Moneycontrol Be a Pro
Get App
you are here: HomeNewsBusiness
Last Updated : Apr 13, 2019 04:11 PM IST | Source: Moneycontrol.com

Meet Devang Jhaveri: An IT head who became an algo trader

A CA who doesn’t like auditing, who gave up his family’s diamonds business and bore the brunt of the 1994 market crash. Meet Devang Jhaveri, a successful algo trader, who shares his recipe to investing in the stock market.

Shishir Asthana @moneycontrolcom

Few give up family’s diamonds business, fewer crack the CA exam under 21 and don’t like auditing, and far fewer dabble in stock markets after getting singed. And Devang Jhaveri exactly did that.


If that’s not enough, Jhaveri quit a cushy job as IT head to play the market using his unique algorithmic trading strategy. Not just that he is a core part of a three-day Algo Convention which spreads awareness about algo trading.


Just out of school in the early 90s, Jhaveri witnessed one of the biggest bull runs in Indian markets, thanks to his father. The 1994 market meltdown singed the euphoria of the bull wave, leaving a scar on Jhaveri’s young and impressionable mind. The fall was enough to convince him to never take a plunge into the market.

But as destiny would have it, he came back and this time putting in all his work experience to good use. The best lessons he learned were from his father, which now forms the core of his ‘Techno Funda’ algo trading strategy where he combines technical analysis with fundamentals to trade in the market.

Close

Lessons from the market meltdown aside, there’s another aspect of his life he implements in his algo-trading strategy – altruism. Every weekend he feeds hundreds of people and during winters he distributes blankets and in monsoon raincoats and umbrellas to the needy. He funds, through his NGO, education and medical expenses for those who cannot afford it.


It is his altruistic nature and the scar of the market meltdown that is one of the reasons behind him spearheading an awareness program on algo trading. Having seen the destructive power of emotion in a meltdown, Jhaveri is helping spread the algorithmic way of trading where emotions are taken out of the equation.

In an interview with Moneycontrol, Jhaveri speaks of his incredible journey and the role Algo Convention in spreading the word of algo trading.

Q: You were exposed to markets at an early age, can you take through your journey?

A: My first memory of the market is when I went in the ring after appearing for Class 10 exams. My father was a jobber and he used to go to the ring every day. Those were the times when the so-called Harshad Mehta rally was picking up. (A jobber is a person who works on a stock exchange.)

My first trade was also from those days – VB Desai Financial, a stock which my father’s friend advised me to buy. I bought it at Rs 100 and saw the stock moving 15-20 percent every day I got in. In nine months, the stock had touched Rs 900, the price around which I exited.


In those formative years, I thought money-making was easy but soon the crash came and for those of us with a ringside view, literally, it was devastating, and that is putting it mildly. In 1994, post the crash I thought I will never come to the stock market again. However, I did come back but only after traveling the world.

Post my market experience, I studied hard and cracked CA exam before the age of 21. I had to wait for the institute to give me my accreditation as they had a policy of minimum qualifying age of 21. I, however, did not like auditing and decided that I will not pursue it as a career option.

So, I pursued my grandfather’s business in diamonds for a while but that too did not suit my aptitude. I was, however, good at computers from an early age and tried ERP which just was a new field in those days. I liked what I was doing there and decided to pursue it.

Between 1998 and 2005, I worked with Oracle and later with GE Capital International Service, which is now called Genpact. My GE experience gave me the opportunity to work with various GE companies across the globe. I interacted with the CEOs of these companies, the finance head and took their requirement and worked with our technology team to code it. This gave me a good insight into various businesses as well as helped polish my technology skill.

While my family had given me a free hand in deciding my career path my father’s ailing health and the fact that I was the only son made me come back to India. In those days the only sector where high-end IT was used in India was banks.

I got a job with ABN AMRO, which was later taken over by RBS. I eventually was heading the IT team for Equity and Private client group.

We worked on various aspects of trading technology on the Fidessa software. Back then one of our traders designed a cash future arbitrage software which we gave to one of our vendors Omnisys to develop.

The period 2007-13 was the best time phase for IT in technology, and I was at the centre of it. My experience in ABN AMRO opened the door of my re-entry in the market.

One of the jobs that I was tasked to do as an IT head was to evaluate products that were to be used by our traders and analysts. I was exposed to fundamental software like Prowess and Capital Market. I played around with them and was soon running various fundamental queries and started investing in the output of these queries.

This period was 2009-10 when the market was just picking up. The stocks that I picked up then have increased by over 20 times.

Q: What were the queries that you were running those days?

A: I had designed seven rules for picking up stocks. They picked out companies that were completing their expansion plans, by tracking their capital work-in-progress (CWIP). I then looked for companies where sales were rising but the debt was coming down, which generally shows up in good companies in a few quarters.

The next step was to look for companies where the operating margin was increasing and the company was able to command a higher operating margin than its peers. This had to be true for 5-6 quarters with margins never falling below the level achieved in the first quarter.

Other filters were companies with Return on Equity (ROE) and Return of Capital Employed (ROCE) above a threshold limit and companies with low leverage, low pledged shares by promoters and improvement in interest coverage ratio.

I avoided commodity plays and banks. Though I worked in a bank for many years on the IT side and knew the mechanics behind the running of a bank I was not confident of investing in banks. It happens to many people that when you work in a particular sector and look at the insides you are not comfortable with the sector.

My filters were helping me select growth stocks and I was picking up companies which were growing at a steady pace but were commanding better margins as compared to their peers, which signals a pricing power and leadership.

On top of these fundamental filters came the entry price. Here is where my father’s teaching helped me. Among the many things I learned from him was to know when to go long and when not to go long, as well as when to go short and when not to go short.

In equity markets, you need to know the role of interest rates. With that knowledge at the back of your mind, you have to pick up companies where the least they can do is earn more than the interest rate in the market. But investing in such companies at any valuation will not serve the purpose. That’s where the concept of Growth at Reasonable price (GARP) comes in. One has to compare the opportunity cost with the interest rate before buying.

In the case of shorting a stock, one has to keep in mind inflation. But in a growing country like India keeping a short position for a long period of time is risky. I generally do not short, I am predominantly long in stocks.

Q: When did you exit from your investments?

A: That question is the reason I am in algo trading. I never used to sell my stocks. I have held on to stocks from the time I re-entered the market. There is a small story to how I learned to exit my position.

After my stint with RBS, I joined Edelweiss as the head of Algorithmic Risk. They created a position to prevent Algo trades from failing, thanks mainly to the episode with Knight Capital where it blew up $440 million in a matter of 20 minutes. The company had tested its trading code but while live trading they used some other code because of which they went in a self-destruction loop and within 20 minutes the game was over.

Edelweiss did not want to repeat such an episode so they wanted to protect their Algo trades. We created a 7 layer perimeter to prevent such ‘mishaps’. These included trading limits given to each terminal, there was also a throttle limit which did not allow throughput to increase in a terminal beyond prescribed limits. Then at an overall level, we had another perimeter set which stopped all trades if it breached a particular limit. We had other checks and balances to prevent a repeat of Knight Capital. For me,  it taught the importance of safeguarding your capital.

My two years in RBS and one year in Edelweiss taught me the importance of exiting. In our office, we used to play mock trading games. What came as a revelation was that my entries were good but I did not have an exit strategy at all. Because of which when markets were down I was cash strapped as my money was blocked in the stocks I had picked up in the earlier cycle and I could not benefit from the next rally.

By this time I was also introduced to technical analysis where it became clear to me where I was wrong.

I had invested in a company called Opto Circuits that was doing very well through 2009-10 and posting excellent results. But then without any reason, the stock started falling. It was only after it had fallen considerably that there were reports of misrepresentation of numbers.

This and the Satyam Computers incident a year back made me realized that the basis on which I was making my buying decision – fundamental numbers, may not be true.

There is something else that I needed to warn me of the impending danger. I was believing and dependent on numbers that the company was disclosing every quarter, but there was another data point that told a story of its own every day and that was price, which I had to learn to respect.

The other thing I learned was on position sizing. The value of stocks I picked up was random. In some cases, I bought Rs 1 lakh worth of stocks in others it was Rs 40,000 or even Rs 20,000. As was more often the case, the stock that moved the most was the one where I had the smallest position. I realized I need to have a rule-based, formula-based strategy if I want to do well in the market.

That’s when I met Vivek Gadodia who was in Algo Trading. A junior from my school Vivek had a well-defined rule-based system for his strategies. Towards the end of 2013, I gave him some money to manage.

From 2014 onwards I started incorporating technical analysis in my trading with clear rules of entry and exit. But exits were still not easy. It was against my nature to exit a position, so even when an exit signal came I used to exit between 30-40 percent of the position. I slowly corrected it and built my ‘Techno-Funda’ system of trading.

During this time I had joined Rabo Bank as the Head of IT and their Chief Information Security Officer since I also had cleared my Certified Information System Auditor certification. As luck would have it the job here was not interesting as everything was controlled by the Singapore office and there was little room to operate. In my case, the familiar saying that people leave managers and not companies were true.

Also in the IT field burnout rate and chances of obsolescence with age is high. You get younger people with the latest knowledge at a much lower cost. After a lot of thinking I decided to give myself a chance and take the plunge into full-time trading.

Q: What is the Algo trading strategy that you currently use?

A: In the Techno Funda strategy, we run a query on all the companies listed on the BSE. To test our hypothesis, we have taken historical fundamental data from Quandl for more than 10 years and ran our queries. This helped us filter out strong companies based on various filters like ROE, ROCE, and operating margin which is important to satisfy my basic premise that the company should earn more than the market interest rate. Higher operating margins gives companies the leg room to earn better returns.


Another filter we use is cash flow. Here if the cash flow of the company is not growing at the same rate as sales and profit growth then there is stress in the financials of the company. We will not touch such a company.

Also, we do not forecast forward earnings of companies and use historical price to earnings ratio pick up our stocks based on valuation parameters. We believe that future earnings are getting captured in the price.

These filters throw up around 300 companies which we then rank on the basis of market capitalization and then pick the top 30 based on percentile.

After the companies are shortlisted for trading, we apply the technical analysis filters to select which companies to trade. We use very basic and simple rules for trading.

A 50-day and a 100-day moving average cross over is the entry filter. Here the moving averages period selected is big but it is essential to my core philosophy of efficient use of funds. The long period moving averages ensure that I do not get in the stocks when the market is in the accumulation phase and money is not blocked.

On the position sizing front, we do not trade in more than 25 stocks at any point in time. Only after a stock is stopped out do we buy the next one where the buy signal has been triggered.

As for selling we use a smaller moving average crossover of 10-20 days. This way we are exiting closer to the top.


Our back-test of the strategy showed that had we run the strategy in 2008 we would have been 80 percent in cash in April. When we were actually trading the strategy in 2018 we were sitting on 90 percent cash in February 2018, just when the fall started.

The system generates a compounded average growth rate of 28 percent with a 22 percent drawdown. Out of every 10 trades, we have 4.6 trades that are winners but our winners are 2.5 times our losers. Since the filter throws up mid-cap stocks also we just need one or two good stocks that will take care of the overall return.


Another thing that the system tells us is of impending trouble, as it happened in January 2018. We were already fully invested in 25 trades but our program was showing many more stocks lining up to enter the market. This signaled an overheated market where many stocks are participating in the rally. What happened in February 2018 is well documented.


The idea of Techno Funda strategy is very clear. I do not want to trade any stock technically if it is not fundamentally sound. And at the same time, I do not want to keep on to holding on to fundamentally sound company even when it is evident in the market that something is wrong.


Q: You and your team have started the Algo Convention, can you tell us what it is all about?


A: Algo Convention is what we call a convention where all the stakeholders of algorithmic trading meet. When we started off with algo trading very few people had even heard of it. It was very hard to convince someone to invest their money through algo trading.


The idea came to me when I was watching TV and saw a company selling ayurvedic toothpaste. The focus of the advertisement was on the word ayurvedic rather than the brand of toothpaste. That’s when I thought that in order to sell algo trading we need to create awareness of the subject. It was like concept selling where we had to create a market place.


So we decided to hold the algo convention as a medium of spreading awareness and not making money out of it.


In the first year, we called 20 people in our office from the different profession – a doctor, a CA, a businessman and similar other people who at some point of time had participated in the market. The feedback that we got after the presentation was encouraging where the participants said they understood and appreciate algo trading. Since all these people had invested in the equity market they appreciated the role emotions play on market performance.


Encouraged by the response we decided to bring in a bigger forum of 40-50 people the next year where we introduced the subject of Artifical Intelligence. However, many confessed they did not understand much on the subject which led us to work more on teaching the concepts.


The next year we organized a bigger event in Mumbai with a capacity of 230 people, to our pleasant surprise we had to close booking after taking in 240 reservation.


We not only have some very good speakers in the event who are imparting knowledge on the subject, but also those who are offering platforms to run algo trading but also actual traders who provide practical knowledge on the subject.


We introduce new topics like last year we introduced Machine Learning. We also have international speakers participating in the event through video conferencing. Plus we have knowledge partners like Quantinsti, CMT among others.


This year it is a three-day event where the idea is to cover everything that a new person who wants to learn about algorithmic trading can find it in one place. Which is why we decided to hold a residential program this year that offers a chance for people to interact with each other, network and the experience of veterans is passed on.




The Great Diwali Discount!
Unlock 75% more savings this festive season. Get Moneycontrol Pro for a year for Rs 289 only.
Coupon code: DIWALI. Offer valid till 10th November, 2019 .
First Published on Apr 13, 2019 03:57 pm
Loading...
Sections
Follow us on
Available On
PCI DSS Compliant