# Last edited on 2017-02-10 06:30:58 by jstolfi

Gemini auction is very new and without precendent in bitcoin.
Reckless to make it the primary reference for the price.

It has low volume, many days don't close (OK);
many days with less than 100 BTC (not OK, could be manipulated,
e.g. by rival ETF.

Fallback option "Sponsor's good faith judgement" makes Gemini etc irrelevant and lets 
Sponsors to set the NAV as they please, without recourse 
by shareholders.

Fact is: due to the many exchanges, mostly irregular and highly prone to fraudulent 
operation and closure, the maket value of bitcoin 
is much harder to define beforehand.  

List of regulations that constrain Gemini does not include the trading in its exchanges.

Coin split cannot join.  

Note the Ethereum example.

In the event of a coin split into Bitcoin-A and Bitcoin-B, the fund
will hold equal amounts of both versions of the coin. This situation
would be similar to that of a fund that holds shares of a company that
creates a spinoff through stock dividends.

The filing should specify what the ETF will do in such case.
The Trust cannot simply ignore one version of the coin, because those coins
are property of the clients who hold shares.  (Cite Coinbase) It could:

 (1) Keep only one one version, say Bitcoin-A, and transfer the correponding Bitcoin-B 
 to the Authorized Participants who deposited the original bitcoin.
 These Participants then would have to be obliged, by the ETF 
 contract and regulations, to return those Bitcoins-B
 or their equivalent in currency to the shareholders.
 
 (2) Keep both coins as being backing assets of the share; that is,
 each share of the ETF would represent 0.01 bitcoins "A" and 0.01 bitcoins "B". 
 The filing should then be amended to say explain that "bitcoin" as held by the 
 fund means "equal amounts of all 
 current versions of bitcoin, identically split among the same blockchain addresses". It must also specify how the NAV will be calculated.
 That requires finding the market value of each version, and adding them.
 One of the versions may have very little liquidity causing unstable price.
 Fuethermore, different blockchain transaction fees may apply to each version.
 
 (3) Terminate the fund.  This may be the simplest course to describe.
 However, a coin split can be executed by anyone, and a relatively  modest 
 investment may be sufficient to give it, even of temporarilty, a non-negligible value.  
 Then this proviso in the filing could be exploited by a hostile entity --
 such as a rival Bitcoin ETF -- to force the liquidation of COIN.

In any case, the filing should also consider the handling of splits for bitcoins that 
"in transit" -- in the hands of the Authorized Participants,
just before the creation of a basket, or just after its redemption.

The "market cap" is actually an investment deficit.

Bitcoin is a fiat currency.   Using "fiat" in the sense of "government-issued",
in opposition to bitcoin,
gives the non-bitcoiner reader the impression that bitcoin is more solid than 
fiat currencies; when in fact it is less solid.  The prospectus must

  (a) point out that bitcoin, not being backed by a commodity, is 
  a "fiat currency", according to the proper meaning of that term;
  
  (b) use "government-issued currency" or a similar unambiguous term
  instead of "fiat currency"
  
  
"I told you" with respect to other bitcoin funds.  Note that there
is being launched an OTC-traded fund that holds the 
crypocurrency "Ethereum Classic" (ETC)
instead of bitcoins.  

A fund based on virtual assets and offers no insurance is very cheap to set up, easily lucrative through the charging of administration fees, and would immediately compete with other similar funds from the start.  The SEC can expect a deluge of requests to authorize such funds, if they authorize one of them; and there will be no objective reason to negate approval for copycats. 

Remind that private keys only give possession; they do not prove property, not even *exclusive* possession.  The Trust can prove that it has control over the bitcoins, but cannot prove that no one else has. It cannot prove that those bitcoins are not encumbered by contract, since there is no mandatory record of bitcoin contracts.   That is unlike a gold fund (say) where the entity that holds the commodity can show precautions that will prevent access to and removal of the gold by unauthoized persons.

The "proprietary, patent-pending" security measures, not having been
reviewed by anyone outside the Sponsors, are less trustworthy than
public methods.

A passive fund based on a material asset, such as gold or oil, can effectively protect and demonstrate possession of the assets that it claims to have.  Theft of such assets would require physical access and removal of the material.  Auditors and insurers can verify that the physical measures in place -- such as vaults, cameras, and guards -- are adequate to reduce the risk of physical theft to a level that can be effectively estimated and insured against.  

In contrast, a fund based on a cryptocurrency can ensure the continuing possession of its assets only by generating truly random private keys that cannot be guessed, and by keeping them secret from would-be thieves.  Both requirements are impossible to verify by any auditing method: there is no way to measure the randomness of a specific private key, and merely ascertaining that a copy of it is safely stored, in a bank vault or the like, will not ensure that other copies are not stored under insecure conditions, or known to unauthorized persons.  Moreover, the risk of those two requirements being violated is impossible to estimate, even with frequent on-site inspections.  That is the main reason why there is no insurance available against bitcoin theft.