Throwing this idea out there for consideration/dissection/comment/criticism. Do people think this would be a valuable, viable and useful alternative to the existing exchanges?
Executive Summary: I'm considering setting up a fully-secured offshore vehicle, modeled along the lines of an Exchange-Traded Fund, which would allow people to safely, instantly, and anonymously "dollarize" their bitcoins at prevailing market rates, thus locking in the US Dollar value of their bitcash and fully protecting them (in US Dollar terms) against volatility of the price of Bitcoins in USD without actually having to convert the BTC to USD (with all of the attendant Know-Your-Customer, Anti-Money-Laundering, anonymity-sacrificing implications that raises). As the underlying vehicle would be fully and transparently backed in full by USDollar cash and cash equivalents, it would also eliminate the counterparty risk which is currently inherent in all dealings with existing exchanges.
___
The (much) longer version: I've been following Bitcoins with interest for the past 18 months or so, and mulling over various ideas for business propositions. For quite a while I've been thinking about the benefits of a Bitcoin "Exchange-Traded Fund" which would allow investors to invest in BTCs by purchasing shares of the ETF on a regulated exchange. (I know others have had the same or similar ideas, as I've seen some discussion on this forum of variations of this idea.) But, while it might be kind of "neat" if it happened, I kept coming to the same conclusion: (1) there are an awful lot of regulatory hurdles which would have to be met to accomplish this. And, while I think we may eventually see such things being traded if the Bitcoin Experiment continues to flourish, (2) I see no particular advantage (and many disadvantages) which I or any other little-guy would have in getting this thing up and running smoothly -- in other words, if and when a Bitcoin ETF were to become an economically viable thing, we will likely see the first BTC ETFs set up and run by the major players in the asset-management/ETF industry, like Barclay's Capital, State Street Bank, etc. (3) Furthermore, at least at the current USD-price of BTCs, there's really a relatively small potential profit in establishing and running such an ETF: basically, the manager/operator of an ETF covers overhead (and hopefully makes a small profit) by charging a percentage of the "assets under management;" and because your assets are BTCs which you're actually taking out of circulation for so long as people remain invested in your ETF, your theoretical maximum is capped by the total BTC money supply (i.e., the value of all BTCs in existence). Finally, (4) I don't see a BTC ETF really adding much to the long-term viability or stability of the Bitcoin Experiment. I mean, there's no doubt people who are holding bitcoin today could see a big upside, as it could well lead to a new speculative bubble as people who have heard about this exciting new Bitcoin thing (but otherwise can't be bothered to figure out how to go exchange fiat currency for Bitcoins today) decide to plough half of their self-directed IRAs and 401(k)s into BTC ETF shares -- but other than helping drive a new bubble I don't see it as actually doing much to change the Bitcoin landscape.
So, then I started thinking about what happens if I basically took the mirror-image of my Bitcoin ETF idea, and I realized that doing so potentially allows for leveraging the unique aspects of Bitcoins and the Bitcoin system in some very interesting ways (and which also, I think, could be up and running quite quickly as the structure obviates the need (inherent in an ETF traded on a US securities exchange and subject to SEC regulation) for perhaps months of back-and-forth with the SEC, first just explaining to them what Bitcoins are and how the system works, and then trying to obtain the necessary exemptive relief to get the listing approved, etc.).
So, as the basic starting point, rather than: (x) "a real-world ETF holding a pool of bitcoins (held in air-gapped wallets which are created offline and saved onto USB thumb drives which are physically secured in a fire-proof vault, and thus (hopefully) immune from online theft), each share of which corresponds to a fixed number of bitcoins within that total pool, and the shares of which are listed on a real-world exchange in a price denominated in USD and traded among and between real-world named-and-identified legal persons," --
instead think: "an online exchange where pseudonymous netizens can freely trade ("virtual"?) shares representing [either (i) an undivided interest in the corpus of a trust or (ii) a preferred equity interest in a legal entity (in either case created in and governed by the law of, say, Bermuda, the Caymans, the BVI, or whatever other offshore jurisdiction is determined to be most well-suited), the price of which shares are denominated in BTC, and the sole asset of which trust or entity is either (i) a briefcase full of USD$100 bills stuffed under a mattress residing in a hurricane-proof vault in downtown George Town, Grand Cayman, or (ii) better yet, interest-bearing secure short-term cash-equivalents, e.g. overnight repos."
___
(In other words, think of it as stepping through the mirror so that our baseline/mindset shifts from (A) "real-world USD-functional-currency legal person, looking to dip his/her/its toes into the Bitcoin economy and/or hedge his/her/its BTC-USD forex risk by forking over USD to lock in the USD-equivalent value of a nominal amount of BTCs at today's BTC/USD rate, by investing in a safe USD-denominated exchange-traded vehicle linked to the price of Bitcoin" to (B) "BTC-functional-currency virtual person, looking to dip his/her/its toes into the global USD-based economy and/or hedge his/her/its USD-BTC forex risk by handing over BTC to lock in the BTC-equivalent value of a nominal amount of USD at today's USD/BTC rate.")
Other than this inversion, however, we still preserve the basic "Exchange-Traded Fund Model" for how the fund operates: i.e., the average investor can freely trade in and out of shares of the fund in exchange for BTC via an online exchange/order-matching system, but otherwise has no right to either create new shares (via a USD capital contribution) *nor* to have any shares it holds redeemed out by the fund (i.e., in exchange for the associated amount of USD cash). Just as with a real-world ETF, as to the small (and/or anonymous) investor, our Fund looks and operates like a Closed-End Mutual Fund.
BUT, just as with a real-world ETF, there will be a special class of "Authorized Participants" who have the ability to create or retire blocks of stock, contributing or withdrawing the corresponding amount of USD cash to or from the Fund's asset base -- and (as with real-world ETFs) it is this mechanism (and the theoretical arbitrage opportunity it creates) which ensures that the shares of the Fund will trade at a price tightly-tied to the spot market price, in BTC, of USD.
___
So, to put some numbers around it: suppose we say that one share of the Fund will correspond to USD$100 held by the Fund. (Just to simplify (although I don't think there's any reason we couldn't easily handle fractional shares) let's say that shares only trade in whole integer amounts.) So at the current price over at MTGOX of about USD$11.00 per BTC, that suggests that 1 share of the Fund should be trading at 9.0909BTC.
Further, let's say that Authorized Participants have the right to either create or redeem blocks of stock in increments of 1000 shares (in exchange for a contribution/cash-out of USD$100,000 via wire-transfer to/from the Fund's Caymans bank account).
(*As an aside, note that this is where arbitrage keeps the Fund's share price from getting out of line with the "spot rate" -- if the market price of BTC/USD over at MTGOX moves down to $10.00 (and assuming sufficiently deep liquidity) but the BTC-denominated share price of our Fund for some reason fails to immediately follow it by moving up to 10.0000BTC but instead stays at 9.0909BTC, then the quickest-acting Authorized Participant can collect risk-free profits simply by snapping up 1000 shares of the Fund (at a cost of 9,090BTC), redeeming them for USD$100,000, using $91,000 of that cash to buy and replace the 9,090 BTC it just spent to acquire the shares, and pocketing the remaining $9,000 of risk-free profits -- or, of course, if they'd prefer to take their profits in BTC, the AP could acquire a full 10,000BTC, and pocket the incremental 910BTC as profit.)
Note that, under this structure, USD only ever moves between the Fund and Authorized Participants, never to or from small (or not-so-small but nonetheless non-Authorized-Participant) investors. And I think this is potentially very important: Because, furthermore -- and unlike the small investor trading in and out of shares of the Fund solely for BTC, who is free to retain his/her/its anonymity being identified solely by a BTC Address within its control -- the small, finite number of Authorized Participants will be known to the Fund and can be vetted by the Fund to ensure compliance with applicable Know-Your-Customer/Anti-Money-Laundering rules.
___
So, what has our small pseudonymous every(wo)man investor achieved by buying shares of the Fund? After all, you may say, s/he hasn't actually cashed out his Bitcoins and turned them into US Dollars, right? True, but effectively, by acquiring an interest in the Fund, s/he has turned her bitcoins into a USD asset. If the price of BTC in USD suddenly plunges, she's fully hedged, and the price of the Fund shares she holds, as denominated in BTC, will go up to reflect the new exchange rate and the fact that each share really does continue to represent an undivided interest in the USD$100 of cash sitting in the Fund's bank account.
And in some sense, if/as the Bitcoin economy continues to expand and grow more vibrant, in some sense her "dollarized BTCs" are even *more* liquid than actual US Dollars (either in the form of Federal Reserve Notes, or as a digital accounting entry in a checking account with ChaseCitiUSBankofAmericaWellsFargo), because if our investor decides tomorrow she wants to buy alpaca socks, she can simply sell some of her shares to reconvert a sufficient amount of her money from "dollarized BTC ETF shares" back into actual BTCs, place her online order, and send the BTCs.
Equally importantly (and this is where the unique aspects of the bitcoin system comes into play) -- (i) the Fund itself does not need to hold, accept, or deliver ANY bitcoins at all; (ii) nor does it transact in USD at all with anyone other than Authorized Participants who will have to provide all required identifying information to comply with anti-money-laundering and related laws and regulations; and (iii) theoretically, it could be set up so that "investors" (other than, of course, Authorized Participants) don't even need to set up an "account" with the Fund in order to trade; rather, an investment in shares of the Fund could simply be tied to a particular BTC wallet address. This is important, because the couple of major "hacking incidents" involving bitcoin currency exchanges over the past 12 months or so have resulted NOT from any fundamental security issue in the bitcoin system itself, but rather (in the case of MTGOX last summer) because the exchange failed to properly secure its database of userID/password combinations, or (in the case of Bitcoinica this spring) because the exchange failed to properly secure its own BTC wallet. With the setup I'm envisioning, neither of those would be a risk here because there are no UserID/password combinations to be hacked, nor is there a BTC wallet of the Fund to be stolen.
___
To avoid the need for unique accounts, here's what I'm thinking: whether you already own shares in the fund or have bitcoins and are looking to buy shares, you go to the Fund's secure website and establish that you are, in fact, the person in control of a particular BTC address with a sort of "BTC micropayment handshake" -- i.e., the website tells you to send some random small amount of BTCs over the P2P Bitcoin network to a BTC address within the Fund's control, and upon receipt and confirmation in the blockchain the Fund will transfer the same amount right back to the investor's address. (So I guess I may have lied when I said the Fund didn't need to have a BTC wallet at all -- but it could literally be holding less than 1 BTC at any given moment in time).
Once an investor is verified and in a secure session (identified by the investor's now-verified BTC address), the website essentially just functions as an online exchange with a standard bid-ask order book -- if that investor (i.e., that BTC address) already holds shares in the Fund, then the investor can enter an offer(s) to sell shares in any amount up to the total # of shares they currently own.
Conversely, an investor looking to buy shares submits a bid to buy. Obviously we need to ensure that investor is good for the BTC funds if the bid gets matched, and there are a couple of different ways this could be done. Probably the simplest (but for a number of reasons, only one of which is a concern about the Fund's wallet getting hacked Bitcoinica-style, a much less favorable solution) would be to toss out the window everything I said about the Fund/Exchange not having a sizable BTC wallet of its own vulnerable to theft -- so, an investor looking to buy shares of the Fund would have to transmit BTC to the Exchange, and would then effectively have a BTC account with the Exchange which could be used to support bids to buy. The Exchange would then be in the position of a central clearinghouse: when an order gets matched, the Exchange/Fund (which already has the BTC funds in its wallet) transmits the buyer's BTC funds to the seller's BTC address, and at the same time adjusts the share register of the Fund to reflect the fact that the (say) 150 shares just transacted are no longer owned by the person who controls BTC address [XYZ123...] (i.e., the Seller) and instead are now owned by the person who controls BTC address [UVW987...] (i.e., the Buyer).
But -- to get a little bit more sophisticated -- I don't actually think the BTCs ever have to even pass through the hands of the Fund/Exchange at all: suppose instead of transmitting BTCs to the Exchange, the Buyer can establish that it has a certain # of BTC in its wallet and can reliably commit to not moving them out (i.e., to not spending them elsewhere) of that BTC address, for so long as the buyer's bid remains outstanding. If buyer cancels its bid, it will be released from this commitment and be immediately free to spend them elsewhere -- by contrast, if the bid order is filled, the BTCs will move directly to the BTC address(es) of the Seller(s), and again the share register will be modified to reflect the change in control of the transacted shares. As I understand it, this will be a fairly straightforward thing once multi-signature transactions are impleemented, enabling automated escrow- and deposit-type arrangements along these lines (where the network itself is serving as the escrow agent).
So in this setup, the Buyer would be able to commit the BTCs by effectively depositing them or placing them on escrow with the Exchange for a set period of time, and then could enter bid orders (which would in any case need to either be filled or expire before the expiration of the deposit/escrow of the supporting BTCs). When an order is matched, the exchange could automatically monitor the blockchain -- and once it detects that the Buyer has released his funds from escrow and they've been transferred into the Seller's wallet, the transaction is fully settled and closed out as the ownership of those shares is transferred from the Seller to the Buyer in the Fund's book-entry share ownership register (which could even be made publicly available online, with share ownership tied to BTC address -- and so, likewise, could the USD assets of the Fund be constantly and instantaneously disclosed online, so that anyone can see that an any given point in time the assets of the Fund always exactly equals (or slightly exceeds) the number of shares outstanding multiplied by $100).
___
While this may all seem a bit complicated under the hood (but I think much/most of the complication could be largely invisible to the end-user/investor), there are reasons why I think this is a good way to go. Among others (but certainly at the top of the list), I want to steer as far as I can from what "e-gold" was doing, which resulted in the indictment and prosecution (ending in guilty pleas) of the owners on charges of money laundering and operating an unlicensed money transmitting business. I think I've largely achieved that, but would obviously want to have an expert in the money-laundering/know-your-customer rules take a good hard look at it.
___
Finally: How, you might ask, is this thing supposed to make money? A couple of different ways it could be set up: (i) the Exchange could collect a small transaction fee (say, 0.5%, which is less than the big exchange sites currently charge) on every filled transaction, and/or (ii) recall now that the Fund itself is now sitting on a pile of USD, which are supporting the total share float on a fixed ratio. If this cash is invested in short-term repos, it will generate a positive (admittedly quite small, in historical terms, given the current interest rate environment) return. (Now maybe a sophisticated investor will say they ought to be entitled to some or all of that interest, but if instead investors are comfortable seeing this as a pure hedge of cash -- i.e., I'm exchanging my BTC cash for an asset equivalent in value to USD cash, and wouldn't expect to earn interest on either -- then we can just keep the USD-per-share ratio fixed (at, e.g., the $100 I threw out there up above), and the interest earnings on the repos could be streamed straight to the Fund management to cover expenses and provide a profit.)
___
Really-truly-Finally: I need to think about it a bit more, but recall that I mentioned above that the potential value and profitability of the original BTC ETF idea is limited by the total value of the entire outstanding "monetary base" of BTCs in circulation. Here, by contrast, it occurs to me that -- analogous to the way the absolute value of derivative exposure can exceed (by orders of magnitude) the total value of the underlying -- it could be the case that, if it really took off, the USD assets-under-management of the Fund are not necessarily limited in the same way (or, rather, the theoretical upper bound limit is not the amount of BTCs in circulation, but the amount of USD in circulation).
I think it all depends on the velocity of BTCs in the "Bitcoin Economy," but (for example) let's say that Goldman Sachs (hah!) signs up to become one of our Authorized Participants, and they see profitable opportunities selling BTCs for USD at a positive spread. So, they pump in $1,000,000 today, creating 10,000 shares which they're able to sell over the course of the day on our online Exchange for ~91,000BTC; they then take those 91,000BTC and sell them to their clients (at a profit).
And suppose they do this every day for 6 months -- and assume the exchange rate is holding relatively steady during this time period -- by which point they've churned through over 16MM BTC, i.e., more than the total # of BTC currently outstanding. So, impossible, right? Well, if their clients are just sitting on the BTCs as a speculative investment, then yes.
But if, instead, their clients are recycling the BTCs, using them to buy alpaca socks and high-quality goods on (ahem) Silk Road, and make BTC-denominated loans to their own client-customers who in turn spend them: well then the BTCs haven't been removed from circulation at all, and each day they may well be a fresh demand for more shares of our Fund, so the USD assets of the Fund could grow pretty much indefinitely large. (Not saying this *would* happen, just that it could).
