The evidence of volatility-price dependence observed in previous works (Karakatsani and Bunn 2004; Bottazzi, Sapio and Secchi 2005; Simonsen 2005) suggests that there is more to volatility than simply spikes. Volatility is found to be positively correlated with the lagged price level in settings where market power is likely to be particularly strong (UK on-peak sessions, the CalPX). Negative correlation is instead observed in markets considered to be fairly competitive, such as the NordPool. Prompted by these observations, this paper aims to understand whether volatility-price patterns can be mapped into different degrees of market competition, as the evidence seems to suggest. Price fluctuations are modelled as the outcome of dynamics in both sides of the market - demand and supply, which in turn respond to shocks to the underlying preference and technology fundamentals. Negative volatility-price dependence arises if the market dynamics is accounted for by common shocks which affect valuations uniformly. Positive dependence is related to the impact of asymmetric shocks. The paper shows that under certain conditions, these volatility-price patterns can be used to identify the exercise of market power. Identification is however ruled out if all shocks affect valuations uniformly.