Betting Kelly on the Wrong Probability

An even-money coin truly wins with probability $p=0.55$, but you overestimate it as $\hat p=0.65$ and bet the Kelly fraction implied by your estimate. What is your actual long-run expected log-growth rate per round? Compare it to the growth you would have earned betting the correct Kelly fraction, and state what the sign of your actual growth implies.