Econometric Project help!

[leyjl2]

Hi there guys,

I'm currently doing a project for my undergraduate studies to test for seasonality patterns in stock market data. Right now I've got about 4 decades worth of monthly returns data of the hang seng index from 1970 to 2014 and im using the OLS first then the ARCH model to test for any significant results in the returns. So ive sorted out my previous issue, but I've hit upon another one: basically, I'm finding issues proving that the data has ARCH effects, I ran a test for ARCH with 1 lag but rejected the null (p-value: 0.17), then I ran the test again with 2 lags and it came out significant. How do I interpret this? Could I still fit an ARCH model onto this?

Any help or suggestions would be highly appreciated!!!

Thanks in advance! :)