Your math isn't faulty, op. At this point, cloud mining simply isn't profitable. The only upsides to using a cloud service like cex.io are that you don't have to deal with maintenance issues (i.e. a PSU exploding), noise from cooling systems, and the large amount of heat produced by all the equipment. Don't get me wrong, those are decent selling points. But if you have the option of buying a physical rig and housing it locally you stand a much better chance to eek out some sort of profit. However, for somebody who doesn't have the kind of money they can spend on a th/s miner, cloud mining might just be the way to go. If you happen to be short on money at the time, you can buy a single gh/s (or even smaller amounts) if you want that will start mining for you as soon as you've bought it. There is no downtime while waiting for your rig to arrive from Sweden or China or wherever it was made. Some would see this as a benefit. And, with cex.io in particular, you can choose to reinvest any amount of your mined earnings at any given time, giving you more gh/s and faster earnings. When I started with cex, I originally bought around 250 mh/s. After a few days, I started reinvesting my earnings to buy more gh/s. And within about 3-4 weeks, I was up over 2 gh/s. You need to do your own cost/benefit analysis to figure out what's right for you. Don't just trust the profit calculators and what other are saying on some random forum. If you find that you have the resources to buy a physical rig and have the ability to host it locally, by all means. However, if you do a careful analysis of all the factors involved, you may just find that cloud mining is the way to go for you. If you decide you want to pursue cloud mining at all, please feel free to use my ref link when you sign up. I could use the extra gh/s.
Edit: Auuuuggggghh... apparently bitcointalk filters cex.io ref links out of their posts. So I guess I'll put it up another way.
https:// cex .io/ r/ 0/ unic0rnp0ny/ 0/ Just take out the spaces