Algorithmic Trading of OTC and Commodity Exchanges with gRPC and a free choice of programming language

When you discuss algorithmic trading functionality in energy trading you experience a great spectrum of reactions. Compliance is concerned about safety, the business is concerned about long term profitability, IT is concerned with hardware requirements, e*star is made to cater to all those needs, but in this essay, we’re going to focus on the needs of the Quants. We’ll discuss how e*star Algorithmic Solutions can make the life of the Quants easier by enabling them to work in the programming language they prefer, by offering a scalable and an exchange independent solution, and by providing advanced out of the box features that help the Quants to simply focus on their complex execution strategy instead of struggling to recreate exact orderbook states in their code.

The e*star Algorithmic Solution is based on protocol buffers by Google and offers an advanced gRPC streaming API through which the algorithmic execution strategy of the Quant is connected to the external markets. This design decision allows the Quant to write in their native programming language, like Java, C#, or Python, but also GO, C++, Dart, or Kotlin would be on the menu if this is desired. To help a Quant to write in the language they are most comfortable in does allow for quicker time to market, but also it minimizes execution risk in production, as the code does not need to be translated during expensive intermediate steps by developers whose sole existence seems to be build around showing the Quant who is in charge and then they somehow forget to take responsibility if the strategy produces errors.

The core concept of the algorithmic trading at e*star relies on leveraging on the e*star Internal Market, this means that an algorithm does not directly execute on the external exchange but generates order proposals on the e*star Internal Market, which can then automatically be routed to any preferred exchange by a Market Access Trader. This setup allows to scale algorithmic use and the algorithmic strategy can be run on any exchange with the same or similar set of products. If multiple strategies are run at the same time the e*star Algorithmic Solution can effectively prevent issues like self-trading on the exchanges, also strategies may optimise their own position with each other without the need to trade externally, hence the more strategies are run in parallel the more profitable the setup may be without compromising on reporting of trades and transparently showing real-time positions. The code which was written by a team of Quants for one exchange may simply be run on another exchange by switching a few core parameters and setting up the routing to the new exchange.

By leveraging on the e*star Internal Market the algorithms are readily offered advanced out of the box features. The Quants do not need to actively manage their own financial trading limits, as this will be done by the Internal Market. Also making sure to trade only on products which the strategy should is quickly setup in the Internal Market. Receiving only a limited depth of orderbooks can also be setup, hence there does not need to be a complex logic to find the best bid or best ask of any orderbook. Making sure to stay up to date with the latest exchange APIs will be handled by the e*star regular release cycle. Allowing the matching of a few strategies (e.g., Power NL with Power GER) while preventing the matching of other strategies (e.g., Power GER with Gas NL) can be quickly setup in the Internal Market. Writing relevant Audit Logs will be handled by the Internal Market.

Yet the best thing about e*star is that we are very much willing to adopt our software to the customer needs, we’re striving to become the best fit for you, so get in touch.