An issue which has risen is the transition from Stage 1 to Stage 2.
The issue with the current model is as follows:
During stage 1, the only way to buy XRD is thought the AMs for 1:1 XRI:XRD trade. However, while the trade is 1:1 ratio, the reserves (Pmin) are actually much lower. This gap between the buy price and the value of Pmin at the end of stage 1 results in a large risk. Investors are asked to buy at a value much higher than what the reserves can ensure them. There is a possibility that this may lead to a panic sell in which the price may drop to Pmin, the moment we transition to stage 2 and people can freely trade on the DEX.
The suggested solution
My suggestion is to implement an adaptive XRI:XRD ratio during stage 1. For instance, this can be set to be max(X, 1.1*Pmin) where X is some minimal value as Pmin starts at 0. This ensures (1) higher incentive to buy early as you may gain more from an increase of value, and (2) when we transition to stage 2, the gap between the buy value and the floor, e.g., the potential risk, is small.
How it would work
User buys FIAT tokens from the AM and converts them to XRI.
The user converts XRI to XRD based on the price of 1.1 * Pmin. The surplus is not distributed during stage 1.
As the buy price is higher than Pmin, the reserves relative to XRD grow and thus Pmin rises.
When we transition to stage 2 the maximal value any user has bought is at most 1.1 * Pmin and in case of a price drop that user is ensured a loss of no more than 10%.
The incentive to buy during stage 1 is based on an increase in value and not redistribution. Stage 2 and 3 will enjoy redistribution effects when the DEX is active.