On token liquidity

Note: I’m not your lawyer, this is not legal advice.

Lately I’ve been hearing a lot about how Crypto-Tokens are a magical solution to startup and VC liquidity problems.

I half-blame Balaji Srinivasan for starting this meme with his very cogent and forward-thinking blog post Thoughts on Tokens, in which he wrote:

tokens are not equity, but are more similar to paid API keys. Nevertheless, they may represent a >1000X improvement in the time-to-liquidity and a >100X improvement in the size of the buyer base relative to traditional means for US technology financing — like a Kickstarter on steroids. This in turn opens up the space for funding new kinds of projects previously off-limits to venture capital, including open source protocols and projects with fast 2X return potential.

I concede that, to an untrained eye, this is indeed what tokens, as we usually encounter them in the wild, appear to do. People make the tokens, people sell the tokens, and voilà, seed-stage startups with big ideas and a handful of staff suddenly have liquidity galore – up to $300 million to spend on precious metals and bonds, launch $50 million venture funds, and give early investors their exits, all before releasing a single line of code.

Because that’s totally normal.

Unfortunately, when real assets get involved, the tokens-are-magic-liquidity-engines thesis falls apart almost immediately.

Equally unfortunate is the fact that Balaji, being more influential than I am, can spread his memes farther and faster than I can spread mine. One recent victim of the token liquidity meme, Stephen McKeon, wrote a blog post on the subject titled “Traditional Asset Tokenization.” He takes Balaji’s thesis and expands:

Tokenizing relatively illiquid assets and creating a market in which to trade these tokens can reduce the illiquidity discount substantially by reducing frictions to trade. Traditional assets will tokenize because they will lose the liquidity premium if they don’t.

The premise of course is that

Since tokens are highly divisible, it eliminates the need for minimum investments. Blockchain facilitates trade in the secondary market.

Consider commercial real estate. Publicly listed Real Estate Investment Trusts (REITs) provide some liquidity, but they are expensive to set up and typically hold a basket of properties rather than a single building. Furthermore, REITs are typically buy and hold vehicles, so most investors are shut out of development projects entirely unless they can meet minimums which are often on the order of $25k and higher. One day you might be able to buy $10 of a single commercial real estate asset like the Empire State Building, or invest $100 in the development of a LEED-certified housing project.

So we’re told that tokens are less expensive to spin up than a REIT (true) and as a result you can access a wider investment market of Bitcoin people by listing your token on, say, Kraken.

We’re then told that structuring a transaction thus

will increase liquidity for asset owners. Real estate appraisers will become more like equity analysts, because the market value of any building will be readily apparent if the token is trading. Tokenization will make IRS 1031 exchanges easy.

EDIT, 23 AUGUST – there’s also  this thought leadership piece from TechCrunch which I came across today:

Tokenizing real-world assets will allow buyers to access assets never before within their reach, and sellers to move assets that were previously difficult to unload. The secret lies in the possibility of fractionalization.

Where I’m left is that everyone is repeating the “tokens make X easier because [buzzword]” mantra ad nauseam but nobody can explain, in a technical way, why.

Which I’m about to do, naturally.

So. Why are tokens cheaper, faster and easier than traditional legal structuring?

Because “tokenization” refers to choice of tech, not choice of legal relationship

The term “tokenization” is legally meaningless. What the asset-tokenizers are actually trying to achieve is something generally similar to debt securitization, the practice of turning boring, illiquid loans into exciting, tradable bonds, which kicked off in the mid-1980s and nearly destroyed the world economy in 2008.

As it happens, this was the line of work I was in before I got into blockchains.

Let me break it down for you.

A) Utility Tokens

I like to divide tokens up into two classes. The first of these are the “utility tokens” where folks are buying the token to consume it in an application, such as Filecoin, or where the token has value based purely on speculation. This would be the case with, e.g., Bitcoin, where the ability to spend the UTXO (something you can do with the tech) allows you to write to the Bitcoin network, but doesn’t entitle you to rock up to a courtroom and ask a judge to compel some third party to render performance of some kind.

There’s no legal relationship between you and anyone else when you play with Bitcoin. There’s just the fact that you have a write permission on a database, and the hope that it’ll still be there and be worth something when you go to use it.

We would say the asset, with something like Bitcoin, is intrinsic to the blockchain. It has value only because other people are willing to pay for it; it doesn’t entitle the holder to make demands of other people. And in this situation, Balaji’s liquidity thesis continues to apply.

Lest we wonder why the “utility token” space currently exhibits the characteristic of having greater liquidity vis a vis more traditional asset classes (and generates concomitantly insane returns), the answer is because (a) they’re sold on unregulated markets to unsophisticated investors and (b) it’s one of the few trades around where banks can write analyst notes promoting the tech to institutional clients at the same time as they make a killing providing short-term dollar liquidity facilities to big players such as the exchanges (unconventional monetary policy + expensive liquidity facility = heaps of delicious margin).

Also (c) there are suggestions of some shenanigans at the exchanges – if we cast our minds back to 2013, the last run-up in the Bitcoin price also coincided with shenanigans at Mt Gox.

Given similar facts, one could probably turn the receivables from a chicken sandwich into the liquid investment-of-the-century too. The money would be worth it as long as you could stay out of jail (protip: you probably couldn’t). But we’re not dealing with securities regulation in this post, so I’m going to leave that to one side for now.

B) Actual investments

The second type of token is an “investment token,” i.e. a token we bundle together with a documentary intangible. If we speak of tokenizing a REIT, what we actually mean is using a token (again, just some tech) to automate or digitize aspects of a relationship between the tokenholder (as unitholder) and some other person which is capable of being enforced by a court order.

With a REIT the asset is the real estate and the income flows therefrom which are paid as dividends to trust unitholders; the arrangements have value because the promises to pay are enforceable. The coin is a mere representation of rights which are constituted by other instruments. Because it is software, it may also be a component in an application which allows the coin-holder to exercise those rights.

The value in such a transaction is therefore extrinsic to the blockchain; constituting the obligation (making it enforceable) is no easier than before and requires all of the same effort and expense. Tokenizing the obligation (turning it into a token) merely allows instantaneous peer-to-peer communications to be made about that obligation. The thing that has value is the enforceable promise, not the coin.

So when we speak of “tokenizing a REIT,” what we actually mean is “using a blockchain back-end to record an ownership interest in a meatspace security.”

Liquidity becomes available by taking the illiquid asset (bricks and mortar) and pooling it with many others like it so you can get a more easily tradable asset (e.g. a AAA-rated note) which allows a bank in, say, Japan to get exposure to some mortgages in southern Florida. Which makes mortgage loans, as a class, more liquid than they were before, because you can get them off your balance sheet before they mature.

This process, in relation to pools of contractual debt being transformed into fixed-income securities, is usually called securitization. 

When folks say “tokenizing” they usually mean “cutting corners”

When we hear people say they want to “tokenize an asset,” most people are talking about avoiding the expense and bother of securitization while achieving the same effect.

The DAO attempted something like this. Remember that? I do. Apart from being the most disastrous piece of public blockchain code ever deployed, legally, it was so broken as to be unfixable. Bear with me for a moment and I’ll explain why.

The DAO promised us that it would be a distributed venture capital fund. Which, in the usual course, should be a set of relationships and steps which look a little like this:

create funding vehicle —> LPs invest moneyz —> VCs use moneyz to purchase shares in FOMO Ltd —> hodl* —> sell shares in FOMO Ltd to Goolag, Inc. —> receive sale proceeds from sale of FOMO Ltd  to Goolag, Inc. —> apply distribution rights to funds realized —> distribute funds to LPs

Each step of that arrangement will be documented fully and be enforceable in a court of law. So if you said you would do something (like, distribute returns to your LPs) and you don’t, your LPs can go to court and get a judge to force you to do it, or provide them with damages in an amount equal to or exceeding their losses arising from your failed performance.

The DAO, on the other hand, promised a revolution in crowdfunding because we could short-cut the VCs and disrupt/disintermediate/[whatever language is en vogue in the Valley this week]. And indeed this has continued to be a major argument supporting the present ICO craze. The steps the DAO used, however, looked something like this:

Instantiate DAO smart contract —> Purchase worthless DAO token from smart contract —> vote on projects with voting rights pro rata to DAO Tokenholding —> projects receive ETH —> DAO might create “reward tokens,” which is attractive to Ethereum folks as they’re a bunch of token fetishists —> projects subsequently issue their own tokens which are not “reward tokens” and not issued directly to DAO Token holders —> ??? —> Profit

The difference between these two positions is

  • VCs sell assets, receive cash, and then distribute proceeds to LPs.
  • The DAO had no such mechanism, the marketing did not describe one, and nobody (apart from yours truly) appeared to notice this at the time.

DAO-funded projects were meant to issue their own tokens, and somehow, magically, at the end of the exercise profits came back up the chain – despite not having any way to automatically transfer tokens or token sale proceeds to DAO investors, or to get a position in a fundraising company’s cap table which, on liquidation, would flow back to DAO tokenholders. Although each DAO proposal allowed for so-called “reward tokens” to be issued to DAO token-holders, it was never really clear how these were supposed to work or what function they would serve in any finished application.

Even if an automatic funds flow/distribution mechanism had been provided for, features like “DAO Splits” (where rebel members of the fund could break off with their funds to create a new fund, while retaining the distribution rights originally granted by the old one) would have made them nearly impossible to implement in practice.

Long story short, all the DAO really allowed you to do was vote about how to disburse ETH. The rest was mostly pointless, as there was no means to get a return from one’s investment in the DAO except to drive speculation on DAO Tokens themselves.

This means the DAO was doomed from the start. Had it not been the subject of a $150mm hack it would have failed eventually due to its legal-technical shortcomings – shortcomings which were easily avoidable with the most cursory diligence. There is, however, a small ray of sunshine in this failure from the perspective of the DAO’s hapless creators/founders. Chiefly, the hack wrecked the DAO so quickly that the world didn’t have a chance to notice how ill-conceived the entire scheme truly was.

So why not try to repeat the DAO? Surely technology fixes all issues?

Not this time. You’re in the lawyers’ world now.

First, if you try to follow the DAO’s example and launch a quickie crypto version of some old-world investment, you’re not going to be selling what you claim to be selling. Selling tokens that purport to be shares in an asset without properly constituting the asset in documentation (i.e. the slow, expensive part) is fraud and will likely expose the seller to lawsuits.

Second, if you’re freely trading an e.g. REIT token, if you haven’t retained a fancy law firm to put together the docs and made all of the filings (i.e. the slow, expensive part) it’s probably all kinds of wrong from the perspective of securities law. I have already written on the consequences of breaking securities law at length. This will also expose the seller to lawsuits.

Third, if you screw up as mightily as the DAO did, rest assured I will still be writing about it a year later as I endeavor to correct the latest token offering craziness.


Tokenizing without “securitizing” properly is like saying you want to use the gold standard but would prefer to dispense with the gold. Or that you want to visit the Rocky Mountains but don’t want to hang out with the marmots. You’ve removed something so essential to the transaction that you might as well not have done it in the first place.

What we saw with the DAO is a phenomenon I call “cargo cult law.” Repeating that experience is inadvisable. When we speak of “tokenizing an asset” and somehow managing to avoid the expense of documenting the legal relationship or full compliance with, say, securities law, we’re not actually creating an asset. We’re creating a problem.

Given the level of sophistication of Bitcoin or Ether money, such schemes might be able to continue on the basis of speculation alone – at least for a little while. But, eventually, economics and legal process will kick in, they will fail, and promoters of any investment schemes that cut corners will find themselves in very hot water. (The jury’s still out on utility coins.)

That said, if you have the right business case and find the right lawyers to advise you, blockchains and securities can mix quite nicely. But that’s a post for another day.

* Yes, this is actually spelled “hodl.”

24 thoughts on “On token liquidity”

  1. Preston, here is a “compromise” position that I’d like to throw out.

    Maybe the ultimate fate of ICOs in the US market it simply to be….identical to IPOs? Maybe they are used to take private companies (with say, a 50 million dollars in sales) public, acting as an alternative to the NYSE or Amex. Tokens, at this point, are simply stock certificates. Maybe they have a bit more function, maybe they are electronic and fancy and digital. Maybe there are technical enhancements, but legally they are just plain old stock shares.

    This is what I think Balaji and David Sacks (another proponent of ICOs for tech firms) are heading towards. This is the goal of the T0 exchange I think.

    Why do ICOs when we already have IPOs? That’s just about politics. The perception in Silicon Valley and other startup-lands is that the IPO process is broken and corrupted. Too many greedy gatekeepers and rentseekers in the process. The “IPO drought” is a threat to national economic growth. The ICO model might be used for purely political reasons, as an alternative competitive stock market. The SEC could hardly in the long run object to that, if all the same rules are followed. There are hundreds if not thousands of perfectly viable, successful companies that simply for whatever reason don’t have what it takes to IPO via the traditional Wall Street-dominated path.

    What do you think?

    1. I think that’s bang-on and probably where we’ll wind up, key language being “if all the same rules are followed.” I also think we’ll probably still call the things IPOs and trade stock certificates. But technological advances pioneered in this bubble will likely make the process of IPOing and ongoing compliance much more efficient and therefore widen access to the market.

    2. While IPOs are ugly and expensive, IMHO the real reason for the drought is the new regulations on public on companies. The hassles and expenses and uncertainty of complying with Dodd-Frank and Sarbanes Oxley are just too high. If the company ends up “public” at the end of the ICO process, then they will face all the same problems.

  2. Not sure if I’m getting the wrong end of the stick here. Are you saying that to make any of the commitments in a tokenisation enforceable you would have to do all the legal work anyway? I see the complications with something as one-sided and open-ended as an IPO/ICO, but what about smaller smart contracts e.g. Financing where both parties are integrated with the tech and timings/deadlines are established and breaches of covenant are automatically enforced? If you can see that, isn’t it then just a question of scale? Or are you saying that the level of work to get everyone integrated and to agree on a framework would mean no saving from status quo? Or is it that the nature of the beast (the reliance on actual human action) means that something like an IPO (and e.g. distribution of any sale proceeds thereafter) could never be enforced by technology alone?

  3. Super post. I agree the jury is very much out on utility tokens (although I’m pretty pessimistic) but I reckon we can close the case on tokenised securities on public blockchains. I don’t see the sense in it from an engineering perspective.

    1. Depends on what we’re talking about. I think the crypto exchanges show that the bar to market entry is falling in the exchange space, at least technologically. The design patterns crypto is pioneering (basically: millions of people competing for mindshare on how to redesign property registers from scratch) may hold some promise to compete more cheaply with existing operators.

  4. I think you and I are in more agreement than is indicated by this post. My article was about expanded securitization of individual assets, not avoiding securities laws. I think you make some good points and there are some important messages here. I’ll post a more detailed response reconciling this article with mine after organizing some thoughts.

    1. Agreed, and wasn’t trying to throw shade at your post – it’s just that your post was the most coherent of the written explanations/applications of the token liquidity thesis to real assets so was the easiest to quote! Enjoyed reading it and I have no doubt I will also enjoy reading the follow up.

  5. Ah, the Big Short to strike again, this time w China and Russia outgaming the new In God We Trust? Were you a Financial Engineer? I worked w the FEs at Pimco on software engineering — saw the whole “see no evil, hear no evil” like you.

  6. Your post assumes a US centric view of the world. The vast majority of countries globally don’t have decent financial vehicles because the markets aren’t large enough. The daily trading volume of Bitcoin is larger than monthly trading volume of the Chilean stock market.

    Most countries in the world lack a decent equities market and decent debt options for businesses to grow. Building those products on top of block chain is relatively trivial in countries without that legal infrastructure.

    Also, options for banking for the unbanked, remittances transmission, creating financial products that could be invested in without minimums would allow people at the base of the pyramid to diversify their personal portfolio and put capital to work for them rather than stick out under a mattress.

    The future of blockchain apps is not in the US. It’s in emerging markets.

    1. The US’ long-arm jurisdiction is to be feared and respected. Wild west token issuances aren’t going to last much longer than they have already.

  7. What are the laws around bearer bonds? Seems like those kind of securities could be tokenized. What do you think of that potential application?

    1. Depends on the jurisdiction. Generally speaking bearer bonds fell out of favour in the 1980s because it was relatively easy to use them to launder money or evade taxes. Plus, the whole Nakatomi Plaza thing.

  8. All agreed on intrinsic versus speculative value. How about a hybrid model where real hard assets are under digital control? Imagine I prepay the rent for a wall of safe deposit boxes using a perpetual annuity. I put gold in each box. I hire a trusted auditing firm to verify the gold is in each box and generate cyrpto combination to lock the box. They never have direct access to the combo itself, they only sign off that the gold was in the box when the combo was generated. Combos one-time convertible back to the hard asset so they are now a direct proxy for it but in all other respects can flow digitally like tokens. Assuming all the above is possible with current tech ( I think it is ), what problems do you see with this system? Maybe the USA comes after you for minting money, but what if you do the whole thing in a safe (for you & the gold) foreign jurisdiction? Tokens are free to flow anywhere regardless of where they come from.

    If this works, then maybe can be extended from gold to any bearer asset. There are still some bearer bonds around, and some counties with bearer instruments for land ownership. You could maybe imagine a box where you keep the deed for an office complex. Box is controlled by collective action of token holders. Works in principal, and only technical and operational limitations prevent it from being practical – at least today?

  9. Question here: does the intrinsic/extrinsic token categorisation perfectly match the utility/investment one? What about tokens that provide usage rights to an otherwise perfectly traditional software system? Are those not both extrinsic in value and utility-driven? Thanks.

Comments are closed.