The Hidden Cost of Volatility Drag

The microstructure of financial markets is a complex and intricate beast, comprising various factors that influence the behavior of securities. One critical aspect of market microstructure is the role of autocovariance in generating returns.

Autocovariance refers to the relationship between two variables at different times or across different frequencies. In the context of multivariate microstructure models, autocovariance can be used to estimate the covariance matrix of the system. This matrix plays a crucial role in understanding market dynamics and making informed investment decisions.

The Autocovariance Generating Function

The autocovariance generating function (AGF) is defined as the polynomial representing the autocovariances. In this case, it takes the form:

g@Θ, Σ2, z_D := CollectAHΘ . L ® zL Σ2 IΘ . L ® z-1M, z

where ΘHLL is the lag polynomial for an MA process xt = ΘHLL Εt.

The VMA Representation

The vector autocovariance generating function (from VMA representation) can be computed from the autocovariances. This version of the notebook does not include this calculation.

A 10-Year Backtest Reveals...

When analyzing the impact of microstructure on returns, it's essential to consider the long-run variance. The autocovariance at lag k represents the coefficient of zk in the MA process xt = ΘHLL Εt. In this case, the autocovariance is given by:

Θ@1D.W.Transpose@Θ@1DD . ΘRulesL.HW . WRuleL.Transpose@HΘ@1D .

Portfolio Implications

The results of this analysis have significant implications for portfolios. One important consideration is the presence of a random-walk component underlying both securities.

Actionable Insight: Consider the VMA Representation

When constructing your portfolio, keep in mind that it's not just about selecting assets with high returns; you need to consider their microstructure as well. This means understanding how autocovariances can impact returns and using this knowledge to make informed investment decisions.