Quote from: xjabc on April 01, 2014, 08:04:26 AM

Regarding the modified proposal, I believe it is not a repo deal any more, it's more like a fx swap deal. As a result, Rr is not real repo rate of repocoin, but a function of the rates of both coins. If so, how could we apply these numbers to adjust repocoin's PoS interests rates?

Yes, it's an interest rate swap, but implemented as a repo in order to reduce the contractual requirements. You're correct that it's non-trivial that the market clearing mechanism will return the right interest rates, as we have two interest rates to pin down from the intersection of two surfaces. But the given market clearing mechanism ought to work. To see this, suppose there were opportunities for investing bitcoin and repocoin with net (e.g. 0.02) risk free interest rates r_r and r_b. Furthermore, for simplicity, let's forget about the lost interest on the locked coins (it can be shown that everything we say ignoring the locked interest on the locked coins also holds in the limit as the time interval goes to zero). Then the seller (no matter how risk averse) will wish to participate in the swap/repo at least if R_b>=1+r_b and r_r>=R_r-1, likewise the buyer (no matter how risk averse) will wish to participate at least if R_r>=1+r_r and r_b>=R_b-1. Solving these four inequalities gives R_b=1+r_b, and R_r=1+r_r. Thus, in the presence of heterogeneity in risk aversion in the market, we should expect to see highest trade volume when R_b=1+r_b, and R_r=1+r_r i.e. when the "repo" interest rates reveal the true interest rates.