Content-Type: text/plain;
	charset=utf-8
Mime-Version: 1.0 (1.0)
From: Eric Voskuil <eric@voskuil.org>
In-Reply-To: <A64C3DCB-10CE-45EA-9A1B-7E3D13DF90EA@gmail.com>
Date: Wed, 3 Jul 2019 17:30:55 -0500
Content-Transfer-Encoding: quoted-printable
Message-Id: <6B9A04E2-8EEE-40A0-8B39-64AA0F478CAB@voskuil.org>
References: <0DBC0DEA-C999-4AEE-B2E1-D5337ECD9405@gmail.com>
	<7A10C0F5-E206-43C1-853F-64AE04F57711@voskuil.org>
	<EB743E28-8486-4CD5-B9AE-09B5693B27FA@gmail.com>
	<708D14A9-D25E-4DD2-8B6E-39A1194A7A00@voskuil.org>
	<1A808C88-63FD-4F45-8C95-2B8B4D99EDF5@gmail.com>
	<83705370-79FC-4006-BA04-4782AD5BE70B@voskuil.org>
	<BF027CD0-FE29-4DD1-AB96-DE92B597AD18@gmail.com>
	<3F46CDD5-DA80-49C8-A51F-8066680EF347@voskuil.org>
	<A4A6099F-F115-4CBF-B7D5-F16581476126@gmail.com>
	<063D7C06-F5D8-425B-80CE-CAE03A1AAD0C@voskuil.org>
	<0AA10217-E1CC-46D1-9B43-038CEEF942CD@gmail.com>
	<E72C4A8E-F850-400B-B19B-7D06B8A169EC@voskuil.org>
	<A64C3DCB-10CE-45EA-9A1B-7E3D13DF90EA@gmail.com>
To: Tamas Blummer <tamas.blummer@gmail.com>
Cc: Bitcoin Protocol Discussion <bitcoin-dev@lists.linuxfoundation.org>
Subject: Re: [bitcoin-dev] Generalized covenants with taproot enable
	riskless or risky lending,
	prevent credit inflation through fractional reserve
Precedence: list


> On Jul 2, 2019, at 00:08, Tamas Blummer <tamas.blummer@gmail.com> wrote:
>=20
>=20
>> On Jul 1, 2019, at 20:52, Eric Voskuil <eric@voskuil.org> wrote:
>>=20
>> I said that I would make no further comment given the belief that no new i=
deas were surfacing. However, after giving it some more thought on my own, I=
 believe I have found the one case in which a person could value such encumb=
ered coins.
>>=20
>> In the case of tracking an asset that becomes worthless at a specific tim=
e, one could value a record of ownership, and the ability to trade ownership=
 of the asset during the period. Consider colored coin type tracking of a th=
eater ticket for a specific show, where the ticket is worthless by the end o=
f the show.
>=20
>=20
> In other words you now see the utility of a register offered by UTXOs that=
 are only temporary availability to current owner. If there is a utility the=
re is also a value in it for them.

In other words I discovered a potentially-valid use case for you. The concer=
n I expressed was that you had not presented one.

> I am glad we are on the same side on this utility

My goal is never to discourage, but understanding of provable behavior and u=
tility. Our space is replete with unsupportable conjecture and hyperbole. Th=
ere are no sides, just discovery of truth.

> and thanks to you and ZmnSCPxj we now have two additional uses cases for U=
TXOs that are only temporarily accessible to their current owner.

Actually you have a single potentially-valid use case, the one I have presen=
ted. The others I have shown to be invalid (apart from scamming) and no addi=
tional information to demonstrate errors in my conclusions have been offered=
.

I=E2=80=99ve noticed that in subsequent posts you continue to imply that the=
re is economic value in such tracking of any asset, and of course here imply=
 the validity of your other use case, monetary lending. This, as I have show=
n, is not the case. Tracking of an asset of value beyond the net compound in=
terest cost of dust is more cheaply accomplished by burning than by renting,=
 and as I have shown, it is not accurate to claim that the encumbered coin c=
an be used as money (or to track any asset of perpetual value). When the coi=
n expires the money/asset holder becomes a bag holder, invalidating any init=
ial value apart from scamming.

In the valid use case that I have demonstrated (tracking of expiring assets)=
, the marketable value of the rented coin is not the market price of that co=
in, but the price paid for it. So for example, 1 coin rented at 10% APR for o=
ne year is worth .1 coin. And when a renter resells this tracking coin it is=
 worth the fraction of this amount for the time remaining. The coin itself (=
i.e. its face value) cannot be used by the renter to purchase anything.

As such this is truly not a loan in the financial or economic sense. Given a=
n actual loan the borrower can use the full value of the amount borrowed to p=
urchase goods that can be used in production. Subsequent generation of produ=
cts and thereby revenue is the source of yield on a loan (economically equiv=
alent to dividend on an equity contract). This allows the borrower to repay t=
he loan with interest. Without *any* usable capital over the term of the ren=
tal, there is no investment possible and the time value of the rented coin c=
annot be realized by the renter.

So the one potentially-valid scenario, a fixed-term tracking rental, is enti=
rely an *expense*, not a loan. A financial loan incurs an interest expense, b=
ut also implies the value of the amount loaned is fully usable (i.e., consum=
ed or traded) during the term (the reason to pay interest). That is money ov=
er time, yielding the time value of money. In this case the value of the loa=
n at any time to the renter is simply the amortized interest remaining. This=
 implies that no income can be generated from the rental =E2=80=9Cprinciple=E2=
=80=9D by the renter. A price is paid for the rental and that value of the r=
ental is fully exhausted by the end of the term, with no other benefit than t=
he tracking that was purchased.

The person renting the fixed-term tracking coin (i.e. =E2=80=9Cowner=E2=80=9D=
) can earn income by selling dust+1 outputs at the cost of capital, limited t=
o a maximum term dictated by the cost of capital and the dust limit (as show=
n previously). Economically speaking, all business returns gravitate toward t=
he cost of capital, including lending, and this is no different. But it cann=
ot be said that the owner is a financial lender. The owner is simply selling=
 non-depreciating (from his perspective) fixed-term tracking space.

The owner can of course trade rights to the controlling output. The rental c=
ontract has been prepaid (by your design, in order to shift counterparty ris=
k). As such the traded contract has no yield and therefore contracting for i=
ts sale is a currency future, not an interest rate future as would be implie=
d by a debt market. Yet FX speculation already exists for Bitcoin, requiring=
 no covenant or rental market. This would seem to undermine any secondary ma=
rket for these more complex and limited currency futures.

Finally, valuation is based on the assumption of a non-zero dust+1, which BT=
C enforces as a 0 satoshi dust limit (i.e. 0 sats is considered dust and is n=
ot valid). Anything above this is policy-enforced only. As such a miner can u=
ndercut the cost of tracking an individual asset down to 1 sat. Given that t=
here is no financial incentive to a higher dust limit for a miner, but a pos=
itive financial incentive to undercut the rental price for the same return, t=
his is economically rational and therefore must be assumed.

One might argue that a lower dust policy would hurt BTC and therefore its mi=
ners collectively, creating an offsetting negative financial pressure. Howev=
er given that the apparent cost is socialized in relation to individual bene=
fit, this is not an economically rational conclusion. Furthermore, as the tr=
acking outputs become unspendable due to the nature of the covenant, there i=
s no actual dust accumulation (implementation dependent).

As such the return on any fixed-term tracking output, given a 10% APR, would=
 become as low as .1 sat per year, assuming such a market could continue to f=
unction at that level. But it is also the case that a 1 sat output can be bu=
rned directly by the tracker and used indefinitely. This would presumably un=
dermine any robust market for fixed-term tracking rental.

Best,
Eric

> Since ZmnSCPxj also raised the question if covenants are needed at all, le=
t me continue my thoughts on this in reply to his mail.

> Tamas Blummer
>=20
>>=20
>>=20
>>> On Jun 30, 2019, at 13:26, Tamas Blummer <tamas.blummer@gmail.com> wrote=
:
>>>=20
>>> My argument does not need the comparison with ICOs.
>>>=20
>>> They were just an example that people pay for the utility of register ev=
en though others think the tokens they keep track of are worthless.
>>>=20
>>> Tamas Blummer
>>>=20
>>>=20
>>>> On Jun 30, 2019, at 22:13, Eric Voskuil <eric@voskuil.org> wrote:
>>>>=20
>>>> ICO tokens can be traded (indefinitely) for other things of value, so t=
he comparison isn=E2=80=99t valid. I think we=E2=80=99ve both made our point=
s clearly, so I=E2=80=99ll leave it at that.
>>>>=20
>>>> Best,
>>>> Eric
>>>>=20
>>>>> On Jun 30, 2019, at 12:55, Tamas Blummer <tamas.blummer@gmail.com> wro=
te:
>>>>>=20
>>>>>=20
>>>>>> On Jun 30, 2019, at 20:54, Eric Voskuil <eric@voskuil.org> wrote:
>>>>>>=20
>>>>>> Could you please explain the meaning and utility of =E2=80=9Cunforgea=
ble register=E2=80=9D as it pertains to such encumbered coins?
>>>>>=20
>>>>> I guess we agree that some way of keeping track of ownership is prereq=
uisite for something to aquire value.
>>>>> We likely also agree that the security of that ownership register has g=
reat influence to the value.
>>>>>=20
>>>>> The question remains if a register as utility in itself gives value to=
 the thing needed to use that register.
>>>>> I think it does, if people are interested in what it keeps track of, f=
or whatever reason, even for reasons you find bogus.
>>>>>=20
>>>>> It was not intentional, but I think I just explained why Ethereum aqui=
red higher market value by being register of ICO tokens.
>>>>>=20
>>>>> Now back to the coins encumbered with the debt covenant:
>>>>> Transactions moving them constitute a register of covered debt and you=
 need them to update that register.
>>>>> Should some people find such a register useful then those coins needed=
 to update this register will aquire value.
>>>>> Does not matter if you think the concept of covered debt is just as bo=
gus as ICOs.
>>>>>=20
>>>>> Here some good news: If they aquire value then they offer a way to gen=
erate income for hodler by temporarily giving up control.
>>>>>=20
>>>>> Tamas Blummer
>>>>>=20
>>>>>>=20
>>>>>> The meaning in terms of Bitcoin is clear - the =E2=80=9Cowner=E2=80=9D=
 of outputs that represent value (i.e. in the ability to trade them for some=
thing else) is recorded publicly and, given Bitcoin security assumptions, ca=
nnot be faked. What is not clear is the utility of a record of outputs that c=
annot be traded for something else. You seem to imply that a record is valua=
ble simply because it=E2=80=99s a record.
>>>>>>=20
>>>>>> e
>>>>>>=20
>>>>>>> On Jun 30, 2019, at 11:35, Tamas Blummer <tamas.blummer@gmail.com> w=
rote:
>>>>>>>=20
>>>>>>>=20
>>>>>>>> On Jun 30, 2019, at 19:41, Eric Voskuil <eric@voskuil.org> wrote:
>>>>>>>>=20
>>>>>>>>=20
>>>>>>>>> On Jun 30, 2019, at 03:56, Tamas Blummer <tamas.blummer@gmail.com>=
 wrote:
>>>>>>>>>=20
>>>>>>>>> Hi Eric,
>>>>>>>>>=20
>>>>>>>>>> On Jun 29, 2019, at 23:21, Eric Voskuil <eric@voskuil.org> wrote:=

>>>>>>>>>>=20
>>>>>>>>>> What loan? Alice has paid Bob for something of no possible utilit=
y to her, or anyone else.
>>>>>>>>>=20
>>>>>>>>> Coins encumbered with the described covenant represent temporary c=
ontrol of a scarce resource.
>>>>>>>>>=20
>>>>>>>>> Can this obtain value? That depends on the availability of final c=
ontrol and ability to deal with temporary control.
>>>>>>>>=20
>>>>>>>> For something to become property (and therefore have marketable val=
ue) requires that it be both scarce and useful. Bitcoin is useful only to th=
e extent that it can be traded for something else that is useful. Above you a=
re only dealing with scarcity, ignoring utility.
>>>>>>>=20
>>>>>>> There is a deeper utility of Bitcoin than it can be traded for somet=
hing else. That utility is to use its unforgeable register.
>>>>>>> We have only one kind of units in this register and by having covena=
nts we would create other kinds that are while encumbered not fungible with t=
he common ones.
>>>>>>>=20
>>>>>>> Units are certainly less desirable if encumbered with a debt covenan=
t. You say no one would assign them any value.
>>>>>>>=20
>>>>>>> I am not that sure as they still offer the utility of using the unfo=
rgeable register, in this case a register of debt covered by reserves.
>>>>>>> You also doubt forcing debt to be covered by reserves is a good idea=
, I got that, but suppose we do not discuss this here.
>>>>>>> If there are people who think it is a good idea, then they would fin=
d having an unforgeable register of it useful and therefore units needed to m=
aintain that register valuable to some extent.
>>>>>>>=20
>>>>>>>>=20
>>>>>>>>> I think you do not show the neccesary respect of the market.
>>>>>>>>=20
>>>>>>>> I=E2=80=99m not sure what is meant here by respect, or how much of i=
t is necessary. I am merely explaining the market.
>>>>>>>=20
>>>>>>> You are not explaining an existing market but claim that market that=
 is not yet there will follow your arguments.
>>>>>>>=20
>>>>>>>>> Your rant reminds me of renowed economists who still argue final c=
ontrol Bitcoin can not have value, you do the same proclaiming that temporar=
y control of Bitcoin can not have value.
>>>>>>>>=20
>>>>>>>> It seems to me you have reversed the meaning of temporary and final=
. Bitcoin is useful because of the presumption that there is no finality of c=
ontrol. One presumes an ability to trade control of it for something else. T=
his is temporary control. Final control would be the case in which, at some p=
oint, it can no longer be traded, making it worthless at that point. If this=
 is known to be the case it implies that it it worthless at all prior points=
 as well.
>>>>>>>>=20
>>>>>>>> These are distinct scenarios. The fact that temporary (in my usage)=
 control implies the possibility of value does not imply that finality of co=
ntrol does as well. The fact that (renowned or otherwise) people have made e=
rrors does not imply that I am making an error. These are both non-sequiturs=
.
>>>>>>>>=20
>>>>>>>>> I say, that temporary control does not have value until means deal=
ing with it are offered, and that is I work on. Thereafter might obtain valu=
e if final control is deemed too expensive or not attainable, we shall see.
>>>>>>>>=20
>>>>>>>> The analogy to rental of a consumable good does not apply to the ca=
se of a non-consumable good. If it cannot be traded and cannot be consumed i=
t cannot obtain marketable value. To this point it matters not whether it ex=
ists.
>>>>>>>=20
>>>>>>> I meant with control the control of entries in the register which I t=
hink is the deeper utility of Bitcoin. Final control is meant to be the oppo=
site of temporary which is the time limited control with some expiry.
>>>>>>>=20
>>>>>>> Thank you for your thoughts as they help to sharpen my arguments.
>>>>>>>=20
>>>>>>> Best,
>>>>>>>=20
>>>>>>> Tamas Blummer
>>>>>>>=20
>>>>>>>> Best,
>>>>>>>> Eric
>>>>>>>>=20
>>>>>>>>> Tamas Blummer
>=20