Basically rate of return makes the assumptions that you bought all your current assets (shares typically) at the beginning of the year and just held.
Personal rate of return factors in that shares may have been bought or sold over the period and as such winds up being a more true reflection of what is your actual return.
To illustrate with a dumb example:
If I dumped 10K into the S&P index 12/31/18 I'd be up 28.86%: the rate of return and my personal rate of return would both equal this.
On the other hand if I dumped that same 10K in 90 days ago, the rate of return would still be 28.86% but my personal rate of return would be 11.88%. Still completely solid but a far cry from the trailing 12 months one.
In my case I threw a bunch of money into my equity accounts this year: my total rate of return was 42.15% this year, whereas my total personal rate of return was 27.23%. Still fantastic just not quite as awesome haha... I focus more on dollar values and regardless of the percentages, this was a stupidly good year for me.
If you can get money-weighted do that; however, it's not easy calculating all of the inflows and outflows when we're talking future growth of a portfolio.
Can be done if you have regular inflows and know the costs (management fees / withdrawals / etc) otherwise it's hard to really get it done accurately but it's a better representation of reality.
If you are just looking to see roughly (or exactly depending on investment) how much 20K will be worth in 15 years doesn't matter, either is fine... if you're at 20K now and putting in 1K per month and there's some expense ratio associated with the investment (index fund let's say) and trying to predict what you'll have, money-weighted is the way to go.
Time-weighted when we're talking investment performance when it's up and to the right is going to paint a rosier picture than it really is for many investments, certainly for any managed investment.
