You under estimate many in the distributed computing community such as Folding at Home, World Community Grid, GPUGRID, M@H, and other BOINC projects. Having my grass roots in those myself and many other's that do such projects at their own expense and enjoy hardware in general. There will away's be those of us backing computational power for projects we believe in regardless of monetary gain. In fact I know many who had plenty of empty pcie slots they have now occupied for mining on already very high end system (EVGA SR-2's etc) to off set there cost and expand even further for F@H, WCG etc. Mining is in many cases helping fund individuals efforts to cure disease's.
I didn't say there wouldn't be people mining, I just said there's no reason for it. If it drops below my threshold I drop off. If the network really looked like it was falling apart I probably would turn on 300mhash/s just to support it though.
As there is NO way to get Euros to an exchange site fast enough befor the price potentially bounces and goes back to profitable, you either need to keep a few day's worth of electricity money on an exchange site and monitor this closely (this costs time, and time is money I loose and have to deduct from these earnings!) or just live with occassional 20 cents of loss on a bad day + shrug it off. If you mine at a constant loss (far below profitability) however, buying is really cheaper and you should stop mining.
Just calculate how much a mined bitcoin costs you (all you need is your MH/s speed, current difficulty + the amount of Watts you use) and decide then if mining is worth it.
Just keep a months worth of electricity costs in EUR at the exchange and setup a buy for a months worth of BTC at your breakeven point. that way you're not even watching the markets.