FishyFin

Logo

Inspiration

We were inspired by market makers trying to visualize trades that align, or not, with a typical order flow. Being on both sides of a book order, it is crucial for them to be wary about those trades. Our projects simulates a real-time order-flow and fishes out the bad apples amongst clean exchanges with the market.

What it does

The code first filters trades based on their vector of events (Ex: NewOrderRequest->CancelOrderOrder, a set of 4 events, etc)
The vectors are then displayed on an interactive 3d graph
Key stats are displayed on the side (total trade passed, total order sent, total order received)
A statistical model based on standard deviation to the mean is then applied to fish out trades that might not align with the typical orderflow as well as trades that are novelties, or symbols that are not typically traded

How we built it

The whole project was built in python using streamlit and dictionaries for fast results and good displays

Challenges we ran into

Finding novel ideas for otherwise typical order flows of markets was the main concern here, hence the idea of fishing out trades that seemed suspicious on the trader side.

Visualizing fishy trades inside of the orderflow