side and more technology than ever. Um so it's kind of um not kind of a straightforward um Career path if you like. Uh but this Python for Financing uh started out

than almost yeah, ten years ago. R roughly ten years ago where we started implementing our first uh solutions using Python. Back then in the days where we even And didn't have a numpy, it was called numeric back then and so forth. You know, we we started out quite early where people said

Python is not really well suited uh to do finance at all, actually. That's when we started out doing it. Right. Okay. Well bring us up to speed even and tell us about what you're doing today. Like what's your what's your current involvement in the industry?

Well I guess uh uh one of the most uh popular and and famous things I guess w what we produce, what I produce is my Python for finance book for O'Reilly. And I think these are the two things that uh make up what we do on a daily basis. So on the one hand side it's Python, and on the other hand it's finance. Actually we work usually on a contractual basis for financial clients. This ranges from

banks, to asset managers, to hedge funds, to exchanges. Exchanges have been quite a quite a bit of a focus recently and over the years. um but almost any kind of financial uh um company that we work on uh projects for them, obviously with Python being involved. So what we don't do is kind of

um do other projects just for the sake of the project and then doing uh focus uh on C plus plus for example in this regard. So what we do, what we are good at And there I guess we cover the whole value chain if you like, from trainings to books to platform, to our library, to the services, which means the consulting and development, this is Python and Python for quantitative finance. And if you want to get a little bit more specific

Uh what we focus on is what I usually call computational finance and financial data science. These might be the best uh two sub disciplines to describe what we do in quantitative finance. Cool. Okay. Well could you give us perhaps Um just an example of some of the projects you might work on on the consultancy side of things.

Yeah, obviously uh more often than not, uh we we sign some NDA uh contracts, or obviously uh non disclosure agreements where we are not allowed to speak about the things we do, but there's Uh at least one good thing that we can talk about. This is something we did for UREX, uh our German derivatives exchange, which is a part of Deutsche Börse actually. Um so um there we did a Python-based tutorial series.

Uh well, we did a couple of things for them. So we wrote a tutorial about the VStocks volatility index and how it is calculated.

uh what you need to do when you wanna implement this in Python, how to recalculate historical values and so forth. It's kind of a very involved thing using lots of option pricing theories and and Not rather complicated but still involved financial algorithms and we and we wrote this tutorial like in a in a book style and actually right now I transformed it to a book so it will come out soon uh and therefore I can speak about all this because it's public and it will be published as a book.

Um and uh not only this, what we also did we wrote uh a backtesting application for them, uh where you can back test strategies uh for the whole history of the VStocks index involving VStocks options, VStocks futures and so forth. And in t in a tutorial we explain the strategies that it can use, we explain how to value futures, how to value options.

Uh and again this is kind of a public thing. Everybody interested uh in it uh can Google it up. It's called um VStocks and and uh variants advanced services. So if you Google URX advanced services you will find uh the right place to start and to go from there. Okay, well we'll dig up a link and we'll be sure to include that in the show notes at chatwithraders.com.

So I know you've been working with asymmetric return capital a little bit recently. Can you tell us any more about what you're doing with them? Yeah, this is kind of an uh emerging asset manager, a hedge fund, uh which is focused on uh volatility uh trading strategies. So I was talking about the V stocks and various things we've done. So there's kind of a few synergies involved in what we do there.

And even more synergies uh with regard to that we apply for them and is one of our first uh big users, at least uh as I'm aware of, because uh our DX analytics library is open source so We don't even know who is using it uh for what purposes, but here we work with our client in a very um A close fashion to apply our DX analytics suite. We should include the link as well. So it's dx-analytics.com.

Uh it's open source so you can download and test it and we use it there for advanced derivatives, analytics and risk management purposes. Uh we have built kind of uh in addition to what the open source library provides, kind of a a complete system around that to to really make use on a daily efficient uh bases that uh when the decision makers like Prime Risk was the founder I'm working with uh since more than two years right now.

uh needs to make a decision, then he has kind of real time uh capabilities to apply what we have uh built with uh DXN. And this is kind of an interesting company to mention in the context we're discussing here because uh more or less uh you can say that R Is is a Python based hedge fund if if you like. So um one of the the the co founders there actually Adam gave a talk at our conference last year in Manhattan where he spoke about how to build a hedge fund with Python. So

This is actually a a very good um example for what we discuss here. Interesting. Is that is that talk available online anywhere? Uh this is not available online, so uh we do the for Python quantities, maybe maybe we cover this uh later on, uh in cooperation with uh Fitch. So for those who have signed up for the online um to get They have access to it but it's not publicly available or forcibly depending on what perspective you have. Got it. Okay. Cool.

So on an individual level, like for yourself, do you actually participate in markets in any way? Are you, you know, actively trading I know you're not actively trading, of course, but um are you investing in in any way, shape or form? Yeah, I mean what I do is kind of more or less for retirement purposes that I have my continuous savings plans and so forth, uh uh in this regard investing in in mutual funds and different uh asset classes

Well what I'm not doing is kind of uh day trading work or whatsoever. I mean I'm involved in so many things and and travelling quite a bit, so sitting sitting on an airplane for eight hours would make me nervous if something would go wrong. And on the other hand, obviously, uh being involved with so many financial institutions which have uh obviously heavy uh heavy trading uh um shops um

and businesses uh makes it uh also a little bit difficult depending on the work that we do, how we get involved and things. And usually we are quite close, for example with what I described earlier with ARK uh and so uh there might be also uh conflicts of interest. But this is not the main reason. The main reason is that

Uh A I'm running uh a few businesses, not only the Python quant, a few other side businesses as well. I'm travelling quite a bit, so I decided for myself don't get even more nervous and and uh have even more workloads uh being involved in the markets on a daily basis, I think. Sure, sure. That makes sense. Now let's spend a little time actually talking about quantitative finance. So

First of all, what does the term quant refer to and how does a quant vary from other types of market participants? Yeah, I mean the term quant is uh as far as I see it used in many different ways these days. So uh maybe in a more narrow sense it's it describes the quantitative analyst.

to certain extents, whereby I mean, uh not an analyst starting out in the business, but an analyst analyzing anything with quantitative means, which means typically lots of numbers are involved, lots of data is involved and and you get uh down to the numerics and to the to the many, many digits that might be important in this regard. So

A quant in a bank uh usually can also have different roles. So you might have uh a pricing quant, which is uh responsible for the pricing of derivative instruments, you have a risk quant.

which is small as in the back office, for example, uh responsible for things going on overnight when it comes to value at risk or X VA calculation. And so you have many types of quants. Uh but in the end what uh what every type has uh in common is that they work with uh with uh many, many uh let's say numbers with data, with formulas, with math and and try to make sense of the markets in the end.

Um and from my point of view, what am what I mentioned before, in a sense of that we do computational finance and financial data science, um there might be a third major world where we are not involved that much is kind of the typical the model quant.

Uh I guess this is the starting point where it all started out with the rocket scientists on Wall Street and maybe the eighties, where they hired physicists and mathematicians uh to come up with better models for instruments already traded, but where sometimes nobody knew for sure how this is to be valued.

Um but I guess when it comes to the to the ratio to proportion of the model quants in relation to the whole quant universe, uh this is less than one percent I would say, people focusing today of uh building and coming up with new models. Uh the majority of people is more or less concerned with implem implementing what the models say and what you have to do in terms of compliance and trading and so forth.

Uh then we come to computational finance. How do you simulate a stochastic differential equation? Um how do you apprice this with numerical methods? Uh and the guys working on the financial data science, more or less making sure that all data is available, the data is consistent.

that it is processed in in an an appropriate way. So this is more or less my world of thinking. Other people might have a different world of thinking and they can cut it differently, uh but from my point of view, I usually think of the model quant, the computational finance guy and the um

and the uh data people, the financial data science people that take care that uh everybody gets the data uh who needs it and in time and with the right quality. Okay. That's a great answer. And I like how you mentioned that. there seem to be many different types of quants because I mean it's kind of become obvious to me that, you know, there's the guys who are very deep into the math and they're coming up with the formulas

And then you have those who are more focused on just analysing big data to develop, you know, algorithmic trading strategies. So no, that was that was a brilliant answer, Eve.

How would you say a quant looks for opportunity? Like just uh generally speaking, how does a quant look for opportunity and areas where they can exploit the market? Yeah, I mean there are quite a few areas uh work once are working on. So if you have a look at the history of kind of investment

Um I mean it all started out more or less with kind of fundamental research combined with speculation. So we always had this kind of Uh yeah, groups of people were the one group was saying, well, we have to deeply analyze things that are going on, the fundamentals of the company, what are the prospects, how does the market look like and so forth.

Uh more or less others that said well, I'm in the markets but I'm simply training. So uh maybe the a great recession in the twenties in the US and so forth uh is like kind of uh a picture of what what does uh what this kind of uh situation might lead to. Uh these days we have many, many more things and and this is where the quants come into play because neither fundamental analysis nor the speculators are That quantity if you like.

Um it more or less started out with the uh Markowitz theory, uh the portfolio theory based on mean variance. But for the first time someone came up and said, Well, let's have a a quantity look at how things are going. uh to keep the word simple and this was already a major breakthrough, uh we focus on certain uh statistics that characterize different stocks actually.

as the expected mean return, as the to be expected volatility. And let's take these two parameters. I mean it's a little bit too simplifying, but still use these days heavily. But let's take these parameters and let's build kind of

uh mathematically optimized portfolios based on our criteria. I I guess this is more or less uh the starting point of of quant finance uh when you have a look at the history. Uh only later on uh we have the other area where you say, well, uh with the Plex Colts theory and option pricing, where we t then start to talk about derivatives, which means uh securities whose price, whose value is derived from

something else, kind of a stock, an index or whatsoever. Um I guess there maybe people started talking about quant, but more or less for me, portfolio theory, mean variance, uh the CAP M for example, are all examples of quant work as well. Coming to the option pricing theory of Plex Cole's um um uh which uh later on awarded them the the uh Nobel Prize for Economic.

Actually uh we're at a point where we say now we need to get back uh to history because uh something like Black Scrolls theory was uh already invented by uh a call a guy called Bachelier, a French guy who wrote his PhD thesis about similar topics. in nineteen hundred but his PhD thesis. in math actually Ben, about speculation and how to price uh options with arithmetic uh brownie motion was only fifty years plus later on discovered. So nobody was aware of of what what he already invented.

for for multiple decades actually. So maybe this was the first real quant, if you like, in our on our model quant definition as of today. Um so we have the people who try to optimize portfolios with regard to certain mathematical methods, quanti methods, we have uh the option the model quant guys. Later on, then the emergence of uh trading strategies where statistics is used, for example, statistical arbitrage.

When it comes to derivatives, the theory at least goes that when you use a strong arbitrage argument, there is kind of a Existing provable mathematical relationship between A and B. So you have option pricing theory, and this gives you the relationship. There should be no doubt in theory, but in practice obviously there is a lot of doubt and and a lot of uh yeah, market so to say, uh playing into that game.

Uh but the statistical arbitrage you are aware of that you might wanna exploit kind of a historical relationship with regard to correlation or whatsoever. Um Where I say well statistically if you traded this and that way like

gold versus gold miners, for example, is a classical trade, um, then we should make money because now we have kind of something which is statistically beyond the normal, beyond one standard deviation whatsoever. We might speculate off uh bring this back to the to the historical means or the to the historical values that we observe.

Um and this goes on until today where we have uh I mean there are many disciplines and sub disciplines in between that, uh where we have today the high frequency traders which are not exploiting kind of financial relationships more or less, but they are trying to exploit the market microstructure.

applaying the system more or less and not going about financial uh economical principles if you like. So whatever what I'm always saying, the whole financial theory when we talk about quant stuff is more or less based on daily historical closing values of NDC stocks and so forth. Uh but what these people are doing right now is still lacking kind of a financial uh background, financial theory j which can compare to the uh daily uh closing price based uh theories that we have.

So many, many types uh of uh of strategies that you can use to exploit the market.

Right. So just narrowing in on the topic of research here. One thing I'd like to ask you about is do you utilize uh machine learning in any way during your your research uh phase when you're looking for um opportunities within markets?

So I must so to say in this regard confess that we are uh due to what we do, uh derivatives, analytics, option pricing and so forth, we are what I mentioned before on the site where we at least think that we are uh in a position where we can more or less Derive prices from what we observe in the market in a not deterministic fashion, but in a rather robust fashion. uh and can prove this mathematically, so to say.

Um so in my daily work uh at least a few machine learning approaches are deployed, but this is nothing which represents kind of a focus as of these days. I know many people are kind of eager to use it and I even get approached by people I don't know from around the globe actually.

um who are who are interested in this topic about applying machine learning to finance and what this implies and so forth. And I mean these days if you hear about Alpha Go uh being one of the greatest masters of the Go game and so forth

Artificial intelligence, deep learning, machine learning, I mean all these disciplines people are so fascinated about and everybody is is uh thinking of well we must have something in there in this uh m r big toolbox which you can apply to financial markets and I'm pretty sure there is something but Frankie, I have at least for my daily work I haven't seen anything where I say, Well, this is kind of a a real area where we get major benefits of applying this or the other technique.

uh to what we do uh and if you speak about we, I'm not speaking for the whole uh industry, I'm speaking for for our uh focus areas actually. Okay. It's an interesting answer. It's interesting. So Still on the same topic here, is everything you do is it based purely on price or do you use other things in combination such as uh fundamental and social data, do they come into the picture at all?

I mean uh generally speaking, obviously what you're mentioning here, price, fundamental, social data uh are all sources that are used um in the industry to base investment decisions on and so forth. Uh but again I would say what we do is at least ninety percent based on

price or market data which is readily observable and where there is typically no doubt about it. Uh so if you think about fundamentals and price earnings ratios which are forecasted and which then influence, you know, there's kind of still lots of uncertainty. uh around that, uh social data as well, about the quality and so forth, all the the people that try to make sense of Twitter data and so forth. But again, I've seen so many talks about this over the last two, three, four years.

Um, but still I I'm not aware of anybody who is using this kind of as a major tool and make uh significant alpha, if you like, as we call it in the industry, uh out of that. Th there might be some people who don't speak about it. Um there are also companies, uh startups that focus completely on that. Uh but I but I'm not aware, at least from what I've seen in in terms of talks and read papers about that, uh where it is. But

As a supporting thing, um I mean this might be really valuable in terms of like risk management. I've seen a talk recently where they said, Well, uh we don't wanna exploit the markets based on social data, but for example we wanna generate kind of uh uh risk relevant information. And this was a bank speaking about that in the sense that they said, well we are lending to so many companies, but we want to exploit the social media to get kind of early indicators with regard to uh the risk.

for default for loans in our portfolio. I mean this makes sense. I there there I can see kind of a strong connection when But many people are tweeting and traders are tweeting and speaking about oh well this company and there's something I don't know you we don't know there's a risk uh on the horizon. Then this might be valuable information for a bank to say, well, we must speak to these people because there might be something at risk.

Uh but this is on a on a longer term scale. It's not like in financial markets that there's kind of an idea and it's exploited by twenty players, bigger players in the market, and then the opportunity has vanished. uh risk of a loan. I mean this is more like a thing over months, quarters or even years where this information might be irrelevant. But I wanna add one thing and there I see kind of a huge opportunity and I'm working close Actually what they are now putting out

It's now in the beta testing phase is kind of uh first of all they they implemented a unified API to the whole data universe, which means historical data, tick data, as well as streaming data. So and that data universe is huge. Uh and around this unified

API actually you you have a Python wrapper available these days. And uh this is the reason why I'm coming to this point. And with this Python wrapper to the unified API, you not only get prices for an index or whatsoever, you get more or less all the fundamentals that you need. need for a company. And you get access to news data, to social data as well. So having this kind of unified tool

um to process such data on l on a large scale. I mean just think of a trader sitting in front of his Bloomberg and top of the Reuters terminal um and trying to figure out this manually. This is It's shit impossible, this doesn't work. But having this unified API and and Python as a language, it's not the only language, uh but one of the languages they support is Python obviously. But having Python available, for example, and being able to now easily combine all the information.

This might bring up something where people say, Well, now we got it. Now we we crack the code and now we got the practice. through with regard to combining all the different uh sources and different types of data that is available. Okay, so is that API available to everyone or is it a a premium service? Like how do you get access to that? Actually not available uh obviously to everyone, but if you are a subscriber to the Thomson Reuters icon uh terminal.

uh then this is uh actually included. But you need to have the the data service, the terminal uh license itself and then you uh get access to the to the API and uh uh to the Python wrapper anyways. So but the API I guess is the is the new the big thing. Got it. Okay. That's very cool. That's awesome.

Yeah, for research it's kind of uh this will be a quantum leap actually from my point of view. Yeah, no doubt. Are you ready to get serious about trading? Then join Tasty Trade, Investopedia's best platform for options trading in twenty twenty six. Stocks, options, futures, and more. Tasty Trade has everything you trade all in one platform. Get low commissions, including zero commission on stocks.

so you can keep more of what you earn. Tasty Trade is packed with advanced charting tools, backtesting, a pre-built strategy selector, risk analysis tools, and more features to help you trade smarter. See equities and derivatives with high trading volumes, dividends, upcoming earnings reports, and more with their pre built watches. Or create a custom watch list to keep an eye on the companies and sectors that matter to you. Manage your positions with speed and precision using Active Trader Mode.

one click trading and smart order tracking. Plus Tastytrad's Stellar Trade Desk team offers live support during trading hours if you need it. Visit Tastytrade.com slash chat for more info. KC Trade Inc. is a registered broker dealer and member of Finra, NFA and SIPC. Have you ever watched a stock explode and thought, if only I had the capital? Or sat on the sidelines because your account balance felt too small to matter. Good news.

With Trade the Pool's limited risk platform, you don't need millions or even thousands to start trading the U.S. stock market. Bypass the PDT and tap into over 12,000 U.S. listed equities. From penny stocks to big caps, ETFs, even the newest IPOs, and short anything you like, with zero locate or hard to borrow fees.

Start your evaluation, get funded with up to$200,000 in buying power so you can go big without risking your own savings. And now you can also have unlimited time to reach the profit target. It's a game changer. Not ready to trade yet? Trade the pool offers a free demo and educational resources. practice on live data, master the platform, and build confidence risk free before you even pay a cent. Click the link in the show notes to start trading with Trade the Pools Capital. So

Tell us a little bit about the Python quants. You mentioned it a little bit earlier, but let's let's spend a little time on it and um yeah, share with us where did the idea come from and and how'd this get started. Actually I mentioned that I founded my first company in two thousand one and for personal reasons I left this company so we are still friendly. The founders back then myself.

Um, and when I started my own company, I mean this was the point more or less when we started using Python. So roughly ten years ago. I started out with a company which is now called Python Quants. Back then it wasn't called Python Quantz, but we renamed it uh uh two years ago, two or three years ago to the Python Quantas. We used this name for quite a while but no or then back then we made it official.

Um, because this was actually or describes the work that we do. We we use Python to do quant work, uh what we're discussing. uh in this podcast actually. Um so uh this is where it all started out and I guess uh what we do right now and I mentioned it before briefly is that we try to cover the whole value chain in this niche market if you like. That people, no matter what uh kind of company they're working for, want to apply Python for their quantwork.

Uh we try to provide everything that they need, which starts obviously with the books, uh three books, my Python for finance book. uh with O'Reilly, there is the derivatives analytics with Python book, with Wiley finance, and the next one coming out, listed volatility and variance derivatives also with Wiley. Um everything obviously Python based. Uh you find all the codes and we are not providing the code.

In the book, we provide uh GitHub repositories providing the books, we provide a plat uh on our quant platform. Uh where you can easily register. We provide executable Jupyter notebooks and all the code. So uh this already the book example already illustrates what we all offer. So so the book. the codes, the QAM platform, also training around that if you like. Uh we organize uh events around that. So we have uh our next big event is um

is in May. It's the first week of May from the second to the sixth. We have five days for for Python Quant's conference in in uh New York, Manhattan. where we have four days of intensive training, starting with the introductory part, then we have a technical uh part which focuses on advanced libraries and performance Python, input output and so forth. We then have a full day of Python and Excel. So how to

uh best combined Python with Excel, which is still kind of a obviously a very popular tool in the financial industry. We then have also an algorithmic financial trading uh day. where we cover the basics from back testing, in and out sample and so forth, implementing algorithmic trading strategies and doing even live trading on the Rwanda platform. So people coming in in the morning, they will do their

uh automated algo trading in the evening. So I guess it's a very intense one day, but uh people can get back home and start their own trading strategies afterwards and then obviously we have the conference day. uh the the the sixth of May this year, uh where we have usually ten talks with high level uh speakers uh out of the industry uh using Python for quant finance and associated things like sometimes also people speaking about R and Julia.

uh which I consider also as very nice um uh open source technology in the space. Stuff we also use, especially R. as a complement uh for Python. In addition to that, just to make uh it complete, uh we also organise meetups. My biggest meetup group is the Python for Quant Finance meetup group in in London where we have roughly thirteen hundred people.

Um as of today we started a little bit more than two years ago and I see London alone is kind of uh although we are really niche, Python for quant finances, a double niche if you like. Um We are very active, usually having meetups with up to a hundred people, eight meetups a year, many other uh things that we do and so forth. So uh many, many things around everything with regard to technology, services that we provide, training, development, consulting.

um uh the books that we have and also the community aspects. So conferences uh the public trainings and also the meetup groups that we run. I'm also running something in in New York as well. Um so it's kind of yeah, in the financial centers of the world um doing Python for quant finance. And this year we will also uh conquer Asia as well starting in June when I will be in Singapore.

Um, so yeah, we try our best to spread the word and to help people. Yeah, that's awesome. And it you know, that five day event sounds really great. I wish I could make it along. Too bad I'm on the other side of the world, but um We're getting back to London as well. Usually we have the London one in November. So later in the year we are in uh in London and in May usually we are in in Manhattan. Right, okay. Cool. So

Who are the type of people who generally attend these types of um you know, your meetups and your events and that sort of thing? Are they people who are like professionals within the industry or do you get like quite a few independent traders and um yeah, w like who attends these types of events?

I mean it's a quite mix. Uh speaking of thirteen hundred people in one group, we of you obviously have a mix of starting with students interested in a topic or or PhD students, uh postdoc uh tour industry players, um startup people and from the fintech space. um people working for the biggest banks at all like Bank of America, JP Morgan, people from hedge funds, uh uh yeah, traders who trade on their own. So I I think we cover quite

Yeah, quite the the whole cross section of of people that you find that are either already using Python for quant finance or want to use it. So we have many, many people and I'm always asking, was it first time here and Uh do you use Python so the typical questions during the video and we we have at least twenty to thirty percent per meal where people coming uh uh for the first time and

don't haven't used uh Python really many some people say, Well, I have done a little bit, but now I wanna apply it really for my finance work. So Uh people who really want to get into that and and yeah, wanna make better use of of this uh fantastic ecosystem that we have for Python.

Okay, well let's spend a little time on Python as a language. So it's clear that Python is your weapon of choice. Why did you choose Python over other programming languages? I mean it's a little bit usually I start out by saying well there's so many languages in the world um but there is English.

Uh I mean speaking English as of today might be the best choice as a foreign language. If you're a native speaker, uh there's not that much of reasoning uh behind doing it because you simply do it. But um

If I could only learn one language, uh this these days it would probably be English. If I could only learn one programming language, it would probably be Python. Um I mean Python doesn't have anything where you say, well, this particular feature or this kind of syntax idiom or whatsoever you single out

um is not found in any other place. So Python is, from my point of view, nothing that special. It makes more or less the combination of so many things. On the one hand side obviously it is the syntax. Which is typically quite appealing.

two quants actually. So if you are mathematically inclined, if you like equations, if you if you know how to code LaTeX in order to produce nice looking documents with lots of formula, uh then you will like Python for sure because it's so close Um uh you would find Yeah, that that you are back home and not in a foreign country.

Uh many people say, Well, Python fits in your mind. Uh you have many languages where kind of the the barriers of entry are so high and you get frustrated only to get the most simple things running. Uh and this actually is not the case with Python. So you can easily start out. Uh it's for even for my son who is nine who's sometimes coding a little bit uh with my guidance.

Um also to the absolute expert programmer. So Python is kind of scalable and then Um but if I would have to single out one thing, why Python over other language and then in parentheses for finance I would say, well um It is probably the ecosystem that we have. All the libraries, everything that it can use in terms of third party libraries if you like.

Scikit Learn, Scikit image, you name it, uh all these uh fantastic libraries that you get for free and which are kind of very, very powerful and especially to single out in this ecosystem one, I guess, pandas for data crunching. number crunching, financial time series analysis, visualization and so forth is something um that brings many people to Python.

Uh the other thing is that actually iPython notebook uh in former days brought many people to Python, our nice web based, interactive uh uh development environment. Uh but this has become language agnostic with the Jupyter project. So um they now have different kernels running and there's no need to come to Python for Jupyter. You can do Jupyter with Julia R as well. So uh but back in the days it was kind of these single libraries which are powerful.

and actually some nice tools that we we had to offer and still have to offer obviously. So they haven't gone but they've become more or less uh uh language agnostic.

Right. Well I can tell you that I wish I certainly started programming when I was nine years old. I'd be much further ahead than where I am right now, but I guess uh now's better than never. So Now, you you mentioned right at the start there that there are, you know, so many languages out there that you could have chosen uh from, but you know, you you went with Python, um

Is it a popular choice used in the finance industry? Like obviously you use it in the finance industry. Is it Um is it like widely accepted in the U.S. one of the most popular languages uh at all these days. Many of the biggest banks in the world now uh do major things in Python. Uh to name two, there is uh Bank of America, Meryl Lynch. uh uh uh where they have uh implemented their core risk management and trading platform, which is called Quart. It's actually a big project.

Uh mainly in Python. Obviously typically doing these big projects uh you don't use a single language only. So you have uh typically a blend of technologies and languages that they use. But Python is the major language driving this big project. And the last number I heard is that uh it's uh with fourteen million lines of code, it's kind of the biggest uh financial system based on Python in production as of today. Um the other one is JP Morgan with Athena.

Um they are they're trading and also risk uh platform uh where Python also plays a major part. Um so big banks are using Python and you can name it, uh pick out any bank, Deutsche Bank and so forth, uh to some extent. They are using Python. I I I mentioned those who have it as a strategic platform technology. Others, for example, Deutsche Bank uses it in many, many places, but not maybe as a core technology if you like. Um so if we now go more to the asset management, the hedge fund side.

Um and I guess the the hedge funds have been here font runners. Uh you will find many, many of the biggest uh hedge funds in the world that make heavy use of Python. uh uh to name just a few, it's it's PET partners, it's AQR capital management, it's uh Two Sigma in New York. So uh some of the biggest hedge funds in the world use Python again as a strategic uh technology for what they do for research, production and so forth.

Um on the asset management side, this is my experience also with regard to the business and trainings that we do. is that asset management, uh the classical asset managers are now catching up and also have discovered that Python is kind of a good choice. But it's kind of a little bit beyond the wave. And what I see next is kind of

uh uh getting more and more questions from the insurance industry. Uh we haven't spoken about that and it's not our focus as of today, but I see more and more content there as well. Um and and in the end I guess Python will will be kind of a strategic technology for almost any uh company. And uh uh getting back to our uh introductory example of of our um or the uh emerging manager in New York. Uh actually uh and I can

There are many, many other examples that it is similar. Once you start out on your own these days, uh and you can start from scratch and don't have legacy codes, I don't know, ten million lines of C code in your pricing library. Then it makes sense to use Python. Uh ten years ago when we started out and did our first steps, it wouldn't have made sense uh to use Python for kind of everything uh and especially not for your heavy analytics task. But as of today, if you are an owner of company

um and it's your money and it's your time and and you have to make sure that everything runs and you do in the best way you can, then typically you choose Python. Because uh especially when it comes to analytics, I guess you have a factor uh probably uh of ten or sometimes twenty in terms of uh reduction of number of lines of code that need to be written to accomplish the same when you compare C plus plus and then moving to Python. Uh so you're much more efficient. You get

many more th or much more done in a much shorter period of time by using Python than other languages. But I I don't speak against other languages, don't get me wrong. I mean C plus and then all the the compiled and high performance languages they have all their place. Uh and Python can easily interact w with these languages and and and libraries uh implemented based on them.

Uh but when it comes to interactively explore stuff, prototype and bringing this as fast as possible and as efficient as possible to production, I guess Python is kind of the right choice.

Yeah. No, that's huge and uh awesome answer. Very insightful. In the past, you've given many talks, you know, over the years, you've you've given a lot of presentations. One of the common themes throughout these presentations seems to be open source. What's the reason for your big push on open source and being such a um, you know, big advocate for for technology and Code and software being open sourced.

Yeah I mean if you think back in history, um then you had a time where open source was something which was exotic. uh we say well uh well when we uh I started out back in in the eighties with the C sixty four and you had to uh to buy your games and this was really expensive, you know it's kinda

There wasn't open source available. You could you could type stuff that you found in in magazines or what you this was open source, but you had to type it by yourself. But I think now if as of far we have come and what this means for our Society, I would even like to say, when it comes to GitHub, for example. How many amazing projects do you find on GitHub these days where you can uh make uh use free of any charge where you can have a look at what's going on and uh where there is the opportunity

for other people to contribute and to improve that thing. And I wanna get back to the example that I mentioned before. Pandas from my point of view is kind of a paramount example in this regard. Uh when Wes McKinney started working on panels, he was employed as an analyst at AQR Capital Management. obviously one of the biggest uh hedge funds in the world there uh there uh his intention was to promote Python

uh but coming from R he was aware of that the data frame object actually which is so nice uh to use in R is missing. He said, well we need this in Python as well and so he started implementing pandas. and started build uh rebuilding if you like compared to R things that have been in R for quite a while and now we introduced it to Python. But this was kind of a proprietary project for AQR. And it was not kind of a core project. I mean by writing a data frame for Python

You don't make money per se for a hedge fund. You make money by how you use it, what what data you crunch and what kind of algorithms and insights you come up with. based on the data you are crunching. So they made the decision indeed to open source it. Um and now we have this is one of the biggest libraries at all. I don't know whether the the uh current PDF documentation stands at uh two thousand pages already. The last count I have in mind is probably eighteen hundred pages.

for this single library alone and it's so powerful and it's uh it has grown um so fast and it's become so huge and immensely uh beneficial for our ecosystem. Uh it Would never have been the case if AKR would have said, well, this is something proprietary that only we should use. uh we don't open source and nobody should learn about what we have in Python and so forth. No, they did from my point of view the the the the correct

uh decision and said, well, this is not our core. If we open source that and provide others with the benefits of what we have implemented, then we might benefit back by others contributing to it. Uh so we can use maybe features we haven't even thought about before, uh that others implement and we have kind of the benefits as well. And this is kind of the perfect win win situation that you see in so many

uh areas and and pandas in our ecosystem is kind of one of the major examples. So now Millions of people use it on a daily, weekly basis. um and have the benefits and AQR also has the benefit because the library is now a a hugely developed highly efficient thing with a gazillion of features Uh they never would have thought it that way. I mean, there are many things that they don't need, but uh overall it's kind of beneficial.

uh for our whole ecosystem, for the whole uh Python community. So this is this is what I think of open source where the major benefits are and there are many, many more in terms of what uh these days typically when you apply for a developer position you typically don't write any typical C V anymore. You just send the link to your GitHub repository and say, look what I've done so far.

in Python and JavaScript and C plus or whatsoever, here's what I can do, what I've worked on, and people can have a a real look at what Yeah, and not uh yeah uh a real look at real things and not at a C V who tries to uh sell something that is not there in the in the best possible manner.

But many others, uh many other arguments apply to this particular topic from my point of view. And it really highlights the the value of collaboration too. So While we're on this subject, I'm I'm keen to get your take and your views on What do you make of the what are the what are your thoughts on the pros and cons of open source languages versus uh traders who use proprietary languages that are native to individual platforms?

Like what are the pros and cons between those two kind of approaches to to programming, if you want to call it that? Yeah, I mean obviously when you when you learn kind of a general purpose language like Python you can apply it in many different places. Uh to learn a proprietary language which you can only apply in one place might be a good investment uh during a certain point in time uh for a specific

uh reason project or what you ever are doing at this point. But I guess in general it should be the other way around. Uh we have uh a couple of examples in this regard. For example, if you think of this big uh front arena um um trading front office uh system. They use Python for example, which is a general purpose language, to describe payoffs on their platform. They say, Well, we have a proprietary platform but we use something as common as Python. W w many people

you can speak of even millions, can already program to do something which is specific to our platform. And I think this is the right way. Uh why would you invent kind of a a a Another spoken written language to write a poem or whatsoever.

uh I mean English is there or maybe you have another uh mother tongue or whatsoever, but there's typically no need. But I guess in the end it's it's all about specialization and payoff and investment. If you say, well What we are looking for and what we need is available out there and there is such a huge business case and this is uh uh sustainable for quite a while.

uh let's do th something on our own, um this might be the way to go. So for example, uh Goldman Sachs has his own language with slang intern internally and so forth. So Th there might be good reasons for bigger institutions for certain environments, but in general I would say if if there is something around which might be

at least as good as what I have in mind to solve the problem, why not using it? I mean, just to have a proprietary it's it's what I usually also use as an argument against proprietary

uh data science platforms and so forth. So you know we have Jupyter notebook, this is used by millions of people. There are I don't know how many tens of millions of Jupyter notebooks are available. But there are still people these days they reinvent the notebook and come up with a new notebook with a with another format used by five hundred people

um and with uh thousand five hundred notebooks available and which will never ever get the traction of what is already there. So if there are good reasons to do that, because your company has specific requirements, you wanna use it internally, fine. But as a startup I would doubt this model actually.

Okay. Yeah that's a really good point. So what are some of the some of the best resources that you'd recommend uh for someone who's keen to to start learning, you know, the language of Python? Uh, you know, someone who's just maybe an independent trader, they might not be a PhD or have any programming experience. What are some good resources that they could um start out with?

I mean, uh in this regard I must come back to what we do, uh obviously when it comes uh to learning Python for finance. There's my book the the trainings we give. The if you go to my private website, hilpish dot com, you find I don't know, fifty plus talks. about different topics, uh in notebook style so you can easily copy that and use it on your own.

Um there are others who provide trainings or like Code Academy and so forth that There are many, many places where you can start out, but so so in the end it I'm it's it's not only for me, I I guess for everybody, it's impossible to name all the the good resources that are out there.

I really think that there are many, many good resources out there. What my recommendation typically is uh to people asking me this question because I'm getting asked this uh quite often, is um learn the basics by reading a nice book or an online resource like the uh

scientific uh lecture notes, uh which are out there on the on GitHub uh for free, I don't know, two hundred pages of well written uh lecture notes about Python uh for scientific applications. So it's completely for free and you can study that. Uh do something like that as a starting point and then pick a project and learn and pick up along the way what you need.

going to Stack Overflow, Google your stuff, uh uh looking left and right, uh so for example, uh I mean I knew about Python for example, but It might be a nice project and I really like uh using the Raspberry Pi for example where you can easily learn Python. You can implement many, many a nice project based on a Raspberry Pi, which is also fun if you like that stuff, uh to learn Python. But not only Python for the sake of Python, but

to implement nice things that you can do with Python based on Raspberry Pi, even including other hardware elements and so forth. Um so this is usually my recommendation. Pick a project that you're interested in. and then try to build this project and pick up along the way all the ingredients, Python ingredients, other technologies.

uh that you need in order to accomplish the goal of your project. I guess from my point of view this is the best learning experience. This is the the very reason why I like to write books, why I like to implement bigger things on my own, uh because this is the best way to learn. I'm uh get getting a producer and and not that much a consumer is the best way to learn. Just produce whatever it is. Really good. So

I mean just uh just another question to bounce off that one. Realistically, what sort of commitment is gonna be required from a non programmer to become efficient with this language? That's also it's it's uh it's a tricky question in a sense. I said the Python fits in your in your brain actually, it's nothing usually the that blows up your your brain. But in the end

I would say if you're really interested in learning it, uh you will learn it quite fast. Just think of more like a child who's interested in playing a certain game or uh when I remember the days when I got my first computer, time didn't play a role because it was so much fun to do it.

um I was like, you know, h no matter how long it takes, I will master it in the end. This was my thing and back then we did assembler and stuff like that. You know, this was much more involved than the stuff we see here.

So it took uh weeks or even months until you got the first results. Here with Python you start out and and you see the first p graphics that you generated on your own like five minutes later. You know, this is a uh a real short feedback cycle. Um but I guess uh picking learning the basics, uh picking up a project, implementing it on in a realistic fashion and doing it part time or whatsoever, maybe half a year and I guess you never

uh have reached kind of a decent level uh of Python. Okay, sure. And just to pick up on what you mentioned um a couple of answers ago.

The course on Code Academy is awesome. I've done that one. Um that was that was really helpful. I think that was actually the first one I first one I did. So that was my introduction to Python. And um yeah, I learned a lot from that. So, all right Evil, let's wrap this up. Um, you know, we've been talking for almost sixty minutes now. Where can listeners go to find out more about you? So maybe your best website and you know, I know you're fairly active on Twitter. What's your Twitter handle?

It's uh D Y J H. So Okay. And best website where listeners can find out more about you? Um this is our company website, which is uh tpq.io, like the Python quans, tpq.io. Uh from there you get kind of directed to the different areas to the book stuff and so forth. And my private website is uh Hillpish.

dot com. There you find all my talks and and stuff like I mean I have even there kind of links to my Raspberry Pi project or what I mentioned before. It's kind of resources which are not kind of directly um associated with uh our company work but um yeah stuff which is obviously to ninety percent about Python and quant finance. Yeah, that's great. Now I'll be sure to include all those links at chatwithraders.com so listeners can find them all there in one place.

Um and just to recap, I know you've written three books. Can you just give us the title of those books in case listeners uh want to check those out? Yeah, the first one, uh which is actually out since uh almost one and a half years right now, is the Python for Finance one, uh which uh has been published by O'Reilly. This was December 2014, then in July 2015, last year my derivatives analytics Python book came out. This is

A little bit more involved. This is really about the finite side and it's an outgrowth of my university lecture. Um so this is more again about quant finance uh and how to do it in Python. The Python for Finance book teaches how to do it.

the other one doesn't teach how to do Python. So but the two together are kind of good companions if you are interested in derivatives and analytics. And the third one is even a little bit more specific in the sense that I focus there on a certain product category which is listed volatility and variance derivatives. uh uh quite popular instruments these days with uh increasing volumes over the years.

um and explaining everything based on uh Python code and more or less on interactive Python sessions. So very interactive, very easy to follow and with all the books come all the resources. Rwy'r rwy'r rwy'r rwy'r rwy'r rwy'r rwy'r rwy'r rwy'r rwy'r rwy'r rwy'r rwy'r rwy'r rwy'r rwy'r rwy'r rwy'r rwy'r rwy'r is reproducible with uh with the code that I provide. So reproducible research and publishing is uh what I'm looking for here.

Good stuff. All right Eve, well let's uh let's call this a wrap. Thank you very much for coming on. I I really appreciate you taking the time to do this and and speak with me uh this morning. So enjoy the rest of your day and let's stay in touch. Thank you for having me, I also really enjoyed it.

Hey, I hope you found this episode interesting. And like I said at the beginning, if there was anything which you didn't quite understand or you'd like more context around Just leave a comment at chatwithraders dot com forward slash eighty four and I'll do my best to point you in the right direction or I might even ask Eve if he can help out as well. Now the second thing I'd like to repeat, which I also mentioned right at the start of this episode, is

I'm running a survey for Chat With Traders, okay? This is a very short survey. I think it's about fifteen questions, all multiple choice answers. Take you about two minutes to complete, Matt. The reason I'm running the survey is because, as you know, the podcast is free and it's financially supported by sponsors. Now I want to make sure those sponsors are a good fit for you listening to this podcast and the sponsors want to know that they're connecting with a relevant audience as well. So

If you can help support this podcast, please go to chatwithraders.com forward slash survey to complete a very short survey. And because I appreciate you doing this, I'm going to give you a free gift. I'm not going to mention what that free gift is right now, but if you go to chatwithstraders.com forward slash survey S-U-R-V-E-Y, it's all there. So please show your support and if we don't speak in the meantime, I'll catch you next week.

You've reached the end of this episode of Chat with Traders, but rest assured there are more episodes.