Thank you for your reply. I mostly agree with what you said. Unfortunately I am not yet comfortable enough with the technical details of fedimint to know exactly how the federation gateway(s) with LN work. I remember reading somewhere in the docs that there can be multiple gateways offered to a federation, some of which may be run by non-federation members, so perhaps this would somewhat mitigate the concern you bring up regading liquidity, but maybe not.

That is confusing to me. Is the minted token meant to be interoperable with lightning...What am I missing?

Thanks for asking for clarification. Maybe it would be easier to think about the following scenario:

1. a mint starts with the as-currently-being-coded full reserve policy where all sats are redeemable on demand
2. somehow the guardians/federation come across one or more "investment opportunities."
3. the opportunites would probably be in the form of debt (e.g. lending fixed amount of sats for a fixed term length and known return). Let us assume the opportunity is for 30 days.
4. the mint somehow notifies its members/depositors of the opportunity, and lets anybody redeem their on-demand e-cash-sats for some new 30-day-locked-e-cash-sats.
5. These new 30-day-locked-ecash sats are basically like savings bonds.
6. Once enough savings bonds have been "sold" to the members, then the mint federation/guardians go ahead and fund the deal (via regardular lightning or on-chain sats).
7. When the 30 days is over, the deal either returned adequatlely or it did not. If it did not, the mint buys back the savings bond from the bond hodlers at a discount. If it did perform, the mint buys back the savings bond at par value and furnishes the bond hodlers with their portion of the interest, etc.

In this regard, the guardians/federation are acting in a capacity more similar to a fund manager -- filtering deals/opportunities.

As a side note, the savings bonds themselves would, presumably, be e-cash (e.g. transferrable/tradeable) as well, similar to treasury bonds. Savings bonds issued by the same mint with the same par value and same maturity would be fungible with one another.

guarantee this will fail then when the bitcoin exchange rate rises and those who understand the risk of this being fractional reserve withdraw.

Everyone would know that this is how this particular mint operates, so the bitcoin exchange risk is information any would-be participants can factor in. Of course they are also factoring whether the mint guardians/federation are going to be able to source/underwrite deals which payoff. On the other hand, if no deals are sourced, then the status quo would be the usual full-reserve-minting.

What I tried to describe here would somewhat insulate member/depositors who do not want to buy any savings bonds (subject to trust in the guardians/federation, but that trust/risk is there regardless). They just use the mint like normal.

Anyway, there are obviously a lot of details that would need to be worked out to even make something like this work. What I am simply trying to show here is that it might still be possible for the mint to stay solvent, in expectation, by matching the term structures of its assets with those of its liabilties.

It is subtle, and I have probably failed to fully explain it, but it is not quite the same as fractional reserve.