How Quant Traders Consistently Generate High Returns in US$ Without Drawdowns from the Volatility on Cryptocurrency Market
Lately, bitcoin has been all the rage. It’s not uncommon to hear stories that someone made millions investing in certain cryptocurrencies. That means markets are highly volatile. If you invested at the wrong time or bet too much on the wrong direction you could’ve lost all your money. But is there a way to profit from market movements without making bets on certain directions? Apparently, such strategies are called market-neutral strategies; they take advantage of market anomalies, security mispricings and other shifts in statistical patterns. Cutting-edge quantitative models are usually required for banking on these opportunities. Also, since these anomalies exist only for short periods of time, advanced IT infrastructure and order execution mechanisms are required.
But let’s look at an extreme example in order to illustrate the concept. Recently, on the 15th of September Chinese regulators announced that they’d shut down all cryptocurrency exchanges in China. This news sent shock to other exchanges globally. Let’s have a look at how bitcoin price changed on 3 exchanges in different parts of the world (and neither of them was in China!):
While it was possible to lose a lot of money during such turbulence, it was also possible to make almost risk-free profit by buying bitcoins at one exchange and selling the same amount simultaneously at another exchange, effectively hedging all the market risks by having total market position equal to 0.Let’s have a look at how spread developed throughout the day:
Should you decide that 2% spread is enough for you to make a profit, you could have entered the market, effectively locking in your capital. In which case you’d have missed much bigger opportunities. But on the other hand, if you stayed away, you could’ve missed such an opportunity if spread closed. So not only do you need to react in a timely manner, but you also need a high-quality predictive model that will allow you to maximize your profit.
Of course, this was an extreme example and spreads as high as 18% between two exchanges are very uncommon even on crypto markets. But smaller price discrepancies between numerous assets across many exchanges, occur every day. Again, not only do you need to monitor all opportunities, but you’d also need to distribute your capital across multiple exchanges optimally if you goal is to maximize your profit.
Now, let’s look at another popular trading strategy that is widely used in HFT (high-frequency trading). It’s called “market making”: when you provide liquidity to the market (and some exchanges even pay you rebates for that!) by quoting bid and ask prices simultaneously (i.e. buying and selling the same asset), collecting the price difference (called “bid-ask spread”). Let’s have a look at daily prices of Bitcoin in Japanese yen, and the bid-ask spread:
(the X-axis is in ‘ticks’(price updates) – there are more than a million in a given day! Hence, the term, “high frequency”)
The trick here is to avoid holding your position for too long, because the price may change significantly (most HFT algorithms hold position for only fractions of a second!) and to control the size of your position, so, for example, if the price is going down, you’re not buying more and more. These strategies are usually very complex and require some advanced math, if you’d like to know more details please subscribe to our newsletter below. For now, let’s see how a typical high-frequency market making strategy works:
It makes a very small amount of money per trade, but it makes thousands of trades per day and shows consistent profit that is not affected by market direction.
Who we are and what we do
We are a group of “quants” with an academic background in Numerical Methods, Computational Mathematics, Game Theory and hands-on experience in High-Frequency Trading and Machine Learning. Our interest was in exploring opportunities in cryptocurrency markets, with the goal of exploiting various market inefficiencies to generate steady absolute returns (not correlated with market movements) with low volatility, or simply put, steady profit without major drawdowns. We have already developed and tested a set of automated trading algorithms and have run them on the real exchanges. We have a fast and reliable infrastructure to execute our algorithms. Our current returns are over 25% per month.
• Combination of statistical arbitrage strategies and market-making strategies with extremely low holding periods that deliver absolute returns that are unaffected by market direction
• Our cutting-edge geographically distributed IT infrastructure monitors market anomalies and shifts in statistical patterns with microsecond delays
• Lightning-fast order execution
• Proprietary risk management models and optimal capital allocation are the core of our trading methodology
• Quantitative alpha models that generate trading signals based on statistical machine-learning and numerical optimization methods
• Order execution models based on optimal stochastic control and reinforcement learning
Once our ideas proved to be working and scalable, we’ve decided to raise additional capital and spread our trading business on crypto exchanges worldwide.
Our Initial Token Offering will be in 2 stages:
-Oct 2nd - Oct 16th, 2017 – presale token offering, target - $0.5 mln. $400k will be used as a trading capital to provide income for token holders (we take effectively 0% management fee+20% performance fee), $100k will be used to continue developing infrastructure and setting up a legal structure to comply with regulators in regions where we currently can’t operate. On that sum, token holders will earn a total of $150k coupon payments, repaid from % of our profit. Payments will be distributed each month.
-starting in Q1 2018 – initial token offering, target - $20mln, 100% of that sum will be used as trading capital, we’ll take effectively 2% management fee+30% performance fee.
Georgy Bulygin,MSc in Applied Mathematics and Physics. Specializing in Numerical Methods and Computational Mathematics, he has been working for 6 years at the Moscow Institute of Physics and Technology. He then became fascinated with the world of HFT (High-Frequency Trading) and joined a boutique firm as a quant developer. As a founder of TensorBox, he’s responsible for the core trading strategies.
Vadim Kuchinskiy, MSc in Applied Mathematics and Physics at MIPT, M.A. in Economics at NES. Investor in FinTech, ex-director & FX-trader at ING and RBS London
Dmitry Shalagin, MSc in Applied Mathematics and Physics at MIPT, M.Sc. in Economics.
Specializing in investments and management of high-technology startups.
He has been working in management consulting for 4 years. Then he was responsible for due diligence at “Skolkovo” Foundation and closed 40+ deals with overall financing of $80 million. Later, he became a partner at “Phystech Ventures” an early-stage venture capital firm. During the last 3 years, he has been carrying out arbitrage deals with fiat-currencies, ROI - 70%+ per year. Nowadays, Dmitry is co-founder & investor in e-commerce, insurance and fintech industries. As a co-founder of TensorBox, he’s involved in business development and fundraising processes.
Rinat Yaminov, PhD in Applied Mathematics and Physics. Specializing in Game Theory and Computational Mathematics. More than 10 years of experience at Dorodnicyn Computing Centre of the Russian Academy of Science. He developed an automated trading system for a boutique fund in the USA. He worked in an investment department, was responsible for the risk and strategic analysis of projects, and managed a total portfolio of more than $3 billion. As a co-founder of TensorBox, he’s responsible for the risk and portfolio management part of the strategy.