Making a Career in Algorithmic Trading

By Milind Paradkar

The advent of algorithmic trading in the late last century caused a massive tectonic shift in the way trading took place in exchanges worldwide. Be it trading in stocks, derivatives, Forex or commodities, trading firms worldwide adopted algorithmic trading in a big way.

The last couple of decades have seen an exponential growth in the algorithmic trading market and it continues to grow at a significant pace. According to the “Global Algorithmic Trading Market 2016-2020” report published by Research and Markets last year, the global algorithmic trading market is expected to grow at a CAGR of 10.3% during the period 2016-2020.

In order to remain competitive and earn big profits year after year, big banks, hedge funds, and other trading firms have been hiring top talent from various universities and colleges worldwide. This is turn has led to a surge in algorithmic trading/HFT jobs. Scores of students, engineering graduates, and developers want to explore and build a promising career in algorithmic trading today. This said, many of the wannabe quants & developers are unaware of the nature of the work in algorithmic trading firms and the skill sets needed to make a foray into this coveted algorithmic trading world.

QuantInsti recently conducted a webinar on “Career development – Jobs in algorithmic/HFT trading”. The webinar offered a unique chance for attendees to interact with a team of quants & HFT developers on a one-to-one level and ask career-related queries to the esteemed panelists. The panel of speakers included Mr. Sunith Reddy (Head of Technology, iRageCapital Broking), Mr. Puneet Singhania (Director, Master Trust) and Mr. Gopinath Ramkumar (Market Risk Quant)

In this post, we highlight some of the most important takeaways from the webinar for job seekers in high-frequency trading jobs, quant jobs, and algorithmic trading jobs.

Key Takeaways:

 

Automated trading is not free from human intervention

A very important point that wannabe quant/developers ought to know is that automated trading does not mean it is free from human intervention. Automated trading has caused the focus of human intervention to shift from the process of trading to a more behind-the-scenes role, which involves devising newer alpha-seeking strategies on a regular basis.

 

Top skill sets for algorithmic trading

Analytical skills – Having an analytical bent of mind is a very important quality for any quant trader/developer, and is given a high weightage in an interview. For example, an interviewing candidate may be given a huge data set and asked to find patterns from the data. Candidates get evaluated on how they approach any given problem and their ability to justify their solutions objectively.

Math and programming skills – As the core of algorithmic trading revolves around algorithms, data, and programming, having reasonable programming skills and a basic understanding of statistics and calculus is important for any job seeker in algo/HFT trading. For example, if a candidate is applying in a firm that deploys low latency strategies, then an expert level of programming would be expected from such a candidate.

To know more on skills sets required, check out this infographic – Top skills for nailing Quant or Trader interview

Understand the strategy development process

While devising any strategy, it is important to understand the risk and rewards associated with that strategy in order to determine whether it has an edge in the markets. This is done during the backtesting of a strategy. The frequency of trading, instruments traded, leverage also needs to be taken into consideration before going live with the strategy live in the markets. Quants should also remember that no single strategy can guarantee profits year-after-year. Quants are required to formulate new strategies on a regular basis using advanced mathematical models & statistics and overhaul the old strategies to remain profitable in the markets. To learn about various algorithmic trading strategies, you can check out this blog post – Algorithmic Trading Strategies, Paradigms and Modelling Ideas

Techies need to understand financial markets

Quantitative trading involves dealing with large datasets, trading in different instruments like stocks, derivatives, Forex etc. Hence, even if you are coming from a non-finance technology background, as a developer in a quant firm, you need to have a fair understanding of the financial markets. Trading firms usually make their new recruits spend time on different desks (e.g. quant desk, trading, risk management desk) to gain an understanding of the markets.

Should I learn C++ or Python?

Python is good for conceptualizing and backtesting of strategies. Python offers many libraries for validation and visualization of results. It can also be used by firms for strategies that are not dependent on low latency. See here to know what makes Python the most preferred language for Algorithmic Trading. On the other hand, C++ is usually used by firms that trade very low latency strategies. Thus, if the objective of an aspiring developer is to get into an HFT firm, then irrespective of the language that he starts with, he will have to finally end up learning C++.

 

Points that recruiters look in a resume

Recruiters appreciate if candidates are honest on their resume, candidates should be able to back any points that they have mentioned. Candidates are expected to demonstrate a strong understanding in the core areas that are highlighted in their resume. Recruiters also tend to give positive weightage if the candidate has undertaken a project work or published any research papers in his/her areas of interests.

Salaries for traders and programmers

It is a known fact that salaries & bonuses are lucrative in algorithmic trading firms. However, there is no common compensation policy followed across algorithmic trading firms. For example, salaries paid to tech guys in similar roles can vary from one firm to another. In some firms, bonuses get equally split between traders and programmers based on the profitability of a strategy. Compensation can also vary depending also on the type of the trading firms (e.g. Family office/bank/HFT firm etc.) and the strategies (low-frequency/high-frequency) that are deployed by the firms.

Conclusion:

These were some of the important points that aspiring quants/developers should keep in mind as they prepare themselves for a successful career in algorithmic trading. You can click here to catch the entire webinar – “Career development – Jobs in algorithmic/HFT trading”. In case you need any further guidance, feel free to contact us.

Next Step:

If you want to learn various aspects of Algorithmic trading then check out the Executive Programme in Algorithmic Trading (EPAT™). The course covers training modules like Statistics & Econometrics, Financial Computing & Technology, and Algorithmic & Quantitative Trading. EPAT™ equips you with the required skill sets to be a successful trader. Enroll now!

 

 

Read more

Candid Conversations with an Algo Trader (Part 2)

Candid Conversation with Algorithmic Trader part 2

If you don’t know who you are, the stock market is an expensive place to find out – George Goodman 

In the previous post, I had a conversation with a few experts in the field of Algorithmic Trading to gain some insights into this seemingly “black-box”. That conversation not only helped me dispel some of my doubts regarding Algo Trading, but it also strengthened my desire to jump headfirst into a world dominated by complex mathematical models, managing positions in stock market, and coding woes.

So, here I stand, ready to take the plunge. I am told that I don’t need a university degree in finance or computer science to become an Automated Trader. Of course, to succeed in the field, I do understand that a formalized education will be highly beneficial, but right now, I am just scratching the surface here. Although, my engineering background does give me certain leeway into the understanding of financial mathematical models and programming.

Yet, the road seems bumpy. I still have a truck load of questions going through my head. How much knowledge of the markets is good enough to start trading using algos? Where can I verify my trading strategies? How can I make my strategies profitable? What are the “Must Dos” and “Rule-of-the-thumbs” of the field that I need to follow?

I guess I need to sit down with an expert all over again, over a cup of steaming hot latte, to clear the air.

So, I have read some literature on Algorithmic Trading and eventually want to set up my own trading desk. What path do you recommend?

If your idea is eventually to have your own setup, you would need to build expertise in quant and/or programming. To start with, I would recommend you to master at least one of the programming language commonly used by Quants and Algo traders (Matlab/Python/R), because that is an essential ingredient. Of course, the skills that you will develop in any one of the languages will be useful for others as well. The best way is to get your hands dirty as soon as possible. Start learning about the markets. Build your knowledge on charting. Read about different kind of strategies used by traders in the market. Perhaps, the important thing is to understand the basics of market inefficiency and to capitalise on it by building a strategy around it. Once the strategy is in place, backtest it on historical data to remove the inaccurate assumptions you might have had. If all goes well, you should be able to trade with this strategy in live markets.

 What do you mean by Market inefficiency?

 A situation where live prices of stocks or securities do not accurately reflect their true value. In an inefficient market, some securities are overpriced while the others underpriced. This gives a certain risk or in some cases, opportunity to make money for the traders.  Ever heard of the term “Statistical Arbitrage”?

Yes. Isn’t it a fancy word for pair trading? I would like to know more about it.

That is only partially true. Pair trading is one of the many kinds of statistical arbitrage. It is also a type of market inefficiency. Being quite popular amongst Algo and High-Frequency Traders, It is actually an evolved version of pair-trading strategies, in which stocks are put into pairs by fundamental or market-based similarities. When one stock in a pair outperforms the other, the poorer performing stock is bought long with the expectation that it climbs its outperforming partner, and the other is sold short. This hedges risk from whole-market movements.

That does seem logical. So what do you opine as the most important things to keep in mind while designing a trading strategy?

There is no direct answer to this question, as it’s a mix and match of several things. You need to understand how the market operates and determine the entry and exit points accordingly. As far as my experience goes, recognising when to exit the market and identifying stop loss is more difficult than the entry position. Every strategy is unique, and therefore in need of its own customised positions.

Once your strategy is ready, it is good idea to backtest it against historical data. However, there is an element of caution to be exercised when backtesting. E.g. a trader wants to test a strategy based on the notion that Internet IPOs outperform the overall market. If you were to test this strategy during the dotcom boom years in the late 90s, the strategy would outperform the market significantly. However, trying the same strategy after the bubble burst would result in dismal returns.

Therefore it is important to prepare a contingency plan, as no strategy will work in all prevailing market conditions. Drawdown is unavoidable, and it should be incorporated in the trading strategy. Remember, a good strategy is not a slave of technology. You can hire a programmer to implement your trading strategy, but not vice versa.

Quantopian, NinjaTrader are free to use platforms in case you want to put your algorithms to test. There are a lot of other free backtesting tools that can be looked into, if you want more information.  

 I like your point about the exits, as I understand that any trading strategy is not devoid of risk. How are the risks handled in an Algo Trading system?

You should understand that risk management and its reduction is one of the most critical areas of Algorithmic Trading. The market is full of oddities and once you realise that you are on the verge of losing money, your algorithm should check out of the deal. That is the reason that exit points of an Algorithm are more important than the entry points.

Algorithmic risk can be categorised into several things: Consistency, Quality, Scalability being a few of them. Consistency means that the data which is being fed to the system is not old or outdated.  Market data packets have time-stamp embedded in them. The advanced exchanges in the world are adopting concepts like time-syncs to atomic clocks. Your algorithmic trading system should be able to track these time-stamps to ensure the data you are getting is indeed fresh.

For any normal trading activity, you have to take care of market risk, financial risk, regulatory risk, liquidity risk and credit/counter-party risk. In addition, algorithmic traders also need to prepare fo the statutory risks. There are two places in an Algorithmic Trading system where the risk has to be handled: within the application and before generating the order in the Order Management System. India, having one of the most stringent regulatory environment, a lot of these algorithmic trading related risks have to be mandatorily checked in the system before an order flows out.

Wow! That’s a whole lot of risk managements that we are talking about here. But I would want to take a step back and ask you about the Order Management System

Order Management System or OMS as it is called, is the pathway through which the orders are routed to their correct destinations. Orders need to contain the security identifier, order size, price limits, order type and conditions, type of algorithms used, to name a few. This information is usually entered via a winform by the end user but for fully automated systems no winform is required, however in both cases it is recommended to have each order object stored in a relational database for record keeping.

Once the order is captured by the system, it needs to be routed to the required destination. As most systems receiving the order have their own proprietary protocol, order routing requires each order to be encoded in the correct format.

There are several checks in place to make sure that order information is correctly sent and received. The venue will run various checksums and order length so that the FIX engine can confirm that the order received matches the expected order transmitted. The order manager also performs risk checks before sending out the order.

Well, well. I guess I will stop my queries for now. I have got some really great insights from your talks. Probably I should just start with some programming tutorials to get me started.

That’s a good idea. Here are a few resources where you can look for. Aggregated Reading list for Algorithmic Trading

Step Ahead

If you had any such conversations with an Algorithmic Trader or Market Expert, do share with us in the comments section below.

If you are a coder or a tech professional looking to start your own automated trading desk. Learn algorithmic trading from live Interactive lectures by daily-practitioners. Our Algo trading course, Executive Programme in Algorithmic Trading covers training modules like Statistics & Econometrics, Financial Computing & Technology, and Algorithmic & Quantitative Trading.

Read more

A Candid Discussion with an Algorithmic Trader

Candid Conversation with Algorithmic Trader

The role of Algorithm in a person’s life is too substantial to be ignored. From a simple coffee-making machine to the music system in his car, from elevators to search engine like Google, all are governed by a set of logical instructions – Algorithms or Algos, which enable them to respond to a person’s specific requirement.

With the advent of the internet, Algorithmic potential has been unleashed in its truest form. Defining trends, identifying preferences through social media and targeting relevant groups with tailor-made services have been possible through sophisticated Algorithms put in place.

Of course, with all the technological advancement, stock markets have been on the forefront of adapting to the riveting world of Algorithms. Algorithmic Trading has gradually become the most preferred way to trade on the bourses, with an estimated 80% of the total volumes of the Wall Street.  Institutional investors, hedge funds and large financial brokerage houses have switched to Algo-trading to stay competitive, cost-effective and cater to their clientele.

So, what exactly is Algorithmic Trading or “Black Box Trading”? Does being a successful Algo Trader carries prerequisites like expert programming skills? What is the investment required for Algo trading desk to be set up? These were some of the questions that I pondered upon while deciding a career switch in favour of Algorithmic trading.

This post tries to answer the questions of a complete novice to Algorithmic Trading.

How to setup of an Algo desk?

Algorithmic Trading is a process of using a set of instructions to place an order of buying or selling script with volumes and speed impossible for a human being. The set of instructions is based on various market metrics like price, time, volume and any other user preference. The good part about Algo trading is that it eliminates human intervention thereby making trading sans emotions and intuition.

A typical Algorithmic system’s architecture entails three primary components

  1. Market Data Handler
  2. Strategy Module
  3. Order Router

The market data handler, as the name suggests, receives the market data and stores it. Strategies for trading, in a mathematical model, are fed to the strategy module. It also serves as an interface between the market and a trader. Order Router or manager sends the order back to the exchange for buy/sell. To set up an Algo desk, as a broker you need to identify your co-location to place the servers in close proximity of the exchange, feed in your strategies to your system after having it backtested and authenticated, have a good internet connection, and I think you are good to go!

What are the steps would you suggest to venture into the field?

Well, first and the most important step is to build a solid base. Learn some programming skills and get a grip on the markets. Being good with numbers always helps. Begin by exploring core subjects like statistics and econometrics. Some books like Quantitative Trading by Ernest Chan or Trading & Exchanges by Larry Harris can elaborate on setting up a “proper” Algorithmic trading system.  When you are through with the above-mentioned steps, get your hands dirty in strategy building, modelling techniques, and statistical tools. Get a hang of various trading strategy paradigms, like statistical arbitrages, execution strategies, bid-ask spread. There are a few free courses available online on Udemy and Udacity which are quite good to test waters. There are other paid and advanced courses available for serious learners.

Which are the commonly used programming languages employed by the traders?

C++ remains the most preferred language as far as High-Frequency Trading (HFT) goes. Reason being that memory leaks and related errors are far less in C++ as compared to other languages. Python has come up in a major way for coding strategies as well as backtesting, because it is easier to master, and is backed by good scientific libraries like Numpy. A number of forums today discuss investing and trading strategies coded on Python programmes.

How is the retail participation in Algorithmic trading?

The setting up cost is definitely on the higher side, from a traditional trading terminal. Obtaining a co-location can be an expensive affair. According to recent statistics, nearly 70-80% trade on Wall Street is done using Algos, chiefly by large institutional investors and hedge funds. However, the scene for retail participants is evolving with the offering of web-based platforms. For someone who is not too concerned regarding latency, its works like a charm. Other than that, firms like Interactive Brokers provide retail clients with API and packages so that traders can code their strategies and trade. Once you get a hang of it, it is like a simple Gmail account. You log in to your account, test your strategy, perform back testing, and after optimisation, trade in live markets. Paper trading or testing on a simulator is also highly recommended.

How is the ecosystem for Algorithmic Trading in India? Are companies readily opting for Algorithmic trading given its niche category and involvement of highly skilled practitioners?

Algorithmic Trading was allowed in India by SEBI in 2008. In a span of 8 years, nearly 50% or of the trade by volume or maybe even more than that is Algo-based. That speaks for its popularity. Indian bourses have adapted to the change very well, with a steady increase in the active participants. Both FIIs and domestic funds use the Algo route to place orders.

How does the global future look like?

Very promising actually. It is clear that automation is the future that is driving the world. Be it in any field, automation is making a tectonic shift from a traditional path, and the same applies to stock markets. In US markets, 70-80% of the exchange volumes is happening through automated systems. Emerging markets like India are witnessing exponential growth in the domain. Of course, the markets are maturing every day, so trading costs would decrease after a certain point. Case in point: Automobile Industry, where after the introduction of robots, it was initially thought that the industry would be unsustainable due to high capital cost.

What are few things that a beginner should keep in mind when venturing into this field?

The most important one is that it is just not enough to have a good trading strategy, but also to have a competitive edge. It can range from having innovative ideas which disrupt to having a low brokerage or the kind of markets you have access to, but you have to have a killer proposition if you plan to be successful. Treat this as any regular business, where you have to develop a strategy to outwit the competition. It is important for someone starting new to have the nuances of the trade figured out.

What should be my next step if I want to understand more about this field?

The best way is to find experts and domain authorities to talk and discuss your doubts. Try out freely available tools and resources on the internet. Be prepared to embrace new knowledge and develop new skill sets!

Step Ahead

If you had any such conversations with an Algorithmic Trader or Market Expert, do share with us in the comments section below.

If you are a coder or a tech professional looking to start your own automated trading desk. Learn algorithmic trading from live Interactive lectures by daily-practitioners. Our Algo trading course, Executive Programme in Algorithmic Trading covers training modules like Statistics & Econometrics, Financial Computing & Technology, and Algorithmic & Quantitative Trading.

Read more

Free Resources to Learn Machine Learning for Trading

Free Resources to Learn Machine Learning for Trading

by Anupriya Gupta

While being a vibrant subfield of computer science, machine learning is used for drawing models and methods from statistics, algorithms, computational complexity, control theory and artificial intelligence. It focuses on efficient algorithms for inferring good predictive models from large data sets and is natural candidate for problems arising in HFT – both trade execution & alpha generation. (more…)

Read more

What They Don’t Teach You at MBA?

Business world is an intense and competitive field. An MBA degree can help develop skills that can be used in a number of situations. In a dramatic shift versus a decade ago, technology jobs are just sought as roles in finance. MBA’s are proving that they can make a difference as leaders in many different industries.

Woo Hoo MBA Done

In the below list we try to help these MBA’s with courses and programmes to do after getting their MBA degree. Any list does involve a certain bit of subjectivity and thus it will never be perfect. Despite all this one must attempt to create a list. After all, a list helps us in the simplest way to know why we decide to take a certain action. (more…)

Read more

What do Hedge Funds Look for When Employing?

What do Hedge Funds Look for When Employing?

By: Brandon Msimanga

One might consider Hedge Funds and Hedge Fund management to be risky business which begs to differ why anyone would find a job at a Hedge Fund to be an appealing prospect. However when it comes to working at Hedge Funds the high risk does come with an unusually high compensation in the form of a very attractive monthly pay cheque. (more…)

Read more

Algorithmic Trading for Technocrats and Engineers

Algorithmic Trading for Technocrats and Engineers

by Anupriya Gupta

Being a pioneer in the field of algorithmic trading education, we often get queries like why should I do this certification? How will it benefit me? Is trading right for me? Many who come up with such queries are students from various fields of study and particularly from engineering and mathematical backgrounds. We will try to address some of these frequently asked questions. (more…)

Read more

How Do You Become A High Frequency Trader?

While the broad contours remain the same, this post is written from Indian market perspective.

HFT is an extremely technical discipline and it attracts the very best candidates from varied areas of science and engineering – mathematics, physics, computer science and electronic engineering. While in the US, they usually look at Ph.D. in CS or physics/maths or an MFE degree, it’s not so much in India. In India, an engineering degree in CS/Maths or MBA in finance from a reputed college along with your zeal for problem-solving and coding can give you a fighting chance to land up quant analyst or a quant developer job in an HFT firm. While the degree makes the resume presentable, it’s not the barrier. If you have done lots of work and have something to show for in your resume, the industry recognizes it. But be aware that getting a quant analyst or a quant developer job will take a significant investment in terms of study and effort.

High Frequency Trader

by Anupriya Gupta

(more…)

Read more

Financial Engineering Courses Evolution in India

Evolution of Financial Engineering Courses

Changing market structure – and the emergence of Financial Engineering

With the fast changing global scenario and increasing competition, large financial banks and brokerage houses are faced with the need to develop customized solutions and complex models to tackle niche and difficult client problems. These led to the creation of a new discipline – Financial Engineering; and thereby demand for a new breed of professionals. Being a nascent and highly paid field, more and more people are trying to acquire requisite knowledge and skills and move to this profession.

Growth of financial engineering courses across all segments

In the wake of the above, there has been a surge of financial engineering courses in India being offered by various colleges and institutions at different levels. These finance courses target students/professionals at different points: intermediate, graduation, and even after post graduation and provide degree/ certificate in financial engineering. In addition, some of them are also offered as skill development programs for working professionals. (more…)

Read more

QuantInsti Invited to Conduct a Session at IIT KGP

IIT Kharagpur

Mr. Gaurav Raizada spoke on Algorithmic trading on 28th January, 2013 at IIT KGP.

The session started with a brief introduction to Algorithmic Trading including an overview of its exponential growth in India in the last 3-4 years. The speaker also discussed a few basic trading strategies leading the discussion on development of a simple algorithmic trading strategy by step-wise development of complex order types.

The talk focused on Asynchronous nature of Market Data Events and Concept of order book in trading. Overall, it provided an insight into the mathematical and programming aspects of Algorithmic Trading.

Read more