Quote from: ZeroNominal on April 01, 2014, 03:16:12 PM

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

the seller would participate in the swap/repo if R_b>=1+r_b and r_r<=R_r-1, only if his gain on one leg is greater than the loss on the other leg. No need to gain on both legs.

In fact, suppose the real interests rates of Repocoin and Bitcoin were r_r and r_b. Currently, Alice holds B bitcoins and Bob holds B*P repocoins (P is the current price of repocoin in one unit bitcoin). Alice and Bob make a one week swap deal, and will get B * r_b bitcoins and B * P * r_r repocoins after the deal respectively. In a equilibrium market, the new price of repocoin one week later will be P * (r_r / r_b).

However, in my opinion, the real interests rates of both coins can't be independently revealed by the market. Even a well functional market can only reveal the relation between r_r and r_b, more specifically, r_r / r_b. When I am putting or accepting an order with 3 quantities: B, r_r and r_b, another order with 0.5B, 2r_r and 2r_b is almost same to me. The only difference is the latter has a leverage of 2 times.