Thread by @GeorgeSelgin

Soooo...I have just heard from someone in Senator Lummis's office who is familiar with the details of the proposed legislation. It turns out that it has practically nothing to do with "bank reserves" or the $3 trillion in Fed assets backing them. Therefore...another thread!

https://twitter.com/GeorgeSelgin/status/1817820745340207467

The plan is in fact the much more modest one of having the gov't acquire 1 million in BTC, or about $64 billion worth. And the Fed would not actually be acquiring any (though it would be involved in the process). The Treasury alone would acquire them.

How? It would take advantage of the gold in Ft. Knox, which is presently officially valued at about 1/6oth of its actual market value, which is presently about $353 billion. Step one is to have the Treasury revalue that gold at its true market worth.

To realize on that tidy gain, the Treasury would give the Fed fresh new gold certificates equal to it. (The Fed originally swapped gold certificates for the Fed's actual gold when it confiscated the latter in 1934, _before_ officially devaluing the dollar.)

The Fed would in turn credit the Treasury General (Fed) account by the value of the extra certificate. That is, it would have an extra $347 billion or so in extra assets backing a like increase in its liabilties, with all the latter consisting of a BIG fresh TGA deposit.

The Treasury could then buy $64 billion of BTC using just a fraction of its windfall. Easy!

But...there is always a but.

First, every amount the Treasury spends out of its TGA balance adds as much to bank reserves. So this plan will add at least $64 billion to those reserves, and could add as much as $347 billion, or more than a 10% increase.

And the Fed must pay interest on those reserves, though it will not have any corresponding interest-earning assets to pay it with, for gold certificates are not interest bearing.

In the long-run, as I noted earlier, the Fed must earn enough revenue on its portfolio to cover its operating expenses, including interest on reserves. Otherwise we run into the Willem Buiter problem mentioned earlier.

In this case the problem is much less serious than in the scenario I considered earlier, in which the Fed backed bank reserves _entirely_ with non-interest-earning Bitcoin! Still, it will reduce the Fed's long-run capacity to remain self-financing.

And that's something to be concerned about, since the Fed is already resorting the accounting legerdemain of booking "deferred assets" to deal with its revenue shortfalls, and that trick only makes sense when as long as it's able to make up for the shortfalls someday. (End.)