High Frequency Liquidity

JEL codes: C58, C12, C13, G12

Liquidity is a key ingredient for the functioning of financial markets. Nevertheless, it remains an ambiguous
concept with various dimensions related to trading volume, transaction costs, price impact and availability of
funding. The difficulty in measuring it and interpreting the results is notorious; moreover, there is no consensus
on the informativeness of the several existing liquidity measures.
This project will advance the state of the art in this feld by developing novel econometric tools to measure the
liquidity of assets traded in financial markets. Our liquidity measures are labeled realized since the proposed
research agenda builds on the realized volatility literature (to which the PI contributed substantially) to open a
brand new strand of research. We aim at characterizing functionals of the observed returns to be analyzed with
a rigorous asymptotic theory when the interval between consecutive prices goes to zero. The interpretation of
the asymptotic limit in terms of illiquidity is justifed in the framework of economic models of price formation
in which rational agents with heterogeneous information interact. The advantages of the proposed
methodology are the precision guaranteed by high frequency data and the neat interpretation of the measure.
Two specific techniques are explored. The first is meant to measure liquidity with unprecedented accuracy for
many stocks on long time spans; the second is targeted to describe cases of financial distress in which a
liquidity shock degenerates into a liquidity crash.
New realized liquidity measures will allow applied researchers to reconsider the existing results in the
literature, for example to assess the magnitude of illiquidity risk premia; pin down the relation between
illiquidity, credit and variance risk; shed new light on the dynamics of liquidity spillovers; provide novel
empirical support for economic theories of price formation; and implement a real-time indicator of liquidity
dry-ups.