Seasonality in Commodities Market and Generalized Seasonality Test via Sequential Regression Testing of Prime Cycle Periodicity
Quant Arb X HangukQuant
I am proud to release our post, starring author/quant of the
blog. In this paper, we study seasonality portfolios in the commodity futures market, and implement novel, generalized seasonality tests using sequential regressions. In particular, sequential dummy variable regressions for prime cycle periodicity are used in the hypothesis test for existence of seasonal patterns against no seasonality in computing the p-value of seasonality in observed time-series, without any prior placed in regards to the structure of the calendar anomaly.This generalized seasonality test can be used on any asset class, to test for the presence of seasonal effects in high-frequency, daily, weekly and other data granularities, for flexible range of values in periodicity, while controlling for the familywise error rate encountered in multiple hypothesis testing environments.
This post is made available for all readers. Python code is provided.
For readers who enjoyed this paper, feel free to follow
‘s excellent blog and discussions on his Twitter feed at @ quant_arb.Download paper:
..Preview
..
Code Script:
Our last formulaic alpha report:
Our last capital market notes:
Happy Trading!
 | A guest post by
|