I created simple program to watch prices and trade at 3 good exchanges: Poloniex, Bittrex and Cryptopia. All of them provide simple API to watch cryptocurrency prices and allow trading. I needed some features to be able to trade efficiently and I wanted to create trading theories a verify their effectiveness and finally make a trading bot.
The feature I needed something to analyze spread. Especially Cryptopia exchange has less popular coins with higher spread and bigger movements. Spread is very usefull information but can be easily misleading. Imagine you have last price of BTC/LTC 0.015 and current order book has highest bid of 0.0149 and lowest ask of 0.0157 and now someone adds new ask order of 0.0152 but only for amount so small it is basically zero (e.g. 0.0000001 BTC), now you get new lowest ask of 0.0152 but it is highly misleading since if you want to trade amount higher than 0.0000001 BTC, the last added order is irelevant for you. Similar situation is when you look deeper into order book and you see lot of small trades that could be easily aggregate with bigger trades. My application aggregates orders into groups that are about 1 % apart.
Graphs of order book at most exchanges seems to me very useless since they look most of the time the same. They are also so small to notice prices that will be harder to break than others.
My application doesn't create sum of prices and therefore it is easily readable.
Another special feature is graph of buy and sell trades sepearated of each other which shows you spread and represents last price more accurately. Following picture shows prices at top and volume at bottom.
The application is by no means finished, I will continue to add new features and fix bugs. If someone would find it interesting I might send you the application, if you have an idea of an algorithm, I might try to implement it and evaluate it or if you wish to continue to work on the application itself, I could upload it on github. It is .NET WPF application written in C#. Let me know in the comments.
