Bearing in mind that we know almost nothing about your game, so any kind of estimates are merely conjecture...
Also bearing in mind that I'm just a hobbyist trying to go indie, and not in the professional game development industry...
Hosting costs are less than development costs. I did some basic back-of-a-napkin hypothetical calculations on server costs at the bottom of this post.
A 500 user concurrent MMO isn't really a Massively Multiplayer Online RPG, just a ORPG - an Online RPG. Though people often mix the terms up, to be 'massive' the term was coined for 100k subscribers, meaning more than about 5-10k active at once. The title isn't fully defined though, what makes a game 'massive' or not is up for debate.
500 concurrent users could probably be hosted (depending heavily on the nature of the gameplay of the game) at less than $1000 a month. I'm assuming a single one of your servers can run 50 users concurrently (optimized, and depending on the nature of the game, they could handle more), and so you'll need 10 servers running 24 horus a day, 30 days a month, at an Amazon.com-hosted price of $0.120 an hour.
For a simple 2D rpg only hosting 500 concurrent users... I would estimate a decent programmer (again, depending heavily on the nature of the gameplay of the game) could do a decent job on it by working full-time for two or three years. This means, paying a wage of $35k a year (far below the average programmer wage of ~70k a year), and hiring two programmers working full time, and taking two programmers two years to make, it'll take $140k for you to pay your programmers.
Toss in another 35k a year for two artists who, if you're lucky, also have musical skills, a quarter of a million dollars isn't too bad of an estimate for a simple 2D online RPG.
Ofcourse, you'll want enough cash to be prepared for things taking longer and for mis-estimations in project scope, and for extra "insurance" against accidentally getting a poor programming who you think is doing good work, who a year into the project quits and suddenly leaves you with the revelation that he was just fooling around not knowing what he was doing. So a half-million would be better. It definitely could be done with far less, but you'll have to really know what you're doing, understanding fully every development detail of the game (which you already said that you don't. You have the design down, which helps, but knowing the development itself would cut costs and reduce risks alot).
If you start to build and can't finish, you lose everything you put into it but don't get anything out of it, so a larger backup reserve is important.