There have been various theories floating around how CoinLab could possibly be justifying this number, with little insight into their actual arguments, a problem not helped by CoinLab's attempts to seal official Japanese creditor records in order to conceal their actions. Recently though I acquired a copy of the last round of petitions to the Tokyo District Court, and have spent the last week dredging through the legal Japanese to get to the bottom of the situation.
DISCLAIMER: While I am more familiar with the underlying subject matter than most people, I am not a lawyer and my legal interpretations may be incorrect. Do not take my commentary as cause to relax or let your guard down when it comes to CoinLab or other bad actors.
MtGox creditors already have no love for Vessenes, who has steadfastly refused to drop or lower this claim, which is now almost two and a half times as large as the sum total of all approved claims combined, despite the fact that the remaining assets in MtGox are customer funds. Basically, if you had bitcoin in MtGox you've already lost 80%+ of your money, and now CoinLab is doing their best to finish the job.
I'll mainly go through two important documents: CoinLab's civil rehabilitation claim assessment petition (filed March 6, 2019) and trustee's response (filed April 12, 2019). A lot of the content is Japanese legal rambling with point-by-point responses from the trustee, so I'll merely summarize the documents together along with my commentary.
The monetary claimCoinLab claims the following in the MtGox civil rehabilitation:
|N5-1||Claim for revenue sharing||27,031,875,695 JPY|
|N5-2||Claim for damages||1,020,160,340,947 JPY|
|N5-3||Determined delay damages for N5-1||14,007,427,733 JPY|
|N5-4||Determined delay damages for N5-2||628,867,640,563 JPY|
|N5-5||Undetermined delay damages||(N5-1 + N5-2) × 12% per year|
The trustee insists that CoinLab's claims all be assessed as 0 JPY, and that CoinLab pay its own legal fees.
Background for claimsOn November 22, 2012, MtGox and CoinLab entered into an exclusive license agreement for offering MtGox's exchange services in the U.S. and Canada. The background for this agreement was ongoing concerns about legal compliance in the U.S. and MtGox's lack of a U.S. banking presence, having historically required customer deposits to be wired to a Japanese bank account. MtGox was eyeing partnering with an existing U.S. company to provide a legal foothold locally, similar to how it had previously acquired Bitomat to gain a presence in the European market.
(This wasn't the first time Vessenes eyed MtGox; he had previously been in communication with Jed McCaleb and expressed strong interest to work together, and after McCaleb handed over MtGox to Mark Karpelès, Vessenes objected and claimed to own half of MtGox based on supposed earlier conversations with McCaleb. McCaleb rejected any such ownership claims, but clearly MtGox remained an obsession for Vessenes for years to come.)
The big allure for CoinLab was a revenue sharing agreement, where CoinLab would be entitled to 40% of all fee revenue for all existing U.S. and Canada based customers (which would be migrated to be handled by CoinLab), and 90% of all fee revenue for new U.S. and Canada based customers that signed up after the transition.
After the signing of this agreement, history sharply diverges depending on who you ask. In the Vessenes universe, CoinLab was standing by fully legally compliant and ready to receive MtGox's customers, but MtGox blankly refused to cooperate and fulfill its contractual obligations, tried to weasel out of the arrangement in order to hog all the bitcoin trading fees for itself, and were it not for MtGox reneging on this contract Peter Vessenes would have become the richest man in crypto.
In the MtGox universe, the one that the trustee argues (and the one I would argue is the one we all actually live in), the contract was probably rendered invalid and/or CoinLab breached it by not being legally compliant in all relevant U.S. states, failing to fulfill the whole point of the agreement, and their claims calculations are absolute nonsense.
Regulatory complianceThe license agreement stipulates that CoinLab had to be legally compliant by the end of the transition period, 120 days after signing, i.e. by March 22, 2013. Just days before that, on March 18, 2013, FinCEN issued guidelines for cryptocurrency exchanges, clarifying that they are "Money Transmitters" and need to register as such. CoinLab had previously registered with FinCEN as a seller of prepaid access (which people had speculated was how cryptocurrencies would be treated), and later updated this registration to money transmitter.
(Note that guidelines are interpretation of existing laws, not the writing of new laws; FinCEN was clarifying existing requirements, not adding new ones. This is worth pointing out as CoinLab's expert witness, Connie Fenchel, stated that no regulatory guidance had been issued prior to March 2013, but CoinLab's claim filing misquotes her as saying no regulations existed prior to March 2013.)
The bigger disagreement and unclear point was over state licensing, with MtGox's attorneys arguing that individual state licenses would be necessary, and with CoinLab (then and now) arguing that they were already fully compliant and that MtGox was just stalling. CoinLab did however later apply for a state license in Washington state and was researching licenses in other states, revealing that they too believed state licenses were necessary. CoinLab also somehow blames their lack of licenses on MtGox, arguing that they couldn't apply for licenses unless MtGox sent them complete customer data first.
The trustee agrees with MtGox's attorneys' opinion that CoinLab lacked required licenses, and makes multiple arguments based on Washington state law (the jurisdiction of the contract) regarding how contracts become invalid if their execution violates the law (Beroth v. Apollo College, Inc., Davidson v. Hensen, Stratford, Inc. v. Seattle Brewing & Malting Co.).
Specifically, CoinLab's FinCEN registration was incorrect as it falsely represented that CoinLab would only provide Money Service Business activities in the state of Washington. Additionally, since the business involved using CoinLab's bank account for fiat deposits/withdrawals for U.S. customers, the trustee argues such operations would have required state licenses in at minimum Alabama, Idaho, Maine, Pennsylvania, Texas, Virginia, Vermont and Washington as of March–April 2013. Providing these services without holding the relevant federal and state licenses is a criminal violation of 18 U.S. Code § 1960 punishable by fine or imprisonment.
DisputesAttached emails paint a picture of a troubled transition period, with each party trying to get the other to provide necessary materials; CoinLab in particular was pushing MtGox to send them data and money since day one (including demanding an audit be performed), whereas MtGox was dragging its feet and was seemingly hesitant to hand over sensitive materials until CoinLab showed that it was ready. CoinLab on the other hand was also slow on fulfilling certain obligations, and seemed eager to skip straight to the part where they get all the business and money.
After the end of the transition period, on March 29, 2013, MtGox sent CoinLab a formal letter calling them out on lacking licenses, stating that the transition couldn't be completed, and accusing CoinLab of having already started to receive bank wire deposits on behalf of MtGox customers (by using a preliminary API to credit MtGox accounts) without forwarding said funds to MtGox, to the tune of millions of dollars. (Out of an eventual 12 million USD deposited, CoinLab ended up forwarding 7 million USD to MtGox, with CoinLab still owing MtGox customers the remaining 5 million USD.)
While not stated clearly in the contract, MtGox strongly argued that no customers should have been handled by CoinLab until a CoinLab Terms of Service was in place and agreed to by the users, a matter that CoinLab seemed to affirm in emails but ignored in practice as it demanded immediate transfer of data and assets.
To MtGox' demerit, they don't seem to have inspected the writing of the contract very closely, instead basing much of their understanding on prior discussions and intentions, whereas Vessenes was clearly on top of every clause and favorable interpretation and quickly insisted on everything he was contractually entitled to.
Following the letter from MtGox, CoinLab sent MtGox a legal strategy for how CoinLab intended to achieve and maintain legal compliance. While this seemingly suggested a willingness to cooperate and fulfill their obligations, going by the letter of the agreement (which CoinLab and Vessenes has insisted on doing for everything else) CoinLab had to be fully legally compliant by March 22, 2013, not "eventually". What's good for the goose is good for the gander.
The trustee argues that MtGox's March, 2013 letter informed CoinLab of a contractual breach and voided the entire partnership. The dispute between CoinLab and MtGox continued throughout 2013 and escalated with a lawsuit filed by CoinLab later that year.
Claim size calculationThis is probably what you started reading this for, and it's a real doozy. Grab a chair.
CoinLab argues that because MtGox breached the license agreement in multiple ways, CoinLab was unfairly robbed of revenue that they would have earned had the license agreement been fulfilled. MtGox collapsed in February 2014, so that's a grand total of a 15 months of revenue for CoinLab to possibly claim, right? Wrong; CoinLab rejects your reality and substitutes its own!
First, CoinLab has a small problem in that MtGox has only provided it with trade data for part of 2013, so it needs to extrapolate in order to estimate their claim. They do this by looking up global trade volume estimates on Bitcoinity and Coin Metrics, and estimating that MtGox trades for which CoinLab were entitled to fee revenue accounted for about 25% of global trade volume for that time period. They then do the same calculation in reverse for other time periods, reasoning that CoinLab trades continued to account for about 25% of global volume.
Second, even for the trades for which CoinLab has data, they apply questionable interpretations. For example, MtGox granted certain accounts (e.g. the Bitcoin Foundation) zero-fee trading, but CoinLab retroactively rejects this and demands a cut of these trading fees that were never collected (instead applying a flat 0.2% trading fee for everything), and accuses MtGox of forging their fee records in order to get out of sharing revenue. CoinLab also makes blanket assumptions about customer sign-up dates, conveniently making it easier to count trades as belonging to the category where CoinLab takes a 90% cut rather than a 40% cut of trade fee revenue.
So far this sounds questionable but not too crazy, but here's where CoinLab's claim starts going off the rails. CoinLab reasons that since the agreement was for a term of 10 years, plus a strangely one-sided post-termination clause giving them continued revenue sharing for an additional 5 years after termination, CoinLab is actually owed a full 15 years of revenue on 25% of global trade volume. They're basing their claim on the assumption that the MtGox collapse never happened, first extrapolating their 25% "share" of global trade volume all the way until present date, and then extrapolating the last 12 months worth of trading all the way to 2027 and taking "their cut" on trades that haven't even happened yet.
Specifically, CoinLab argues that they are owed 27,031,875,596 JPY ($242,590,645.21 USD) of actual revenue, that is, for the period MtGox was actually operating. They then calculate that between March 2014 and September 2018 they are owed damages for lost revenue equal to global trade volume × 25% [CoinLab's "market share"] × 0.2% [minimum MtGox trading fee] × 40% [CoinLab's revenue share], both on the USD and BTC side, amounting to $537,575,951.9 USD and 95,702.8777 BTC. They then find that this works out to an average of $39,610,903.18 USD and 4,313.140943 BTC per month for the last year, and extrapolate those levels all the way through 2027, resulting in total damages of $4,855,164,398.44 USD and 565,835.33386 BTC. At the civil rehabilitation exchange rates, this translates to 1,020,160,340,947 JPY ($9,155,167,737.12 USD).
We're not done yet though! With creditor claims come delay damages, in the case of bankruptcy counted from the cessation of operations until the commencement of bankruptcy proceedings. Normally these don't matter much since they are awarded pro rata and it's still the same fixed amount of assets being liquidated and distributed. CoinLab however insists that rather than the standard 6% yearly interest rate they should get delay damages of 12% per year (based on a Washington state law that doesn't seem applicable) for the full period from the cessation of operations until claim payouts.
They therefore figure their delay damages for the time up to the commencement of civil rehabilitation in 2018 works out to an additional 14,007,427,733 JPY ($125,706,073.17 USD) for their revenue claim and 628,867,640,563 JPY ($5,643,611,599.78 USD) for their damages claim. Further, they are claiming a running tally of 12% per year for the delay from the start of civil rehabilitation until payouts are actually made, in theory setting themselves up to benefit from dragging out proceedings as their claims grow faster than other creditors'.
All in all, CoinLab is claiming $15,167,076,055 USD plus an extra $1,127,731,005 USD per year that the proceedings drag out past June 2018. Not only that, they have the gall to refer to these numbers as "conservative estimates", and warn that as they get more data from MtGox the claim may need to be adjusted upwards.
The trustee plainly calls CoinLab's numbers impossible and completely groundless, and disputes every aspect of this joke of a "calculation method". When the trustee did the calculations to estimate the revenue share CoinLab would have been entitled to under the license agreement he arrived at a figure of approximately 636,000,000 JPY ($6,100,000 USD), suggesting CoinLab's figure for the same time period is 40 times too large.
Note that the contract also contains a liquidated damages clause for $50,000,000 USD, something that has been previously mentioned publicly. CoinLab does not in fact appear to be currently arguing for these damages, and the trustee is emphatically stating that this clause is not applicable since it only refers to breaches licensing rights exclusivity.
Other argumentsEach filing is tens of pages long and full of legalese so I can't adequately describe them in full, but some of the other arguments being made by each side include:
- CoinLab claims that the sudden bitcoin price increase in Q1 2013 was thanks to their partnership with MtGox being announced on February 27, 2013. See if you can spot it in the graph!
- CoinLab claims that MtGox has been trying to get out of the agreement for the purpose of keeping all trading fees for themselves.
- CoinLab argues that since MtGox at one point asked if Vessenes would let MtGox simply buy CoinLab instead, MtGox was admitting that CoinLab was legally compliant. The trustee counters that MtGox was simply exploring all options for a legal foothold in the U.S. market, since creating a new entity from scratch would be a much larger effort.
- CoinLab argues that by continuing to offer its service directly to U.S. customers, MtGox violated the exclusive license agreement. The trustee counters that the exclusivity clause in question is about licensing the MtGox service, not offering the MtGox service to users (and MtGox doesn't need to license their own service to themselves). There is another clause prohibiting MtGox from offering the service to CoinLab customers, but this is not the material clause that CoinLab is trying to invoke.
- CoinLab argues that MtGox breached its representations and warranties (e.g. to live up to highest industry standards) as evidenced by MtGox becoming bankrupt, Mark Karpelès being arrested, and people saying bad things about MtGox. The trustee counters that such representations are not requirements to be infallible, and that CoinLab has not pointed out anything specific industry standard at the time that MtGox failed to live up to.
- The trustee argues that under Washington state law, a material breach of a contract excuses any further performance by the other party (DC Farms, LLC v. Conagra Foods Lamb Weston, Inc.). Since CoinLab breached their legal compliance covenant (which was the whole point of the partnership), whatever other things MtGox did or didn't do is irrelevant in that situation.
- The trustee argues that since MtGox had reasonable grounds to expect CoinLab would fail their duties under the contract, MtGox was allowed to excuse their own contractual obligations under the principle of Restatement (Second) of Contracts § 268. Specifically, MtGox's attorneys had advised them that CoinLab required additional licenses in order to be legally compliant.
Conclusion and other commentsThe trustee appears to be thoroughly fighting every part of CoinLab's claim, with no indication of negotiation or compromise. While this will probably please a lot of creditors who don't want to see Peter Vessenes profit a single penny from people's misfortune, the fact that CoinLab also isn't showing any signs of backing down reinforces fears that we'll probably see a long and drawn-out process until CoinLab exhausts every legal avenue to try to argue their case.
Other things I noticed while reading:
- CoinLab's claim contains a number of typos, and parts of it feel like literal translations from English. CoinLab's Japanese counsel is likely being directly instructed from the U.S. rather than bringing a locally grounded case.
- Compared to the trustee's arguments, CoinLab's arguments feel winding and insecure, where every mere mention of alleged wrongdoings by CoinLab comes with a mandatory parenthesis (to remind the reader that THIS IS NOT TRUE! MTGOX ARE THE BAD GUYS!).
- CoinLab's claims are held to be self-evident and are merely listed (or ranted) rather than argued, and their "research" is full off self-serving assumptions and extrapolations that no self-respecting scientist could ever make with a straight face.
To continue to argue this frivolous claim at the direct expense of tens of thousands of people who actually lost their own money is utterly shameless. Peter Vessenes and anyone currently associated with CoinLab are blatantly trying to rob honest creditors and deserve to be publicly shamed until they do the right thing and drop this claim.