I am pretty sure it will, for two reasons. One, the inefficiencies will continue to exist. Two, the model is not a result of data mining, parameter optimization, neural network training, etc. - it will not deteriorate with time.
Apparently other market participants detect the same inefficiencies using their own methods and enter orders to take advantage of them.
It is not yet a complete trading system that can be backtested. It is a method of indicating trading opportunities - an edge.
Why did you start this thread about a trading system that's not complete...you asking for help in completely it or you're trying to find out if your design is going in the right direction or you have questions about what your exit strategy should be considering you mentioned it hasn't been designed ??? Also, after you complete your trading system...what trading instruments do you plan on testing it on or trading ? By the way, I'm a little confused by your use of word its "an edge" when you also say the trading system isn't completed.
Please ignore the following if you're a veteran or/and an expert: One of the first things I did, shorty after getting into trading was discover momentum indicators. I subsequently wasted a lot of time and effort developing a system which was primarily based on them. You can fiddle with their parameters, combine them or even create an improved version, like you may have done, but they do not take S/R/volume into consideration, so (IMO) should never be thought of as anything more than secondary - i.e. a filter.
Let me clarify for the brethren who may be getting a bit confused. Actually, "the edge" ain't some guitar playin' multi-millionaire. The edge is following the trend. Reversals are okay, but not for beginners or the faint of heart/pocketbook.