Good question. It seems Bangladesh overall price levels are higher than in Pakistan. The main reason I would suppose is they have fewer natural resources per capita compared to Pakistan. This means they are more interlinked with the world and reliant on trade for price sensitive and logistically important commodities (which they have to use USD buffers to pay for many times). They are also very reliant on RMG exports which are consumed mostly by people who pay in USD.
This means their PPP mulitplier base level will be somewhat lower than a country like Pakistan or India (who have a more diverse basket of exports, more internalised logistics/supply chains for crucial consumption items, or in Pakistans case also trade somewhat less overall). China is an interesting one who's PPP multiplier is quite low now given its massive presence and linking into world trade (even though it is very high in local natural resources per capita).
Lets compare with China who's economic story is known very well. In 2000 it had a PPP multiplier of around 3. Now it is around 1.7 (43% drop).
Report for Selected Countries and Subjects
That is what more "international reliant" high growth does over many years. The currency "corrects" itself as more suppliers, producers and consumers get access to multi-origin goods and many markets and those markets inflation. Part of this is also explained by the massive USD reserves China holds.
In a way, PPP defines the "frontier potential" for GDP and arises because price levels in a country are different from that of the US (where USD is printed). As you trade more and more with the US (and other countries that use USD for trade), you generally get closer to a nominal and PPP per capita that match.
This is one reason. The other reason is just inherent inflation (especially supply side) within a country that makes its goods more expensive over time relative to the inflation of goods within USA (or more specifically worldwide that use USD for transaction).
We can see this latter happening within Pakistan to a larger degree than the first reason. The trade has not jumped significantly (in terms of what China has experienced for example) as % of GDP....but Pakistans PPP multiplier has dropped from a factor of 5 in 2000 to around 3.4 in 2015 (32% drop):
Report for Selected Countries and Subjects
For India the drop in the same period was from 4.4 to 3.7 (16% drop).
Bangladesh was from around 3.3 to 2.85 (14% drop)
This explains partly how Pakistan was a full double per capita in PPP in 2000 compared to Bangladesh but its now only 35% higher. India's lower fall rate in PPP multiplier also now means it has a lot higher PPP per capita than Pakistan (i.e local price levels have not increased so much compared to Pakistan).
It is not a hidden science on how this PPP multiplier is calculated. Basically IMF takes a basket of goods across a representative cross section of the country and compares their prices to the "same" basket of goods in the US using local consumption patterns and references. I say "same" because it can never be perfectly identical...but as close as possible (hence a source of error). Consumption patterns/preference also vary adding another source of error when comparing across countries (but thats an issue with GDP overall anyway)
It is also a source of error within China notably, that they dont allow such data to be gathered outside their cities. One would assume this keeps their PPP multiplier depressed (since price levels are generally lower in rural areas).
