In our latest Forex Focus podcast, UBP’s Global Head of FX Strategy Peter Kinsella interviews Saeed Amen of Turnleaf Analytics on the evolution of market research with artificial intelligence.
Advances in artificial advance are taking the financial industry by storm, and nowhere is this more evident than with the use of AI in economic forecasting techniques. Forecasting methodologies have become increasingly complex over the years, with macroeconomic researchers using a wide variety of models to calculate trends in everything from inflation to GDP data.
The use of AI means that economists can now use significantly wider pools of datasets which feed into larger and more complex models, giving them a real-time view of what is happening in the economy. Increasingly, these AI models rely on higher-frequency data from a wide variety of sources – especially alternative data (known as ‘alt data’ in the industry). Examples include Google Trends data, social media sentiment and even the use of iPhone data to track mobility. The advent of AI means that huge datasets can now be constructed and analysed in a matter of minutes, with machine learning and AI now highlighting new trends in the data to researchers.
Several prominent hedge funds have long since utilised AI trading models – particularly in the use of algorithmic and high-frequency trading. However, the use of AI within the wider finance industry is at an early stage and the industry is only now starting to contemplate its possibilities. This will have several early-stage implications – from an increasing use of AI in the construction of asset allocation strategies to real-time risk management which can run several risk management models at any given time. The ultimate aim is to enhance investment outcomes, and so far, we can see that AI’s use in finance has had positive results. Trials have shown that AI models are more accurate than traditional economic forecasting methodologies – something that we think will revolutionise traditional economic analysis.
We were joined by Saeed Amen of Turnleaf Analytics to discuss these trends. At Turnleaf, Saeed and his team have been pioneers in the use of AI in economic forecasting. We discussed the use of alternative datasets, which can incorporate up to 500 data series just to forecast US inflation data (interestingly, pricing data about Broadway shows is a leading indicator for US inflation developments!). Forecasting European inflation data is a slightly more complex affair, and his models use up to 1,000 separate data series – which is explained by the different taxation and economic regimes across Europe. His team expects that US inflation will rise in the second half of the year – by more than current market expectations, with European inflation data expected to rise only marginally. This podcast covers all of the latest developments in the use of AI in economic analysis, and it is a must listen for anyone interested in this area.
