The $12B figure seems to be a bit too less though.
I recall them finishing a number of projects worth more than that in the power sector alone.
For now, the projects in its first phase have failed to usher in the level of prosperity that was promised to the people.
www.dawn.com
Early harvest projects worth $19bn have been completed, and CPEC is now in its second phase, in which most of the projects will be based on a public-private partnership model.
There's also the critical M5 highway that was worth nearly $3B. There was also the rather expensive Orange Line metro in Lahore. An unknown amount has been spent on Gwadar as well, quite a bit of which is outside the CPEC framework and is seeing direct Chinese involvement.
And then there's some which have barely begun but will eventually be completed, like that gigantic dam. There's also the nearly $10B ML-1 upgrade program along with the engines and coaches. Only these two combined would be about $25B+overruns and are pretty much guaranteed to go through.
I take my ~12 billion from what I have seen in the general debt statistics at World Bank (for loans) and UNCTAD for FDI...but I might have to check later, its off top of my head.
My understanding is the larger amounts we see from time to time (for specific projects or combined project classes) are valuations (X+Y+Z), rather than precise investment/loan injection amount (Y).
The land + other capital resources used itself has some X value already, you develop it (add buildings, logistics + labour + capital goods + capacities etc) with further Y amount of money and its worth (you say) X+Y+"Z" in the end in local currency (PKR)....Z being added utility+functionality from predicted long term output etc. The "total market cap" of that project lets say.
Then you convert it (X+Y+Z) verbatim to USD reference and imply that conversion holds the same average world quality (when the scale, intensity and makeup of exchange rate of USD-PKR is far different to average world norm to begin with).
But if it actually is that valuation needs Pakistan to integrate far more with the USD/World references+standards (and the growth of high frequency markers to back this) by larger trade, investment metrics etc... and demonstrate commensurate growth in its total market cap too so we see what the actual hard equity and asset numbers are that materialised...from say 2015 onwards period (when CPEC started).
Simply it might only be known on that front in 2025, 2030 etc.
But all that stuff needs far far more to be done than just CPEC (or at least what I have seen of it)....especially to even get a sound+better+relevant valuation (that can be leveraged internationally) in the first place rather than just a number to print in media/brochure.
Because that hard asset + equity number (gross capital formation) so far is way underperforming compared to the CPEC brochures and whatever valuation claims floating around.
The real problem to address is why are Pakistanis not saving enough money. You ask that and you are back to cabal issue and practical realised trust deficit of aam-admi. That is larger thing to handle, there is no good economic study of it, it involves human psyche, hidden (but real) inflation and various shakedowns done by those in power on the less powerful (and thus the populace's perceived net benefit from investing honestly into the system....with whatever the hidden inflation leaves behind to begin with).