Sticky liquidity

[Fran Zayn]

Hey guys,

i want to do a regression where I show that liquidity is sticky.
So if we have high liquidty today then it should also be high tomorrow unless some unexpected shock occurs.
Do you know how i can illustrate this the best in a regression model?
is this model the right step?

spread_t = a + b * spread_t-1 + controls + error term_t-1

or do I need to add time dummys?
I am assessing an Event which occured on a specific date and test wheter it had an impact on liquidity

Thank you!

[Phil Bromiley]

This sounds like you're likely to have panel data - see the documentation on panel data. If you have a lagged dv, then you need to look at instrumental variables, xtabond, or xtdpdmls.