Indeed, and again, I never suggested that it could indefinitely prevent this from happening, but that it simply postpones the "crash" in case the cause is actually a whale/s colluding as per an agenda. But the same can work in reverse surely, if the backing currency crashes, then the value of the asset it is backing is also in jeopardy.
It can, but there is no dependency between the two, which is critical. A dependency can cause the spiral of death as devaluation causes further devaluation and market panic.
Right, you don't want dependency, but even without out any, the effect can still occur.
Right so that brings me to the crux of the issue that needs to be addressed. You are right, these informed traders were not accounted for in the test, I considered it and maybe wrongly determined that their effect would not be great enough to deplete the buffers.
However, as its been raised here, I'll revisit that thought process and look to include their effect into the test and rerun it. Should the result then be that the price is still stable and the buffers are healthy, perhaps we can begin to look at this model in a more serious manner?
You can actually get an estimate for the proportion of informed traders present in the market from your data - it can be approximated by orderflow, so take a sliding window of trades and calculate the buy volume and sell volume, then the VPIN = (buy-sell)/(buy+sell). This goes between -1 and +1, so just abs() it to give you your normalised proportion.
Ok great thanks for that. I might call upon you as I'm implementing the modifications if thats OK with you , as you've looked into Informed Traders and other such phenomenon in more detail than I have.
I largely considered the activity of buying low on the internal exchange and selling high on an external one as irrational, as the barrier to entry cost to trade on the internal exchange is minimal, I (perhaps incorrectly) assumed that no one would buy at the higher external price. (Are there even any valid reasons why they would?)