Why do some quant strategies fail? It has to do with all of the human decisions that go into setting up the machine.
The top three addressable data problems with quant models are data cleansing, data normalization, and data ranges, wrote Nicholas Abe, Boosted.ai’s COO, in a byline for TabbForum. In the piece, he explores these three blind spots in machine and data-driven investing, and outlines ways portfolio managers can program around these potential pitfalls.
In Nicholas’ experience, a combination of fundamental and quantitative experience is key to bringing the right knowledge and data to train any investment machine. Nicholas suggests cleansing data manually by finding problems and through system checks that ensure data makes sense to both the human and the machine. For a quant strategy’s models to work, incorporated features must be normalized. Applying normalizations makes values comparable across an entire universe. And when it comes to data ranges, quant strategies will do better with more data, since machines are driven entirely by data.
The real-life performance of a quant strategy is only as good as the human running it. Nicholas has found that it is not the machine or a portfolio manager alone that produces the best results. It is the union of the two. Taking time to ensure the proper data and machine construction at the onset is often the difference between success and failure.
To read the full article, visit TabbForum here.