[refurb]

As far as I know, US laws require that the `x` number of trials demonstrate the effectiveness of the drug and failured dont need to be reported.

That is not correct. When you submit an NDA to the FDA, you need to include every piece of clinical trial data you've obtained so far. This includes trials that you've either ended early or failed to show any effect at all.

The FDA doesn't look kindly upon a company that hides any clinical trial data.

How many years exactly does one need to subvert the market so that a satisfactory level of innovation is maintained ?

We have some data on that, but we can't predict the future. I can't find the stat right now, but only about 1 in 3 drugs approved by the FDA actually produce an overall positive return.