A few thoughts...
This is a pretty slippery discussion. To what is your organization considering "Investment"? Just your salary? Salary and benefits? Salary, benefits, and all of the space/utilities required to allow a purchasing department to function?
This is a tricky question. When you get this total idea of what the "Investment" is, then you can start to calculate your ROI.
Imagine that your purchasing headcount, with the 'worst case scenario' per head is 150K a year.
Your first and most obvious return is any cost savings you may encounter. That may be as simple as finding a cheaper (but same quality) box of pens. This may also include rolling out vendor managed inventory, supplier consolidation, consigned and/or committed inventory (where applicable), or any purchasing/stores methodology.
You may also want to look at your role as a Risk Manager, in which you may have your contractor/service providers sign contracts which push the burden of the risk upon them, and not on you. Think about this; imagine that you do (all numbers are arbitrary) 25 projects a year in which there is a 1% risk of a worker fatality. We know that the total cost of a worker fatality, through payouts, investigations, regulations, etc can approach (and many times exceed) $1,000,000. Using "The Formula", of # of opportunities (25) multiplied by the likelihood of the incident (1%), multiplied by the cost of an accident (1,000,000), you get 250,000 worth of risk avoided by signing that contract.
(For those of you who recognize this from pop-culture, it's something that is discussed roughly in "fight club" in the car accident investigation.)
And finally, I don't mean to sound negative, but...
Your organization may be backwards, and doesn't understand purchasing's role in the modern world. Look at the top tier entities in the world; robust purchasing departments that reap billions upon billions of dollars of savings that immediately and directly the bottom. But you could take these numbrs, and show it to your organization, and they may say "meh" because that's management's choice. It's not a smart one, but a decision that can be made, nonetheless.
There have been more than once in my (short) career that I have demonstrated, both mathematically and theoretically the error of management's ways, but they choose the "good ol' boy" or "easy" method.