Amongst other things, I employ modellers in banking / financial services industry. Some random thoughts:
1) No-one will give you a modelling job because you know excel quite well.
2) When considering your career, distinguish between modelling and analysis. Successful analysts do things like automate things for fun. Successful modellers read new developments in probability theory for fun.
3) Without direct domain experience, you need **** good, proven, well above average modelling skills to get a foot in the door in a modelling role that's not paying peanuts
4) Modellers in banking do a lot less modelling than you think (imagine all the regulatory stuff they have to deal with, before you even start on the documentation, meetings etc etc etc)
5) Modellers in firms that contract to other firms that would like to have more modellers than they can afford full time tend to do more actual modelling
6) There are other routes in. In banking, models need to be validated so there are model validation teams. You can, sometimes, get into validation teams more easily with less domain expertise, then move into 'proper' modelling once you've proved yourself.
7) Think creatively about how you demonstrate your modelling skills. If someone came to me, said they hadn't really had the chance to demonstrate the range of their excel skills in their current job, but had 8000 posts on MrExcel and had recently been recognised as an MVP, I'd pay quite a lot of attention. Ditto modelling. There are 'model-off's out there, as you have identified. You could, without too much total effort, find other 'non-traditional' ways to prove your interest and skills.
8) There are a very large number of people who call themselves 'modellers' who work in marketing firms. (Think propensity to respond to marketing offers etc). Some are very good. But, as a general rule, the less regulated the firm using the model, the easier it's going to be to get into the modelling department.
…which may or may not help.
