I've been making a working assumption of about 2% for quite a while now. Of course it's difficult to test this theory over the short term because of variance issues (the variance is larger than the trend). The only thing that would change this would be a major shift in the BTC price again.