Order-flow and long-memory in a simulated financial market

<p dir="ltr">This is the dataset used in our Honours Project on "<a href="http://10.25375/uct.30453044" rel="noreferrer" target="_blank">Order-flow and long-memory in a simulated financial market</a>". It can be used hand-in-hand with our <a href="https://github.com/STA-HONS-PROJECT-RSSMIC-SLVSHA-LMF/STA_Honours_Project_LMF" rel="noreferrer" target="_blank">code</a> to replicate our findings. The data consists of intraday quote and trade tick (trade-by-trade) data for the top 42 stocks listed on the JSE, from January to December 2015. This data does not contain Trader IDs (a way of knowing which trade is associated with which trader) and thus one needs to apply a method for generating synthetic trader IDs for these trades. We used the <a href="https://arxiv.org/abs/2503.18199" rel="noreferrer" target="_blank">Maitrier-Loeper-Bouchaud (MLB)</a> algorithm once we had processed this raw data. Trader IDs are required in order to calculate metaorder lengths, which is a key component needed when estimating the microscopic exponent 'alpha'.</p>