15 Seconds to Alpha: Higher frequency risk pricing for commercial real estate securities

This paper introduces a generalizable method to estimate reduced form risk decompositions at daily and intraday frequencies applied to CMBX. We estimate partitions for the risks of default, liquidity, excess liquidity, and interest rate volatility at daily and intraday frequencies. Our new estimation technique combines previously simulated risk partitions with current market data using principal components and OLS methods. We find liquidity and excess liquidity risk partitions are significant in explaining daily effective bid-ask spreads historically, from 11/2007-4/2019, and in 20-day forecasts. During the Covid pandemic, we extend the model from daily to intraday frequency, estimating intraday in 15 second intervals over the period 4/2020-4/2021. During Covid, we find regular patterns of risk partition volatility in the cross-section and exploit those insights in the related, and more frequently traded, REIT sector in automated trading strategies. In our 54 long/short day trading strategies, 96% showed significant alphas, and 63% produced abnormal cumulative returns between 0.73% and 48.74%. These results support pricing risk with risk partitioning at higher frequencies for commercial real estate securities.