I would like to re-examine this problem in light of Jordan’s Finalized evolution of liquidity operations proposal, which proposes to provide deep liquidity across many major exchanges. If the proposal is implemented, the scenario with PPC can well happen in the NBT/BTC market in a proportionally bigger scale.
In a major market crash, sellers could be collectively dumping several 10k btc in a short period. Daily volume on BTC-E alone had been 70-150k btc, so having 10k exiting volume on one excahnge is very possible. If Nu operates NBT/BTC bots with deep liquidity, I think we could see our reserve shrink by a significant amount quickly. Bitcoin had gone up to $30 before crashing down to a few dollars and stay there over over several years, then to $260 back to $60. Now bitcoin price is coming down from $1000 to who knows where.
To run some numbers to get a feel of the problem, suppose a total of 50k btc exits, from several exchanges, via NBT/BTC, in a market crash that went from $320/BTC to $170, with an average selling price of 270 NBT per BTC. In the end the bots lost $5M worth of reserve (because they sold 50k x $270 worth of NBT and in the end have 50k x $170 worth of BTC in hand). This could put Nu from deep liquidity to deep in debt overnight. Due to its inherent problems, bitcoin price could well go down to below $100 forever. It could be very hard to recover to previous price level.
On the other hand only operating NBT/USD bots will insulate Nu from such market crash because anyone want to sell BTC for NBT has to buy NBT from someone who has paid $1/NBT to the bots.
I do recognize the value of providing BTC/NBT liquidity by Nu bots. Maybe the best way is capping the liquidity of BTC/NBT to limit the exposure to BTC price that Nu reserve would face?