I am currently reading John C. Hulls' "Risk Management and Financial Institutions" and came across the following passage related to the efficient frontier and combinations of risky and risk-free assets:
It is a short step from here to argue that the portfolio of risky investments represented by M must be the portfolio of all risky investments. Suppose a particular investment is not in the portfolio. No investors would hold it and its price would have to go down so that its expected return increased and it became part of portfolio M. In fact, we can go further than this. To ensure a balance between the supply and demand for each investment, the price of each risky investment must adjust so that the amount of that investment in portfolio M is proportional to the amount of that investment available in the economy. The investment represented by point M is therefore usually referred to as the market portfolio.
The following part confuses me:
No investor would hold it and its price would have to go down so that its expected return increased and became part of portfolio M.
This makes little sense to me at the moment. I understand that by excluding a particular investment (say $I$) in $M$, its demand will decrease as all risk-adverse investors will purchase portions of $M$ to produce a linear efficient frontier. This will decrease the price of $I$ but I am not sure why this will increase its expected value, and thereafter be inserted into the market portfolio.
I am also confused about the proportions part of the passage.
Any help would be truly appreciated.