The Purchasing Power Parity (PPP) Puzzle refers to the difficulty of reconciling the enormous short-run variability of real exchange rates with their longer-than-expected deviations from equilibrium (Rogoff, 1996). Recently, Imbs, Mumtaz, Ravn and Rey (2005, hereinafter IMRR) have argued that much of the PPP puzzle is due to upwardly-biased estimates of persistence. According to their view, the source of the bias is the existence of heterogeneous price adjustment dynamics at the sectoral level. As they put it, "the aggregate real exchange rate is persistent because its components have heterogeneous dynamics". This paper re-examines this claim in two steps. Firstly, we demonstrate that IMRR's measures of sectoral persistence are systematically downwardly-biased because they are based on an inaccurate definition of the "average" Impulse Response Function (IRF). We then show that standard estimates of shock persistence are recovered after this bias is corrected. Secondly, building on the results in Mayoral (2007), that prove that aggregate and micro models induce similar shock persistence behavior, we show that estimates based on aggregate and sectoral exchange rates are in fact highly compatible. Therefore, aggregation does not solve the puzzle and further research is still necessary.JEL classification: C22, F31, O11.