with Shamim Ahmed, Daniel Tsvetanov, and Xiaoxia Ye, 2024, working paper.
[Paper] We compare factor models using “test-asset irrelevance” of Barillas and Shanken (2017) but
under regime switching. In doing so, we show that unconditional squared Sharpe ratio is
no longer appropriate for model comparison when returns are subject to regime shifts. But
conditional on a given regime, model comparison can be performed using conditional squared
Sharpe ratio. We find that the Fama and French (2018) model employing cash profitability
performs the best under the bull regime, irrespective of accounting for transaction costs
as per Detzel, Novy-Marx, and Velikov (2023). But no model performs better than other
models under the bear regime.