A New Paradigm for High-Frequency Finance

The financial econometrics literature relies on a well-established paradigm: log-prices of financial assets evolve as a jump-diffusion process contaminated, at high frequency, by microstructure noise. This paradigm omits a fundamental feature. Intraday prices of financial assets are stale: they show a large incidence of zero returns.

PRICE aims at overcoming this imperfect paradigm by proving that the transition to a new one is unavoidable.

We will progress in several directions:

1. The definition of a general reduced-form model for asset prices featuring genuine (due to lack of trading activity) and spurious (due to price rounding) staleness. Novel testable implications and estimation methods will be proposed.

2. The definition and estimation of a micro-founded model of price formation capable of explaining the economic determinants of staleness. Derivation of model-implied estimates of market and funding liquidity and new insights on their interaction.

3. Theoretical and empirical evidence that popular moment estimates based on high-frequency data (realized moments) and the related (classical) inferential methods are affected by staleness. A new asymptotic theory for staleness-robust estimators and new empirical economic implications.

4. Empirical evidence that the cross-sectional pricing power of realized moments is distorted by staleness. We will revisit the pricing of realized moments in light of the new paradigm.

5. Empirical evidence that jump identification tests based on the old paradigm are not reliable, since staleness generates spurious jumps. A new theory capable of distinguishing between real and spurious jumps will be proposed.

6. Early-warning indicators, useful for policymakers, of market instability through staleness and implications of staleness for risk management.

7. A new framework to price options written on stale assets. Estimation of the pricing model and derivation of illiquidity risk premia and pricing errors associated with the old paradigm.