Lykke Exchange Open Data

By Sergey Ivliev

As a part of its DNA, Lykke is committed to releasing its own intellectual property to the public domain: the Lykke Exchange code base is available at Github, all competitions results are published under Creative Commons license, market making methodology is made available in research papers.

From February 2018 Lykke has started to publish the log of all the trades that were matched on the Lykke Exchange

Tradelog download page screenshot

The tradelog report for each trading date is published on the next day after midnight. Let’s look inside — it normally contains a zipped text file with the following fields that describe the trade:

  • DateTime field with the millisecond timestamp (UTC+0) of each trade,
  • TradeId field with a unique trade identifier (each trade is presented with four lines: buy and sell trade legs for the buyer and the same for the seller),
  • Asset contains symbols of instruments exchanged with the corresponding Volume,
  • OrderType contains the initial order type and the moment when it was placed,
  • WalletId has the identifiers of the counterparties.

In the example below trader #cea147ac has placed a limit order at 1:23:32.657 to buy 42.92 ETH for 9002.819 CHF (@209.758 ETH/CHF), approximately 2 seconds later it was matched with another limit order sent by the trader #ea7348c2, which crossed the book and triggered a trade.

Trade fields of the trade in the tradelog

Why do we think it’s unique? Many crypto exchanges disseminate their tradelog information, but 1) normally this is only available via API with a limited period, 2) it is practically never available with traders ID resolution (except for decentralized exchanges, where tradelog can be extracted from the blockchain itself).

This is also not a common practice for conventional exchanges. In most cases, tradelog studies are held on a limited scale by the universities who are granted access to these data by exchanges or regulators, for example:

  • Liquidity provision and market making by HFTs on the Canadian equity market (link)
  • Identification of clusters of investors from their real trading activity in a Nokia stock (link)
  • Market impact of large orders (metaorders) executed in the U.S. equity market between 2007 and 2009 (link)

Overall market data transparency is not only good to study market regularities but also key to provide transparency and ouster manipulative behavior, which remains one of the most unpleasant phenomena on unregulated cryptomarkets.

About us

Lykke Research Hub is an attractor for research related to crypto markets and blockchain tech and playground for Lykke to experiment with new products, financial instruments, and UX.

We welcome academia and industry practitioners to submit research proposals. Best proposals will be funded. More details on Lykke Streams.

To stay up to date please follow Lykke Research Hub twitter and lab page on ResearchGate.