α52
Alpha 52
Where AI meets Quant

Frequently Asked Questions

Performance & Results

What is the purpose of this website?

This website showcases how AI and Quant can deliver exceptional results in active asset management by overperforming the benchmark.

Could I rely on shown signals to build a portfolio?

Yes, using pyramiding techniques to allocate asset weights. Real-money results are published on LinkedIn.

What is the average trade lifetime?

It ranges from 4 days to 5 weeks on average.

Could I rely on shown signals to place swing trades?

Possible, but entry and exit are different based on your risk appetite as detailed in User Guide.

Technology & Methodology

What is the Technology Stack?

TensorFlow (AI framework), PostgreSQL (database), Railway (deployment platform), Supabase (backend-as-a-service), and DeepSeek (LLM for news extraction).

What are the website creator's qualifications?

PhD in Mathematics, Financial Risk Manager (FRM) certification, Chartered Market Technician (CMT) Level 1, and TensorFlow developer certificate.

What is the AI doing?

The AI model is trained on historical price movements and market-switching regimes to predict future stock movement, determining if it is above or below the money market threshold. The AI model returns a probability that is plugged into another model to generate the signal measure.

What is the Quant doing?

The Quant uses advanced mathematical and noise filtration techniques to produce an output that is passed to another model, along with the AI model output, to produce the signal line.

What is the meaning of the signal measure?

The signal is a number ranging from -100 to +100 that describes the stock movement as: Trending Up, Slowing Down, Trending Down, or Recovering.

How could I convert the signal to trades?

Monitor the signal line and consider buying when the signal becomes positive if you are risk-averse. If you are willing to take higher risk, you may buy the stock when it shows a Recovery signal.

Data & Coverage

What are the inputs to the model?

Only market data (OHLCV) is used, as no other data is available.

What is the projection period?

As we use only market data (OHLCV), models are trained to project one week ahead. If fundamental data (Earnings Reports) and alternative data are included, projections could extend to around 3 months and beyond.

Where is the financial analysis for company earnings?

Initial analysis shows promising results for applying AI to analyze company reports, but it is not included, as it requires more storage and processing power, which is outside the allocated budget for launching this website.

Why is coverage limited to the US market only?

Because data for other markets is either expensive to access, not in a standard format, or of poor quality, requiring extensive cleansing.

Why is the US market not covered in full?

Because servers are not free!

Home Stock Hot Lists Insights
Support expand_more