I just had an idea about reconciling Bitcoin with the Regression Theorem. Imagine invisible envelopes (of various sizes) were invented which allowed you to discreetly ship gold or USD between any two points in the world. You’d “pack” $100 in a “$100 size envelope” or 10oz of gold in a “10oz envelope”, enter the destination and off it goes. I would imagine such envelopes would be subjectively valued some non-zero $ amount in the market – say, 1% of the gold or USD value they are able to ship. So a “$1000 envelope” would be worth $10. If I shipped you an empty “$1000 envelope”, wouldn’t have I actually made a $10 transaction with you – using the envelope itself as money? Also, due to their obvious non-monetary utility, wouldn’t these envelopes also satisfy the Regression Theorem? Since I could use Bitcoin to send you gold or USD today (I trade USD => Bitcoin on my end, send you Bitcoin, and you trade Bitcoin => USD on your end), we could view Bitcoin as an envelope as described above. After a while, why not just exchange empty envelopes whose value has been derived from their previous utility as transaction vessels for (old) money?
This hypothetical example wasn’t meant to exactly replicate the Bitcoin situation as it stands today and the envelope I described does not have to be worth 1% of the content it is able to transport, or even an amount proportional to the amount it is able to transport. I just picked that number as an example. The envelope will be valued at whatever people would be ready to pay for it, and that would be determined by how useful it is to them (demand) and how many envelopes are available for sale (supply).
The above shows how a method/system/pattern for transacting money (or anything else already widely valued by others) could become widely valued itself – hence become money as well. It would satisfy the Regression Theorem due to it’s previous utility as a private/secure conveyance for money.
Today anyone (even someone who doesn’t believe a Bitcoin is worth more than $0) could use Bitcoin to privately “send” fiat currency to anyone in the world without exposing themselves to fluctuations in the $/Bitcoin price for too long. Party A exchanges $'s for Bitcoins at the current market price then immediately sends Bitcoins to Party B across the world upon which Party B immediately exchanges those Bitcoins into $ at almost exactly the same market price. Businesses could pop up (BitPay?) offering this transaction in a package deal, even guaranteeing a fixed $/Bitcoin exchange rate at both ends for a fee. The more people do transactions like these, the faster the transactions would become (lowering the risk/exposure to the $/Bitcoin market price) and the more efficient the $/Bitcoin market would become, leading to a more stable $/Bitcoin price, which in turn lowers the risk/exposure for the people unsure about the value of Bitcoin in $ terms.
As long as transactions like the above are demanded (even by people who would never hold Bitcoins for more than a second as they believe Bitcoins to be worthless) there would exist a non-zero $/Bitcoin price, praxeologically so.
Thoughts?