Traditional quant trading firms find alphas by using technical indicators and hand-crafted features, which cannot explain the overall market actions efficiently.
They then apply traditional time series analysis, statistical or financial models that cannot fully capture the underlying market dynamics, easily overfitting their models and resulting in fast decay in their alphas.
Quantrend Technology takes advantage of modern end-to-end deep learning models, which can automatically extract robust and high-quality trading signals (Alphas) that cannot be captured by traditional approaches.
Instead of predicting market trends without considering the volatilities involved, our models can learn the whole market dynamics as a probability distribution, allowing risk-adjusted returns to be effectively predicted in all time scales.
At the same time, inter-market dynamics can also be learned end-to-end, which is often intractable using traditional approaches.
Quantrend Technology applies machine learning-based risk management and portfolio diversification system to help clients of different risk appetite to reach maximum risk-adjusted returns under the exposure of controlled risk.