Forecasting Consumer Discretionary ETF Performance with Composite Quarterly Data

Summary

The article proposes forecasting SPDR’s XLY consumer discretionary ETF with quarterly macroeconomic and consumer data rather than relying mainly on the ETF’s own price and volume history. It reviews candidate datasets covering economic indicators, interest rates, employment and inflation, consumer surveys, retail sales, and demographics. The rationale is that income, purchasing power, borrowing costs, confidence, and retail activity can influence demand for non-essential goods. The author compares feature importance across datasets and suggests combining the most important features into one composite dataset for a model.

The article describes using an MLP and discusses data consistency, alternative model comparisons, cross-validation, and broker tick data as considerations for system development. It reports that the backtest and forward test were not performed on XLY and that data quality needed improvement; the reported results therefore do not establish predictive performance for the ETF. The composite dataset is presented as a potentially useful research direction, not a validated trading system. The available text is incomplete, so the detailed dataset construction and results cannot be fully assessed.

Key ideas

• Quarterly macroeconomic and consumer datasets can provide candidate features for forecasting XLY.
• The article considers economic activity, interest rates, employment, consumer surveys, retail sales, and demographics.
• Feature importance is used to select inputs, with the leading features proposed for a composite dataset.
• The reported tests were not conducted on XLY, and the author notes data-quality limitations.
• Alternative models, cross-validation, and broker tick data are relevant to further validation.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.