I am a good test case for this question. I was very active for the year following my joining - 470 answers between June 2011 and July 2012. For the last five years, I've only posted 45 additional answers, and haven't posted any since August last year. Questions too - out of my 133 questions, 128 of them happened during that first year. My last question is nearly three years old.
So my reputation gain over the last year reflects only votes on posts at least a couple months old, also keeping in mind that over ninety percent of my posts are over five years old.
On Oct 9, 2016 I had 42,677 reputation, and as of Oct 9, 2017 I have 44,844 reputation. That's somewhat over 2k, or just about exactly 5% per year. I have not been active, so this number has not been affected by bounties or my downvoting.
I'd say then that you want to have around 20K rep in order to idly fund 1K rep spending per year. Maybe 21K if you want to be sure to retain 20K privileges throughout the year.
Edit: OK, so curiosity is killing the cat. I knew there must be a SE data query that would reveal "passive" reputation in some meaningful way -- and sure enough, there is, now. Check it out. There's quite a significant margin of error between users. Hopefully it's pretty self-explanatory, but what we find is about exactly what my case seems to show. All users earn an average of over 1000 passive rep per year with the exception of two users, until you get below 21K rep total. The chart shows I average over 3K passive rep per year, but remember that is an average over all six years I've been here. Apparently, it slows down as my posts become enshrouded in the ever more distant past.
To address your own standing: you have averaged... wait for it... 999 annual rentier income. KaShazam!