Thoughts on Cryptocurrency Lending

There’s two basic models, custodial and non-custodial. The vast majority of options are custodial, and that means a third party has your coins. This entails counterparty risk.

What is a reasonable return to lend out your Bitcoin? Is 6% good? Is 15%? It entirely depends on the counterparty risk. You’re not earning money for lending out BTC, you’re earning money for lending to a specific counterparty like BlockFi.

So, what’s a reasonable yield? Junk bonds in the US have typically paid 4-6% over treasury rates. For a risky start-up, uncollateralized loans would typically be more like 15-25% over treasury yields.

Short-term treasury yields are currently 2.5%, so I’d look for ~22%+ yield to lend my BTC to a start-up at a minimum. Now, this superficial analysis is unfair to Blockfi and others who would rightly argue this isn’t a purely uncollateralized loan.

On their website, Blockfi says they “typically lends crypto on overcollateralized terms.” The specifics matter a lot here. If that collateral was legally owed to you as a specific lender, these should probably be viewed as collateralized loans. They’re probably not.

If you as the lender have no legal right to specific collateral, then you’re simply a creditor to a single company, Blockfi (or whomever else), and you might not even be a senior creditor (for Blockfi, I believe you are a senior creditor, but this should be confirmed for each platform.)

So the right way to think about making a loan on a platform like this is very similar to if a crypto start-up asked you for a loan and offered you some interest to borrow your money, and promised to keep a pile of cash representing their general obligations to investors.

Non-custodial solutions are very different. With these, you may potentially have zero counterparty risk, rather you’d be subject to some sort of liquidation risk and protocol risk. If the market gaps lower, the proceeds from liquidating collateral might not pay you back.
Additionally, using things like multisig solutions for a lending platform have serious risk of bugs. Just how serious? Ask Parity, undoubted experts on ethereum smart contracts, who twice lost tremendous sums to bugs in the simplest type of multisig smart contract.
TDLR: The growth in crypto lending markets is awesome for the ecosystem (separate topic), but is extremely risky today however it’s done, and I think the yields are far too low to price that risk from my perspective.
Posted in Cryptocurrency | Tagged | 5 Comments

Some high level thoughts on cryptocurrency governance

This essay is an overview of some high level concepts in cryptocurrency governance, followed by some specific examples.

Governance: Control and Power

Governance is the process by which the members of a community or organization make decisions.  More than any other topic, the question of “who runs Bitcoin?” is the one that takes newcomers to cryptocurrency the longest to grasp.

Power in cryptocurrency is often indirect.  The executives of Coca-Cola have great control over the product they release, but ultimately serve the demands of customers or the company will go out of business.  Similarly, cryptocurrency miners with specialized hardware (ASICs) have strong incentives to support the long-term demands of cryptoasset buyers since their mining is only profitable if they mine an asset that someone else wants to buy.  When thinking about power in cryptocurrency, it’s useful to note both who has elements of direct control, and also who sets incentives that strongly influence those with direct control.  Those setting the incentives may have more power in most scenarios.  In the case of Bitcoin, miners collectively have control over which transactions to include in blocks, but the community of current and potential BTC buyers may have more power, since miners must cater to their desires to be profitable.

Developers are sometimes more ideologically motivated, but still want to develop a protocol that will be used.  Usage is supported by service providers like exchanges, hardware wallet manufacturers, and custodians.  Developers also want a secure protocol, which requires the support of miners or validators.  Developers often have powerful influence over changes to the protocol, but must serve the other stakeholders or their changes are unlikely to be adopted, or unlikely to result in a valuable cryptocurrency.

Most stakeholders benefit if the value of the crypto asset and size of the network is maximized, which requires support from other stakeholders. This produces a “Keynesian beauty contest”, in which stakeholders bet on what other stakeholders will support.  This results in lots of public debate in which parties attempt to convince other stakeholders that their own position has the strongest support.

Who are the stakeholders?  Anyone who contributes to a cryptocurrency’s value and benefits from that value.  This includes holders and buyers, developers, exchanges and other liquidity providers, some types of service providers, and those securing the network (miners, stakers, etc).  Stakeholder are aligned in wanting to maximize the value of the asset, but may differ on important specifics.  For example, miners generally want to maximize revenue from transaction fees, while service providers and users may want to minimize transaction fees.


All governance introduces incentives for ‘corruption’ and rent seeking behavior.  The more complex the governance model and the more it supports swift decision making and implementation, the greater the potential attack surface.  We can minimize the effects of bad governance and minimize the uncertainty that stems from unknown future governance decisions by choosing to have as little governance as possible, resulting in a relatively static protocol.

Alternatively, the lack of governance at the base layer may slow down valuable innovation.  Developer resources may not be well organized, and may not be directed to the needs of the broader community.  Additionally, even when there is strong consensus supporting a change, a lack of governance processes may make implementation of the agreed upon changes difficult.

Perhaps crypto assets should adopt the mindset of federalism and have minimal governance at the global protocol level, and more active governance for “local” protocols that can exist as layers on top of, or connected to, the global protocol, or for more specialized ‘local’ use cases?   There may be a lot of interesting ways to combine different governance models for various use cases, similar to how a US citizen is subject to US national governance, and can then choose which state and municipality governance they prefer.

A few types of governance

As a rough mental model, I think about governance as divided into the following categories, but many examples are some blend of the below:

  1. On-chain governance: decision making that is built into the protocol itself.
  2. Formal off-chain governance: decision making that occurs outside the protocol in a pre-specified way. Authority may be vested by law, by direct control (like the ability to directly edit a live smart contract), or by strong social consensus.
  3. Informal off-chain governance: decision making that occurs by weak community consensus.


Bitcoin’s governance is primarily informal off-chain.  Changes to the Bitcoin network follow a process supported by a weak form of community consensus.  Specific proposals usually begin life as conversations among developers on technical Bitcoin mailing lists or other mediums.  A developer will then produce a specific BIP (Bitcoin Improvement Proposal), a standardized format for proposing changes.  This will be subject to public peer review, and may then be integrated into a future release of the Core client.  Individual node operators and miners may choose to install the new client or not.

Bitcoin’s governance may best be described as “governance by exit.”  Changes to the consensus rules produce a chain fork.  There is a strong community norm that the protocol should not be changed without near unanimous support.  Governance in Bitcoin often means “exiting” the existing network and creating a new community as we saw with the birth of  “Bitcoin Cash” (BCH).

Bitcoin has made use of on-chain signaling mechanisms for miner activated and user activated soft forks (e.g. BIP 148).  Miners or users may update their Bitcoin clients to take action automatically if a critical threshold of mined Bitcoin blocks contain a signal for a change, or at a pre-specified time.

To learn more:


Ethereum’s governance is formal and informal off-chain.  Ethereum has a proposal process, EIP, similar to Bitcoin’s BIP process.  A key distinction is the strong leadership of founder Vitalik Buterin and the coordinating and financing role of the Ethereum foundation.  Vitalik and the Ethereum Foundation have no direct control, but have strong community support.

Ethereum has community norms supporting relatively frequent major protocol changes, and many in the community support the idea of at least a temporary “benevolent dictatorship” to coordinate development for faster innovation.

In July of 2016, Ethereum underwent a contentious hard fork in response to the DAO hack, that led to the creation of “Ethereum Classic.”  This changed the makeup of the Ethereum community as crypto investors supportive of the fork stayed, and some of those opposed left the community.  It is easier to “move” from one cryptocurrency to another than from one political regime to another.

This ease of movement may lead to a self-reinforcing governing dynamic in which a community becomes more similar over time via self-selection.  In the case of Ethereum, self-selection after the fork may have strengthened norms of immutability in the ethereum classic community and weakened that norm in ethereum.  This is not intended as a criticism of either community.

To learn more:


EOS governance is primarily on-chain and formal off-chain.

EOS utilizes on-chain voting by token holders voting for block producers (1 token, 1 vote.)  EOS also has a formal constitution governing behavior by block producers and empowering them to make off-chain governing decisions, like coordinating to freeze accounts that have engaged in criminal activity.

EOS is a new protocol (launched June, 2018) and one of the many governance experiments in cryptocurrency.  Ideally, token holders will vote for block producers who are “good actors” for the ecosystem, which could include building general tools like wallets and block explorers.  We’ve already seen evidence of cartels forming amongst block producers, and of “vote buying”, but also of block producers responding to incentives and investing to strengthen the ecosystem and increase token value.

As with any political system that involves voting, one challenge is complexity and voter knowledge.  It is difficult for EOS token holders to intelligently monitor, analyze, and then vote for a slate of block producers who are likely to be “good actors.”  This is something that fascinates me personally in both politics and cryptocurrency.  How can stakeholders make informed decision under antagonistic conditions?  In a conflict between stakeholders (or in an attack by external actors), we can expect disinformation campaigns and confusion in both cryptocurrency and politics.  This applies to all cryptocurrency governance models, but is exacerbated when stakeholders have to make frequent decisions about a wide range of topics with a diverse set of choices.

To learn more:


Storecoin is an early stage project that aims to primarily make use of on-chain and formal off-chain governance.  It is loosely modeled after the US representative democracy with an emphasis on the checks and balances reflected in the various branches of government.  Storecoin aims to implement a “one entity one vote model”, in contrast to plutocratic proof of stake systems that are “one token one vote.”

Like the US government, Storecoin aims to balance responsiveness with steadfastness; change should be possible, but difficult in this community’s view.  Additionally, different types of responsibilities (e.g. validation, network security) are managed by specialists, but subject to review by other governance ‘branches’, aiming to balance the benefits of specialization while limiting the potential for abuse of power.

To learn more:



I view all cryptocurrency governance models as experiments.  We’re just now leaving the first decade of live crypto networks and still have much to learn from trial and error.  We may also find that some forms of governance perform better in different contexts.  Perhaps on-chain governance makes sense for some use cases and off-chain governance is preferable for others?

My intuition for cryptocurrency is the same as for politics.  The more diverse the stakeholders, the simpler we want the governance.  In politics, this suggests something like the federalism model we see globally today with minimal global governance, stronger national governance, and stronger still local governance.  The smaller and more homogenous the set of stakeholders, the more likely a strong governance model is to produce more good and less harm.

Imagine if we had a global democracy and voters in Asia could assert their will over voters in the rest of the world combined?  We’d likely want to minimize how much control the majority could assert over the minority.  In contrast, a stronger form of governance in a small town might be fine, since the residents are more likely to be closely aligned in their values and similar in their resources.

I look forward to learning from these real-time experiments.

Disclosures: BlockTower is a cryptocurrency investment firm that may take long or short positions in crypto assets, and our positioning may change suddenly.  At the time of publication, we have long positions in each of the assets discussed.  Nothing in this essay is intended as investment advice.

Posted in Cryptocurrency | 3 Comments

Early stage valuations fall to earth

I initially wrote the following as an internal memo to our team this morning, but thought it might be interesting to the broader crypto world.

We’re seeing something play out in real time (but slow motion) that we’ve long expected.  The crazy valuations of early stage projects are falling to earth.

6 months ago, a decent team with an idea on a napkin were raising at $100m+ valuations, sometimes $250m+.  This happened because investors mistakenly extrapolated the amazing returns of 2017 ICO investing forward.

The average return on a pre-ICO or even ICO investment prior to around April 2017 was outrageously high.  The projects from before this period were generally valued at <$10m pre-ICO and <$40m at ICO time.  They were generally high quality projects, since it wasn’t easy to raise $20m in an ICO prior to 2017.  The ‘get rich quick’ entrepreneurs mostly weren’t on the scene yet, so the ratio of good projects to bad was relatively high, and the valuation entry points were relatively low.  It’s far easier to earn a 5x on a $10m valuation than on a $100m valuation.

Combined with the general market run-up in the second half of 2017, every one at every stage had a chance to profit.  The earliest pre-ICO investor got a mark up to the later stage pre-ICO investor, who got a mark up to the ICO investor, and all were able to exit on exchange listing to an exchange buyer, who often also profited.  In the last 6 months, the exchange buyer has consistently been losing badly, so they’ve mostly stopped buying.  This means that the late stage pre-ICO investor also started often losing.  Until now, the early stage pre-ICO investor was still winning, but it’s like dominos, and we’re nearing that last domino falling as well.

As a result, many investors have been more public about their aversion to investing at ‘crazy’ valuations, even when offered deep discounts to those crazy valuations, since those deep discounts still look overpriced and there’s likely to be no one to sell to in the near future.  For example, a 75% discount to a $200m valuation is still $50m, which is still aggressive for an early stage project with limited usage, network effects, and switching costs.  Sometimes I find myself looking at a project thinking, “I’d never invest at in this at $50m, but maybe it’s interesting at $50m if that’s a 75% discount to the $200m valuation others are or will pay?  That line of thinking isn’t inherently irrational, but it is dangerous.  It’s a trader’s mentality, not an investor’s mentality, and it relies on market timing, not investment underwriting.

These things play out in slow motion, because both investors and projects are reluctant to realize mark-downs, and the lack of a liquid exchange price means they can fool themselves.  One way this plays out is what we’re seeing – much smaller raises at the high valuations.  For example, consider a project that had an initial angel round raising $2m at $10m, and now raising $10m at $200m.  If they find the $10m of investment at the higher, those early investors think they have $40m worth of tokens.  But while the $200m valuation is ‘fair’ from an accounting perspective, it’s not real from an economic perspective – there’s only $10m worth of liquidity at that price, so holders of the $40m worth of tokens can’t exit at that level.  If even half of them tried, it would likely crash the valuation 50%+, maybe much more.

As a result, there’s been a long lag in pre-ICO valuations coming down to earth to match the public markets and changing investor sentiment.

This may produce some attractive investments in the near future if investors ‘panic’ and look for OTC liquidity to exit their pre-ICO investments, but since many of these valuations have so far to fall, and investors are very reluctant to realize 80%+ losses, this will likely both take some time to play out, and willing sellers at attractive levels may be scarce.

This market dynamic is not a surprise to most of the investors in the space – many understood that they were playing musical chairs, and just hoping to be able to find a chair before the music stopped.  Timing the music is far harder than identifying the game.

The very best projects will survive their valuation write-downs and ultimately thrive.  And hopefully we’ll see new projects come to market at reasonable levels, 1/5 the valuations of the recent batch soon that will provide attractive investment opportunities.

Timing these cycles is challenging, and the illiquid nature of the assets means that to successfully time them, you have to anticipate the cycle by 3+ months (maybe 6+ months now given longer lock-ups.)  Doing this at the margin makes sense (e.g. deploy more capital in bear markets, less in bull markets), but I think it’s generally best to do so as shifts to an underlying consistent investment strategy.  In other words, rather than deploying nothing for 9 months, and then racing to deploy, I think it makes sense to have an underlying “slow and steady” approach to allocating capital to top decile projects each quarter, and to maybe cut that pace in half when valuations seem high, and to double it when valuations seem cheap.  To the extent an investor wants to further time the general market, this can be done by increasing or decreasing liquid cryptocurrency exposure.

That’s the end of the internal memo.  I’m getting asked by a lot of projects what this market dynamic means for them.  There’s no “one size fits all” advice I can give – it depends on your project, the current size of your treasury relative to your roadmap, and your ambitions.  But there are a few specific suggestions I can give. 

First – be realistic.  A down round (aka raising at a lower valuation than your last raise) is optically bad and you may reasonably choose to avoid it by simply not raising at all if your treasury is sufficient.  But…if your project needs that influx of capital to thrive, swallow your pride.  Whether you exchange $5m for 10% or 30% of your project’s tokens is trivial relative to maximizing the odds of your project surviving and succeeding.  30% of 0 is still 0.  And 10% of $1 billion is still $100m.  Momentum and optics matter, but they ultimately matter far less than actually building a product/service/network that offers value to users.

Second – think about what you really need.  Sure, it might’ve been nice to build a $50m warchest, but what do you really need to execute on your vision?  Think critically about your roadmap and what you need financially to execute on it.  

Third – love your investors.  A great many investment agreements in our industry are on questionable legal and regulatory footing.  Some are explicitly ‘donations’, others are utility token agreements with lengthy contracts that exist in a regulatory gray area.  Many ‘good actor’ projects likely violated legal and regulatory fine print.   JP Morgan and Credit Suisse routinely violate contracts with one another over delivery of treasury bonds for example.  This *could* result in a lawsuit, but almost never does – rather they resolve things amicably instead of racking up huge legal bills, incurring negative publicity, and damaging relationships.  When the market is up – no one looks too closely at contracts or thinks about lawsuits.  When markets are down, this becomes an issue.  Recognizing that early stage investing is largely based on trust and that this industry is largely one based on relationships, projects should be careful to treat investors as valued partners in all regards.  One example – many projects are currently re-writing their investment agreements to better comply with current SEC guidance.  Such re-writes can be viewed as exploitative by investors or as necessary for the project’s success.  Pivots, contract re-writes, and corporate restructurings may be important for a project’s success which every investor desires for their portfolio holdings.  Project leaders must communicate both the substance and intent of such changes.  This is important to ensure that investors view such changes as for their own benefit as well as the welfare of the project.  





Posted in Cryptocurrency | 15 Comments

Education – University of Chicago Alumni Event

BT Educationpdf

Posted in Cryptocurrency | 2 Comments