Automated Cryptocurrency Trading Engine in dotnet - 08 Aug 2022

White Paper on Automated Cryptocurrency Trading Software Development using Microsoft dotnet

Disclaimer

This is not advice on how to trade, an incentive to trade, a recommendation as cryptocurrency of a way to make money. Please do your own financial research.

Background of Info Rhino's interest in cryptocurrency

Our CEO has worked in financial technology for many years prior to and during consulting though Info Rhino. With a keen interest in cryptocurrency as a new innovation to the solving challenges in the traditional financial system, and as a mechanism to simplify many commercial agreements - cryptocurrency opens up finance to the many.

Having said this, we are still in the speculative phase of cryptocurrency. We liken it to the gold rush, there are many areas where financial growth spots will emerge and others will die. It will be a balance of working with the most durable cryptocurrency technology, and knowing how to move between different blockchains and cryptocurrency services. Many liken cryptocurrency to the Wild West, but we have to be mindful that money can get lost inside our existing financial system - legitimately.

We have purchased a domain called www.cyptostatto.com and we are planning to release a new platform publishing regular market insights to it.

If you wish to learn more about our work, either fill out our contact form www.inforhino.co.uk/contact or email us here solutions@inforhino.co.uk

Can we build a money printer?

Cryptocurrency seems to offer the ideal ecosystem for making money;

• Asset fundamentals
• Price volatility
• Bear Markets
• Bull Markets
• Secular Bull Markets
• Secular Bear Markets
• Price feeds
• Online tools for trend analysis
• APIs for interacting with trading exchanges and other asset exchange platforms
• Advanced trade types, limit orders, stop loss, take profit, leverage/margin, short trades

It looks as if the cryptocurrency market offers trading capabilities normally open only to hedge funds and investment banks for stocks and other assets.

Evidence of trading potential

If you have a trading strategy, and it returns profit, we have the potential candidate for automated trading. With cryptocurrency, we aren't always expecting to get fiat profit. Sometimes, we may try to grow our "stack" to then capitalise as their fiat value rises.

Learned observation on trading cryptocurrency

From conducting trading over a number of years to developing a model that grows cryptocurrency quantities, it appears that smart traders don't need to keep adding more fiat capital into the market. The reasons for this is because sensible strategies seem to accumulate more cryptocurrency.

One major risk with cryptocurrency is drawdown risk - when the market collapses. It is essential to be able to take capital outside of cryptocurrency before the price crashes, and it is essential to put more fiat into the asset class in preparation for rises to avoid upside risk.

Optimal positions universal law?

If we think of this simply - there is a an optimal number of positions to be in at different stages of the cryptocurrency bull/bear cycles. We should be prepared to move between assets to try and maximise our returns, manage risk, and/or stabilise/maintain holdings.

A challenge from manually trading is our tendency to anchor ourselves to belief systems that may themselves cause significant losses. Our intention is to set up automated trading that moves capital between assets and instruments.

A simple example is;

• BTC will lose or gain by a certain amount in a set period
• Altcoins will often lose or gain by a certain amount in a set period that is higher than BTC
• Fiat (Standard Currency)

Cryptocurrency Automated Bot trading? Are you planning to build a bot?

We hear a lot on bots "sniffing out the shorts", "liquidating the longs". One way they do this is to look at Fibonacci levels and setting sells just below what seems to be a natural stop loss. This is because adding volumes to what is a natural buying level will squeeze the shorts into buying back in as the price increases. This makes sense, but does it work? Are bots adding enough volume to trading?

Whilst our platform could be configured to act like a bot this is not our intention.

How our automated trading application will work

There are four main phases to our cryptocurrency trading application;

1. Position Management
2. Market Analysis and Signal generation
3. Trade Rule Generation
4. Goal seeking and Backtesting

Position Management

We cannot discuss our detailed strategy or technology approach to doing this right now, but we think that Position Management is the most important. Position Management lets us set up buy and sell rules, and then manage those positions as those orders fill. We can manage the state of our positions to apply different strategies. In an ideal scenario, these position management trading strategies will run perpetually.

We are aiming to build a system that can mirror how we manually manage an active portfolio. Our interest is to keep rolling our portfolio positions to grow assets.

Market Analysis and Signal Generation - RSIS (Rate Slicing on Interval Snapshots)

As experts in Multidimensional Modelling and Relational Data Modelling, with querying of OLAP cubes in MDX and databases in SQL, the idea of partitioning data by different timeframes is a common experience of ours.

We have noticed that cryptocurrency market technical analysts and many online cryptocurrency trading exchanges focuses on price movements over hourly, daily, weekly, monthly, and yearly time periods. It is our opinion that markets present data to their audience in a format that can often be counterproductive to informing decision-making by a trader. A simple example is Daily Highs and Lows. We often see these highs and lows vary quite substantially during the day but we are presented with a single percentage figure that is supposed to represent our current price relative to that. This figure may be useful to helping us understand the market but given that we use many indicators to interpret the market - we feel that more is needed.

We have our own series of indicators that we use, and have suspicions will help us to enhance our ability to minimise risk and maximise gains.

Our Data Calculation Engine to help support Rate Slicing on Interval Snapshot Analysis

You can read about our Analyser Processor Data Calculation Engine here https://www.inforhino.co.uk/beta/data-calculation-engine . We wrote it to provide NoCode timeseries analysis on cryptocurrency and house prices. Importantly, we can compare and test values to determine whether conditions are hit or not. Having revisited this excellent application we are confident we can reuse this software on cryptocurrency price data as long as we do a small amount of effort to prepare the data for the engine.

Trade Rule Generation

Our Position Management application is designed to allow trade rule configuration to be dropped in to the application. The application works with whatever the latest trade rule configuration is. This means that we can generate trade rules from any application or define them manually. Suddenly, rather than having to integrate with the application, we can instead simply generate the trade rules.

Goal Seeking and Backtesting

Really, this is about testing to see whether we make profits, grow coins, and calibrate our models. Some trading exchanges offer test accounts, but in theory we may be able to replicate our own trading exchange.

The technological landscape

Should we write this software ourselves?

We have looked into software written in C# that could perform this trading but they always focus more on interoperability with exchanges. Whilst great pieces of software, we are more concerned with developing the application that will intuitively deliver what we need.

Rule based exchanges and automated trading bots - how about those?

A lot of thinking has been undertaken to determine whether trading bots where rules can be set up to perform trading could indeed avoid the need to write our software. A couple of platforms stands out in particular, our CEO is likely to be using for himself independently of Info Rhino. The costs for these platforms are also not expensive.

Why would we write our own automated trading solutions?

From our analysis so far, and our review of other platforms we are focused more on how we have modelled the market. How we see it as working. Whilst we cannot avoid the obvious synergies found in conventional financial and cryptocurrency markets, our inquiry is into how we can frame our trading strategies in a way that has less technicality to it. We are looking more to open up our platform than expecting users to log onto our portal to do things exactly how we expect them to be done. As with our other software, simply configuring files and options is enough to get our applications working. If a third party wishes to create a front-end, or generate trade rules using AI from price data - feel free.

The commercial landscape - what are we seeking to achieve with our cryptocurrency trading software?

At a very primitive level, we want to see if we can apply heuristics and technology to be a more efficient form of a human, to save time, and increase gains. If this can be achieved then we can start to look to train these applications with AI.

Another objective is to provide trade signals to our upcoming CryptoStatto platform. We don't expect this platform to be anything like a TradingView or a CoinRule, or any one of the major cryptocurrency exchanges.

Risk Management

Risk Management is a major concern with cryptocurrencies. Indeed, it is a subject all in itself. Repeatedly, cryptocurrency goes through major peaks and troughs, always on the hope it will follow the long-term upward trend. Many critics of cryptocurrency quite rightly point out "How can people build markets upon cryptocurrency when the price is so volatile?" One answer could be risk management, and applying how some successful trading software works may hold the key to stabilising cryptocurrency assets.

Partnership potential?

We do see partnership potential from our software we are building in the cryptocurrency space. We are not considering to open source the software yet.

Do contact us if you wish to find out more. Perhaps you wish to collaborate with us - https://www.inforhino.co.uk/article/Collaboration/Partnership/Collaborate-with-Info-Rhino

Commercialisation - what is your business model for this software?

We are not intending to create a B2C trading engine. Our intention is to stabilise our position within the cryptocurrency space, and seek to use technologies we develop and discover to strengthen that position.

Why are you not choosing a blockchain and developing on that?

We have explored a number of blockchain, DAOs, and Smart Contract based blockchains. We have explored NEO, Ethereum, Cardano, to name but a few. We have also liked the runtime languages that compile to different blockchain virtual machines. We still feel price discovery is a big part of blockchains, and so are staying in this area at the moment.

Breaking the mould - not following convention

Building Business Intelligence Platforms, applications across many financial products including swaps, Repos, Equity, SLB, is really has been a pleasure. Learning about new ways risk mitigation has been attempted in liquidity, market, and credit following large systemic shocks within the financial system, has been very interesting. It is a concern that certain players within the cryptocurrency space seem too keen to get us on a Financial System 2.0. We saw this with Celsius recently for example - their company filing for bankruptcy.

Thank you

There is a lot packed into this article, and a lot we had to leave out. We are now focusing our efforts on building out decent prototypes before deciding how to progress this cryptocurrency trading technology.

Thanks again.