Investment Analysis: MakerDAO

BlockSpace Solutions Inc.
20 min readMay 8, 2018

--

By Aviv Milner & Victor Hogrefe

Abstract

Maker (MKR) is one of the earliest implementations of a tradable company on the blockchain. The Maker product, a type-II stable coin (the Dai) smart contract platform and user interface is digital, auditable, decentralized and separate from the volatility of the MKR price. Although the infrastructure of the Maker/Dai system is a breakthrough for crypto assets, the success of the Dai stable coin opens the door for a community standard for tokenization of company shares, beckoning legacy stocks into the blockchain space. This report aims to outline the business infrastructure of the Maker DAO, the relationship between product success and Maker valuation, and the risk vectors associated with the business model.

Maker Infrastructure

Maker DAO, Dai and Ethereum

The Maker DAO, Dai and smart contract interface is a locally closed system to the Ethereum Virtual Machine. This relationship can be visualized as follows:

In this diagram, the ethereum virtual machine (EVM) acts as the blockchain platform, where MKR and DAI are both ERC-20 standard tokens that interact with each other on the blockchain. The roles of these three tokens is as follows:

  1. Ethereum (ETH)

Solves distributed consensus (BFT) and secures the blockchain

Ether token acts as collateral for a CDP

Ether token is spent in order to interact with contracts and move funds

Ethereum platform allows other tokens to potentially interact with Maker/Dai

2. Maker DAO (MKR)

Governance token allows user to vote on system settings

MKR can act as a fund raising mechanism for the Maker team

Token is burned as part of the fee for settling a CDP

MKR token acts as a lender of last resort in the event of insolvency in Dai

3. Stable Dai (DAI)

Stable Coin pegged to the USD, that can be traded on any exchange

Token that is instantiated by users in the form of a CDP

Token required to pay-off the debt in a CDP, in order to reclaim Ether

Token used to bite undercollateralized CDPs by third parties

Maker/Ether/Dai User Interface

Whereas Type-I and Type-III StableCoins allow for the governance arm to instantiate new stable coins, the supply and stabilization of Dai is entirely left to third parties who are free to lock their own Ether in a Collateralized Debt Position (CDP) and instantiate and withdraw Dai insofar as the CPD contract is not below a defined threshold. The entire supply of Dai, which currently is just over 27 million, is the sum of the one thousand or so currently open CDP smart contracts held by voluntary users. The collateral of the outstanding Dai is conveniently held in a single pooled Ether contract (PETH) and includes more than 160,000 Ether. In order to help users evaluate token metrics, audit and assess risk from data recorded on the blockchain, the Maker team has created three valuable UI tools:

  1. A User Interface for Interacting with Smart Contracts (https://dai.makerdao.com/)
  2. A Systems Metrics Overview for Assessing Risk (https://mkr.tools/)
  3. A Decentralized Exchange for Price Discovery (https://oasisdex.com/)

User Interface Images

https://dai.makerdao.com/

This site is used to interact with the Maker/Dai smart contract. Users can see the total pooled Ether, the total Dai balance and the MKR that has been burnt as a fee to the shareholders. Users can also see their own balance via a metamask account.

https://oasisdex.com/

Users can trade any of the three tokens (MKR,DAI,ETH) by placing bids on the blockchain in the form of smart contracts on this decentralized exchange, often called a DEX. Although a decentralized exchange can be more cumbersome and expensive when compared to the centralized counterparts, it solves three crucial problems — reducing central points of failure, introducing an oracle mechanism and permitting the aforementioned bailor of last resort.

https://mkr.tools/

Users can check the market price of MKR, the total collateral of existing CDPs (the pool of Ether) and their standing as well as the total number of Dai tokens that have been issued. Currently a total of 1620 CDPs have been created, of which almost 500 have been closed. Below is a distribution of Dai creation by CDP in the form of a pie chart.

Maker Valuation Metrics

CDPs — Creating New Dai

The product of the MakerDAO is the smart contract that enables users to instantiate stable Dai by locking away Ether, in exchange for a small fee. The MakerDAO sets a liquidation ratio that instructs users on the maximum amount of Dai they are allowed to create for every Dollar of collateral they have locked away. Since users are incentivized to ensure that collateral remains above this threshold, most will have as much as twice the collateral needed. It is for this reason that we observe an average of 400% collateral of PETH when a liquidation floor is set at 150%.

As more and more people hear about the MakerDAO and the Dai stable coin, those who are sitting on Ether may find it wise to refrain from selling their holdings for liquidity and instead use the Maker/Dai business as a personal bank from which they can get a loan. In a (crypto) commodities bull market this can mean that all CDP owners would have an increased ability to spend (or loan) more money, while in a bear market (as has been the case for four months) users will tend to add additional collateral to their CDP and refrain from withdrawing large sums. From this perspective, most users (most of the time) will either increase their collateral or increase the money supply with volatility in commodity prices, suggesting that the Dai supply may expand faster with higher volatility.

Closing a CDP — Burning MKR

In order to tie the success and adoption of Dai with the price of MKR governance token, the MakerDAO sets a settlement governance fee in the form of an annual percentage. As of now, this fee is 0.5% of outstanding Dai per year, and is required to retrieve the collateral locked away. So if a user has created 100,000 Dai and maintained the CDP for 6 months, then a settlement would require $250 USD in fees. Since OasisDEX provides the blockchain with a price feed of MKR tokens, this fee is repaid by calculating the amount of MKR of equal value and having the user burn it in the global burn address. As of writing, 8 of the 1 million MKR have been burned in this address.

The most important thing to understand from this model is that the revenue generated from fees is not dependant on the price of MKR. If the price of MKR is very high, users will burn smaller amounts and vice versa.

Modeling MKR to Circulating Dai

The price of the MKR token, like any other stock,replies to forces of supply and demand. The MKR burn model acts as a supply contracting mechanism and can be modeled with the following parameters:

Let PM(t) denote the price of Maker at time t, in years.

Let FG denote the governance fee, and let us assume that it is constant at 0.5%

Let D(t) denote the amount of circulating Dai at time t, in years

Let M(t) denote the amount of circulating Maker at time t, in years

Let dM(t)dt denote the change in Maker supply as a result of burn, in MKR/year

Let b denote the (constant) annual burn rate of MKR as a percentage/year

Let c denote the (constant) rate of exponential growth in circulating Dai

First observe that the amount of MKR burned is only dependant on three parameters: the governance fee, circulating supply of Dai and the price of Maker. We can describe their relationship like this:

Since we want to describe the price of Maker as a function of all other parameters, we can rewrite the above function by rearranging variables:

Next, we want to describe the growth of circulation Dai over time. We will need to make two assumptions in order to get a clearer picture:

  1. Circulating supply of Dai is one billion at time zero.
  2. The growth rate of circulating Dai will follow exponential, rather than linear, growth.

The first assumption is justified on the grounds that Dai becomes a serious competitor in the stable coin market, which as of writing totals more than 2.5 Billion USD in circulation, of which USDT makes up 2.3 Billion USD. All investment models in this paper are based on the success of Dai as a competitor in this market.

The second assumption has to do with the nature of a positive feedback loop — as more Dai enters circulation, it allows for more exchanges and users to have Dai for the purpose of commerce. As more exchanges, users and businesses get more onramps to Dai, the demand

will rise and continue to put upward pressure on the creation of new Dai. Here is the model:

Here, “c” is some constant that defines the amount of time before the circulating Dai doubles. The larger the number, the more time it will take for the circulating Dai to double, and vice versa.

Next, we need function that describes the total current supply of Maker tokens, we can use the fact that the Maker team has instantiated 1 million tokens at time zero, and the supply is burned according to function (1).

Now we can simplify the Maker supply function by assuming that the rate of burn, that is, the yearly percentage of Maker that is being burned from the total supply- is constant. We can then model maker supply over time like so:

Now that we have a reasonable function to describe the supply of MKR over time by assuming that burn is constant, and a justified equation for modeling the circulating supply of Dai, we can take the derivative of price and measure how price changes over time. We take equation (2) and plug in (3) and (7*) and take the derivative:

We then replace all of the constants with their known values and we are left with this:

However, what the above equation describes is the yearly predicted change in the price of MKR in Dollars, as a function of time, Dai growth and burn rate. A more useful result for predicting price would be to observe the return-on-investment, or ROI. We take the derivative of price and integrate over a year of time to measure price growth, and then divide this by the price of MKR at the beginning of the year. We get this:

We make an interesting observation with (9*), namely that annual ROI is independent of time if we make the assumption that circulating Dai grows exponentially and burn is constant. Instead, the only factors that influence ROI are the percentage burn rate, and the coefficient of growth in the Dai supply. This result is not entirely unintuitive, and is what makes the MKR model so special. Now we need to consider what possible values for b are plausible.

We can partition values of b into 4 rough ranges:

The first scenario describes an extraordinarily slow burn rate, and this leads to unreasonably high valuations of the MakerDAO, with an initial share price of $2,000-$5,000 at time zero, and a $10,000 valuation of the share within 1.7 to 3.8 years. We can take the burn rate as a sort of price-to-earnings ratio that we typically use to value conventional stocks, with that in mind, we consider this rate of burn to be a “speculative” valuation of the MKR token. Below is visual representation of price spread given a speculative valuation. You can see that if the burn rate remained this low, the value of the MakerDAO would exceed 20 Billion Dollars within 3.5 to 7.6 years, which would place the MakerDAO as a large cap company among traditional publicly traded companies, a result we consider overly optimistic.

Constants & Assumptions:

  1. Annual Rate of Burn, constant: 0.1%
  2. Exponential Growth Rate of Dai, doubling 1.7–3.8 Years
  3. M(0), D(0) and F_g as stated above (4) & (6)

Now we consider a more accelerated burn rate, subject to (ii). We believe that this range provides us with a reasonable, or “expected” valuation of the MakerDAO. We plot two distinct curves for three burn rates inside of this range, giving us six curves and three shaded regions:

Here we get much more conservative, although still optimistic result. The price of a MKR share is valued between $500 and $2000 at time zero, and hits $5000 sometime in the next 2.3 to 8.6 years, as the circulating supply of Dai steadily increases. The green shading, which displays the lowest burn rate of 0.25%, suggests that the price of a MakerDAO share would hit $10,000 within 4.0 to 8.7 years. The shaded area in black uses a burn rate of 0.5 % annually, while the are in red uses a constant 1% burn rate.

Next, we plot the same conditions but for a higher burn rate, which suggests the price of MKR is undervalued. Here we observe an initial value of MKR between $150 and $250.

Perhaps the most interesting result that demonstrates the unique properties of the MKR burn-to-use model is the relationship between the annual expected return on investment, and the burn rate. Below is a graph of ROI (as a decimal) plotted against burn rate (as a decimal). We took four possible growth rates of Dai — no growth (c ), doubling every 7 (c=10), 3.5 (c=5) and 2 (c=3)years, and this gives us three distinct regions.

Note that as more MKR is being burned (i.e. as we move along the x-axis towards the right) the more the annual ROI increases. This is an expected result, since the more MKR that is being burned, the more the remaining un-burnt supply grows in value. Furthermore, note that even if the circulating supply of Dai remains constant at one billion units, as modeled in the bottom green dashed line, the ROI grows almost linearly with burn rate, where a 1% burn rate is equivalent to a 1% ROI, and a 3% burn rate is met with a 3% ROI.

This model gives us some excellent insight into the expected burn rate of MKR based on the expected ROI. We can compare the ROI of investing into MKR against less risky ventures both within and outside of the blockchain economy and consider what an equilibrium point or range for investors would be enticing. Investing national currency (like USD) into bonds or a savings account give investors a 1–3% nominal ROI with almost no risk, whereas a typical large-cap publicly tradable stock would offer 7% real ROI, although bears more risk. Given that MKR is both a small cap company, with many possible risks and much uncertainty (as is the case with emerging tech) we expect a nominal ROI equilibrium between 12% and 28%.

Conclusion

Deciding whether to invest into MKR requires answering the following questions:

  1. What is the expected growth rate of circulating Dai this year?
  2. What is the expected burn rate this year, given the current price and fees?
  3. Does this burn rate produce a worthwhile ROI, given all of the risks?

Using the above models, we believe one can arrive at a reasoned answer to these questions.

Attack Vectors and Investment Risks

Current Type II Competitors to the Dai (BitShares)

If Type II stable coins like the Dai overtake Type I coins (namely USDT) in circulating supply and volume, a possible risk for MKR holders is that a competitor in the Type II stable coin space has a larger following and adoption rate. Currently, only one other project — BitShares — is offering a collateral-backed stable coin. Their decentralized exchange (BitShares) allows users to collateralize the BitShares governance token and instantiate the following stable coin assets (dubbed “smart coins”) on the platform:

It should be clear that the platform allows more variety in stable coins than MakerDAO, although nearly all of the stable coin supply is from the Chinese Yuan and the US Dollar. As of writing, Dai has already passed BitUSD and sits as the #1 decentralized stable coin pegged to the USD. If we look at the historical price of the BitUSD (which attempts to peg itself to the USD) we see that BitShares has not yet succeeded at properly incentivizing outside actors to keep the price pegged, and the BitUSD token has continually swung between $0.90 and $1.10.

For this reason, we don’t think that BitShares will be a serious competitor to the Dai, and instead turn our attention to projects that are soon to be released.

Future Type II Competitors to the Dai

Several projects have decided to implement Type II stable coins, although this is by no means an exhaustive list, these groups have attracted the most attention:

  1. Havven
  2. SweetBridge
  3. Augmint

Havven, much like BitShares, allows users to create stable coins by collateralizing them with the Havven governance tokens. This system benefits token holders in the short run, as the demand for governance token increases to meet the stable coin demand, but in the long run presents a serious solvency risk for those who hold the native stable coins. We believe that collateralizing stable coins with independent assets like Ether will prove more successful and reduce the risk of insolvency that these proposed new models will inevitably face. In addition, users are more readily convinced of the stability of Dai, when told that it is backed by Ether, more so than they would be if it were backed by Havven governance shares.

Sweetbridge presents the most plausible successful competitor to the Dai, with one serious advantage — the stable coins will be backed by a larger variety of commodities — including Bitcoin. However, the project is bit over ambitious, and it claims to let users put physical assets like homes and cars as collateral, without demonstrating a firm solution to the oracle problem.

All Type II stable coins that may serve as competitors to the Dai stable coin in the future also share a common setback — since the companies themselves do not control the money supply, it will take the slow adoption and acceptance of the wider crypto community to produce growth by adding collateral. This means that Type II competitors will be easy to spot and are unlikely to overtake Dai very quickly. Below is the total Ether collateralized (PETH) in the Maker system:

Future Type I & III Competitors to the Dai

The landscape of stable coin implementations is diverse, as we have discussed in our previous article “Introduction to Stable Coins”. We argue that in the short to medium term (2–10 years) users will tend to exit Type I models which incur counterparty risk, and will be too risk-averse to jump into a Type III stable coin, until more research and adoption is present. Please read our previous publications for more information.

Adoption Ceiling

All of the price models above have only considered a world where the Dai stable coin is already at one billion circulating units, with various degrees of exponential growth. There is, however, an argument that can be made that Dai will never hit this target due to an adoption ceiling. For example, it may be that few major exchanges, ATMs and businesses adopt or recognize the Dai stable coin, and the adoption hits a point of niche saturation — limited to the smaller subset of use cases. We believe that this is unlikely for 3 reasons:

  1. The demand for a stable coin is enormous and an analytics show that it is increasing.
  2. MakerDAO intends to release multi-collateral Dai in the summer of 2018, allowing a much broader consumer base by including many ERC-20 tokens as collateral types.
  3. A growing number of decentralized applications (DAPPs) including exchanges, betting and insurance markets have emerged, suggesting that a decentralized stable coin has an expanding market.

Dai Collapse, MakerDAO Governance Failure

The lenders of last resort — if all else fails — are the shareholders of MKR. If the incentives placed by the MKR shareholders do not entice enough third party liquidations of bad debts, then new MKR is printed (the only instance when MKR is printed after the first day) and is sold on the Decentralized exchange for Dai, which is then used to settle bad debt. This scenario would likely only be triggered by a combination of circumstances:

  1. A “black swan” event that causes the price of Ether to drop very quickly in a short amount of time. This would likely have to be as much as a 30–50% drop of the price within a 6–12 hours.
  2. An average collateral level of less than 300% before the black swan event. Recall that the current overcollateralization ratio is set at 150%. If the collateral level is above the 300% mark, it is unlikely that even a black swan event could trigger a massive liquidation.
  3. Too few independent “keepers” (users who liquidate bad debt for a profit) that cannot provide sufficient liquidity against the bad debts.

We do not believe that a combination of such events are very likely to occur.

Weaknesses in the Ethereum Infrastructure

Maker and Dai both rely heavily on the growing adoption and success of the Ethereum Virtual Machine, the community and the totality of all projects on the blockchain. As a result, any risks to the Ethereum Blockchain are also simultaneously risks for any holder of MKR. Below are just a few of the possible risks:

Bugs/Hacks

If Ethereum is discovered to have a critical bug in the coding language, as was discovered recently with the creation of ERC-20 tokens, the bug may either trigger a black swan event as described above, or cause the code of the MakerDAO smart contract to be corrupted. This concern can only be addresses with strict audits of the Maker smart contract and added scrutiny towards the Ethereum ERC standards. We think that this concern is not nearly as improbable as the Dai collapse, although in most circumstances these bugs would not be devastating to the Maker business.

Contentious forks

In the event of a contentious hard fork, that is, a hard fork where a majority chain is not immediately discovered, or where the minority chain retains at least 20% of the hashrate of the majority chain, the Maker model would be in jeopardy. Because the MakerDAO cannot exist simultaneously on two parallel chains, the resulting hard fork would cause a sizeable drop in the value of the Ether in collateral as well as a reduced consumer base (those who have left the majority chain) . If the hard fork could split the community in half, this would likely cause a huge concern in the MakerDAO, as the governance token holders would not be able to firmly ascertain which forked chain is the majority.

Although the probability of minority and super-minority forks is very high, these will not pose a serious concern to the MakerDAO, as they can very well be ignored with only a minor loss to the ecosystem. A more contentious hard fork is highly improbable.

New Generation of Blockchains Overtake Ethereum

The most recent concern that plagues the Ethereum infrastructure is the increasing growth of new generations of smart contract blockchain platforms that employ scaling technologies that could appeal to a much larger community and siphon developers and consumers from the Ethereum network. As of right now, Ethereum maintains more than 90% of the top 100 tokens in the blockchain space, with Bitcoin, Neo & Cardano each holding a few tokens each. With the release of EOS in just a few weeks, we are likely to see more developers branch out onto different platforms.

Scalability Issues

The EVM currently has the capacity of no more than 1.4 million unique Transactions per day. In the beginning of 2018 this limit was reached, as is clearly visible in the graph below. When throughput was limited, the price of executing transactions began to rise very rapidly, and the average transaction (21000 gas/txn) cost $3-$4. This would be even more expensive for users who want to send Dai, as a transaction requires 78.2% more gas (37426 gas/txn). If Ethereum is unable to solve it’s scaling issues, Dai would be limited to high volume transactions and would not be suitable for merchant adoption.

Concluding Thoughts & Future Projections

In light of the 2013/2014 Bitcoin bubble, many programmers and economists in the Blockchain ecosystem began to ponder of a solution to price volatility. As Bitcoin tumbled from $1150 USD to $150 USD over the next two years, a large number of Bitcoin advocats lost motivation and hope. In March of 2015, just as the first centralized stable coin — Tether — made it’s initial release, Rune Christensen set out to solve the hard problem of a decentralized value-stable cryptocurrency and created the MakerDAO. In his interview on the Epicenter Bitcoin Podcast one year later, in a quiet and humble tone, Rune told the interviewer “The business of money is the best business in the World”.

The MakerDAO project is unique in many ways, our hope is that this review has demonstrated that. In a time where an incomplete and untested project can get millions in ICO revenue for prematurely releasing unfinished projects onto the market, Maker stands out above the crowd for a slower, more cautious and meticulous approach. You won’t find pushy advertisements on social Media that promise unreasonable returns for investors in the Maker governance token, and unless you have been introduced to stable coins, or know the project developers directly, you likely haven’t heard about MKR yet.

Every morning, as we go over the various analytics of dozens of projects, we take a moment to check on the market price of the Dai stable coin. Every morning we stare in awe and disbelief as once again the Maker team has proven themselves, proven that programmable money is without limits, proven that finance is governed by laws of mathematics, and we observe — the Dai is stable.

As of writing this paper, the price of the MKR governance token sits at $1,000 USD, more or less, and the circulating Dai has continued to rise, with another million Dai being created by voluntary users roughly every four days. We believe that Maker will attract new users and achieve a 1 Billion circulating supply of stable Dai within 24 months.

The future looks promising for Maker, with a list of new partnerships recently announced as well as the second version of the Dai (the Multi-Collateral Dai) scheduled for release this summer we expect that it will not be long before the Maker project is known in the rest of the community.

___________________________________________________________________

This article was written by Aviv Milner and Victor Hogrefe of BlockSpace Solutions, with help from Mechanical Engineer and Blockchain Enthusiast Guowei “G” Huang of the University of British Columbia. Guowei is a friend and voluntary contributor to BlockSpace Solutions, and his help with the modeling and valuation metrics of MKR was critical.

--

--