I predict a reasonable rise in NuBit interest rates will actually lead to panic selling not parking.

This would make for an interesting Best Response analysis!

Let's say I hold NuBits, and I believe that the interest rate is positively correlated to the likelihood that NuBits is about to collapse. In other words, I believe there is a function

**t(r)** that describes how many days before the NuBits collapse as a function of the interest rate

**r** (to make the math easy, assume

**r** is in %/day). Then if I have

**N** NuBits, for any interest rate

**r**, I have two options:

1) sell now and profit

**N**.

2) park for a little less than

**t(r)** days, sell right before the collapse, and make over

**N*(1+t(r)*r)**. (I'm not compounding interest; this is a lower bound)

Well look at that! Unless

**t(r)=0**, option (2)

*always* gives me a higher payoff than option (1)! So we should expect that even when the NuBits collapse is imminent, there will probably always be some risk-seeking individual(s) out there who are going to park for

*one more day* in the hopes of making away with

*one more dollar.* Of course, at some point the risk-averse people, who have been selling all along, will reach some critical mass of selling volume that the market-makers can't keep up with and the peg will break and the risk-seekers will lose all their money.

Interestingly, the custodians (aka the fat cats who make NuBits possible) will also have a choice to make:

1) Keep buying excess NuBits, trying to maintain the peg

2) Abandon their responsibility to buy excess NuBits because the collapse is imminent and the longer they wait the more they lose.

Clearly, some custodians will abandon earlier than others; the ones who hold out longer will lose more money. So notice here that in a collapse scenario, the custodians have a strong incentive to abandon the holders of NuBits!