Modeling Selected Asset Spreads with Time Series Methods
Summary
The time series approach begins after a pair or group of assets has already been selected, for example through cointegration testing. It models the resulting spread to produce trading signals, shifting the focus from finding related securities to deciding how to trade their relationship over time.
The document illustrates predicted and observed spread values from an Auto ARIMA model used in a quantile pairs strategy, and lists other possible tools: state-space and Bayesian models, Ornstein–Uhlenbeck processes, and nonparametric methods such as Renko and Kagi. It offers no performance results or implementation details, and the plotted example alone does not establish that forecasts are accurate or profitable. Model choice and signal quality therefore require further evaluation.
Key ideas
- The time series approach assumes related assets have already been selected.
- It models the spread to generate trading signals.
- Auto ARIMA is shown as one way to predict spread values.
- State-space, Bayesian, Ornstein–Uhlenbeck, and nonparametric methods are listed as alternatives.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.