Originally Posted by Midnight_Sun
The Saka (Scythians) who founded Bactria, and before them, bronze age Andronovo cultures who displaced the older Elamitic civilizations that were found in Mundigak near Kandahar, and elsewhere in Afghanistan, Pakistan (Harappan), and Iran (Shahre Sukhta). Kushans came way later, after the Greek settlements. They were made of confederation of Yue-Chi (most likely Tokharian) tribes, Tokharoi, Sakaroloi, and Assiani, and others that were pushed out of the Tarim Basin by the Greater Hun or Xong Nu.
The Bactrians and Soghdians were contemporaries with one another. Neither the Bactrian nor the Soghdian language belongs to the Scytho-Sarmatian linguistic continuum despite being NE Iranic.
Andronovo (~2000 B.C.) was an offshoot of the Sintashta culture from around the Ural Mountains, which itself was an eastern extension of Yamnaya. These people had settlements and there's even evidence of warfare between them in some regions (multiple layers of burnt and rebuilt fortifications in the archaeological strata). As many people know, these people were semi-nomadic pastoralists who rode on chariots and kept horses.
The important step you missed was the involvement of the Bactria-Margiana Archaeological Complex (BMAC). These people were resident to South-Central Asia for thousands of years and were agriculturalists who also originated further west (Anatolia/Caucasus/NW Iran).
Archaeological evidence of interactions between the people of Andronovo and BMAC can be found in the peripheries of both domains. After some time, however, Andronovo material goods begin to make their appearance deep within BMAC settlements. Conspicuously, the horse does not accompany them. I have interpreted this as evidence the Andronovo newcomers had assimilated themselves into BMAC society to an extent. Central Asian Archaeologists seem to hold a similar view (e.g. Kuz'mina paraphrased; "Andronovo individuals exerted elite dominance over BMAC natives but eventually mixed after their language and ideology overrode the natives").
Little else happened in our region until the Migration Age, which you have partially described; tribes along the Eurasian steppe were displaced westwards beginning with the Xiongnu, an East Eurasian nomad confederacy who warred with the Chinese. The Yuezhi (another confederation with a Tocharian component) were but one of their victims. The rest is (well-attested) history.
In addition to all this, I should point out that South-Central Asia (i.e. Pakistan and Afghanistan) had, since the BMAC era, been much greater recipients to the Eurasian steppe than the Iranian plateau. The few Scythian raids in Iran were focused in the northwest of the country and were sandwiched in-between the expansion of native empires at that time (Assyrian, Median). Based on this, I would not be very surprised if Pashtuns and the Pashto language do show associations with specific developments in the region (for instance, Tocharian loanwords?).