We show that the interplay between investor behavior (LETF fund flows) and index return autocorrelation (“see-saw effect”) plays a central role in either moderating or amplifying the portfolio rebalancing demand of levered and inverse-levered ETFs (LETFs). Rebalancing, in turn, affects the underlying index’s volatility. Divergent investor expectations of LETF and other market participants, channeled through LETFs, leads to increased volatility but can also improve price discovery. LETF trading can provide a channel for noise-reduction in the price discovery process, especially during market downturns.