AREIX, your AI-powered financial assistant

AREIX helps you to manage your personal finance and spending effortlessly using artificial intelligence and behavioural economics on your smartphone. AREIX is all you need to take control of your financial destiny.

Contact Info
1/f Lion Rock 72, Innocentre
Tat Che Road, Kowloon Tong, HK
Follow Us
AlphaGen > Details

Practical Decision Fault Tolerance (PDFT)

- by Bigto Chan   Ronnie LI   Hunter Zhan  

Link Copied to Clipboard

Cumulative PNL


Daily PNL


Model Accuracy


Real Time Model PNL

Model Scoring Contest kicks started in
Days Hours Minutes Seconds

Investment Objective

Investment Objective: As a team made of computer science and finance students from City University, we strike for the best balance between quantitative and finance perspective for our trading strategy. It’s an all-weather trading strategy to maximize our return on cryptocurrency on no matter bull, bear or range-bound market. Trading Strategy Description: We select MSAD (Moving Average Convergence & Divergence) as our trading decision indicator. When the MACD number turns from negative to positive, we place an all-in buy order. When the MACD number turns from positive to negative, we place an all-in sell order. We will only trade on XRP/USDT and our trading interval is 1m. Model development philosophy: We think it’s quite hard to capture the fundamental side of cryptocurrency though we believe major news breakout will have a significant impact on the price movement. So we decide to leverage technical analysis. Out of mass technical indicator including RSI, EMA, Stochastic, MACD and so on, we decide to use MACD because it provides the most prompt and accurate buy/sell signal. With MACD, one sell order always comes after one buy order so we shoot all bullet and no need to control position. Then, we run our strategy on many cryptocurrencies (BTC/USDT, ETH/USDT, DOGE/ISDT, XRP/USDT...). It turns out XRP/USDT has the highest average profit. Based on that, we tried different trading interval (1m, 5m, 15m, 30m, 1h, 2h) and the result is that 1m gave the highest average profit of 22033.0202.

Model Creator

Bigto Chan

City University Computer Science Graduate

Ronnie LI

City University Computer Science Graduate

Hunter Zhan

City University Finance Graduate