New York-based asset manager, qplum, has launched a multi-strategy AI managed futures program (QMAP) for qualified institutional clients.
QMAP aims to give investors access to a diversified investment strategy that trades across different geographies and asset classes. It trades futures such as fixed income, equity indices, FX, commodities and volatility. There is a drawdown control-based risk management in place.
The strategy is built using qplum’s proprietary, deep learning framework that already powers other portfolios offered by the firm. Large amounts of market, economic and other structured data are used to train the models and the entire trading pipeline is fully systematic. Under the supervision of CIO, Gaurav Chakravorty, qplum’s in-house team of more than 15 engineers and quantitative portfolio managers are responsible for the oversight and improvements of the models.
“There aren’t many truly diversified, AI driven investment strategies in the managed futures space,” says qplum CEO, Mansi Singhal. “Instead of betting on one signal like trend following or mean reversion, QMAP diversifies across roughly up to 11 signals that allows it to react and evolve with the market.”
A combination of traditional and data-driven signals are used; examples include systematic macro, mean reversion, feed-forward neural network and autoencoder. QMAP is designed to target very little correlation to both the stock market and trend following strategies.
According to qplum, the QMAP strategy could be a good complement for investors who already have exposure to trend following CTAs and are looking to diversify.