I'm testing the efficiency of the capital market of my country; when I used the autocorrelation test for the index returns, the Q-test shows that the market is not efficient at one and two lags levels (rejecting the null hypothesis of randomness), and after that, the returns are independent, and there is no correlation between them (from 3lags up to 30)

So I got confused about what can I conclude about the overall efficiency of the market?