Our Network: Issue #15

Updates on Bitcoin, Tezos, Cosmos, and Decred.

Editor’s Note

Our Network is live on Gitcoin Grants. We are grateful for everyone who has donated so far who believes in our mission of free crypto analytics. Right now 1 Dai donations are being matched with 33 Dai, so it doesn’t take much to move the needle for us.

Click here to learn more or make a donation, and thanks everyone for your support!


This week our contributors cover the following cryptocurrencies:

  • Bitcoin

  • Tezos

  • Cosmos

  • Decred

📌 Bitcoin

Contributor: Nate Maddrey, Research Analyst at Coin Metrics

  • Bitcoin has rebounded relatively well after its precipitous price drop from about $7,600 to under $5,000 on March 12th.

    Although hash rate temporarily plummeted along with the price, it appears to have started growing again after hitting a local bottom. Sudden hash rate drops can be painful in the short term but can be healthy in the long term. Less efficient miners are often knocked out of the market after large price drops, but are typically replaced by more efficient miners over time.

    The below chart shows Bitcoin estimated hash rate over the last year, smoothed using a seven day rolling average.

  • The percentage of Bitcoin untouched for at least 30 days (in other words, the percentage of the total Bitcoin supply that has not moved on-chain as part of a transaction in 30 days or longer) dropped from about 89.4% on March 10th to a low of 86.0% on March 20th. December 2018 was the last time the percentage of supply untouched for at least 30 days dipped below 86.4%.

  • But the percentage of Bitcoin untouched for at least 2 years has actually increased since the crash, from 41.96% on March 3rd to 42.50% on April 1st. This signals that longer-term holders have (mostly) not been panic selling, and that recent price movements are likely mostly driven by shorter-term holders.

  • Bitcoin market cap briefly dipped below realized cap on March 12th through 18th.

    Realized capitalization is calculated by valuing each unit of supply at the price it last moved on-chain (i.e. the last time it was transacted). This is in contrast to traditional market capitalization which values each unit of supply uniformly at the current market price.

    Sudden changes in realized cap can signal that older coins are suddenly being moved. If market cap drops below realized cap and does not recover, it could signal that investors are increasingly underwater and unable to sell their coins without taking a loss.

  • But if market cap drops below realized cap and then later bounces back above it, it could be a sign that the market was oversold during that period. Historically, the periods where market cap has dropped below realized cap have been some of the best times to buy Bitcoin.

📌 Tezos

Contributor: Alexander Eichhorn, Founder at Blockwatch Data

  • Protocol Upgrade: With world-wide spotlight elsewhere, the Tezos community quietly focused on hardening its base layer with a third protocol upgrade (codename Carthage) that went live on March 5. Voted for by 46% of all bakers and backed by 72% of network stake the participation rate was slightly down from earlier votes (46%/84% in Athens, 45%/83% in Babylon). Like expected for a bugfix release the on-chain upgrade went much smoother this time. >85% of all bakers had upgraded their software ahead of time and baker churn was limited to <3% (down from 10% in earlier upgrades). Note that a protocol upgrade in Tezos is not backwards compatible like in Bitcoin, so all infrastructure including indexers, wallets and dapps need to follow.

  • Staking: Custodial staking at the top 3 exchanges Coinbase, Binance and Kraken continued its strong 2-digit growth trend during March while smaller exchanges saw almost no change. On April 1st a whopping 65.7M tez (10% of all staked coins, 7.9% of total supply) was staked at Coinbase alone. Together the top 5 custodial exchanges manage 136M tez (21% staking supply, 16.4% total supply). At these numbers monthly income from staking fees for Coinbase should be around 82.5k tez ($140k USD).

  • Growth: Month over month growth slowed down during March to about +6% (25k accounts) with the strongest growth in the account range between 100 and 1000 tez (+2k, +20%). With custodial staking on the rise centralization grows and the top 1k accounts now hold 64% of total supply with the Gini index for balances > 1 tez sitting at 0.9647. Large accounts with balances between 10k and 1M were in decline, while exchanges and a few whales are continuing to accumulate. 6% of total supply (50M tez) was on the move last month while 77% of supply haven't moved since 3 months ago.

  • Staking: 3.3M new tez were minted last month, network wide staking ratio dipped back 1.2% to 78.5%, staking rewards are at 6.07% annually, future expected yield above inflation increased slightly to 1.3% (note: we changed the way we calculate yield since last newsletter). At first sight it seems as if the downtrend in price from mid February heights could have caused delegators and bakers to liquidate some of their stake.

    A closer look at the data reveals, however, that the opposite happened. Total delegated supply (the part of staking supply that is not owned by bakers) actually reached an ATH of 522M towards end of March. Reason for the network staking dip was baker churn. Around 3% (10M staking supply owned by bakers and another 7M delegated to them) went inactive during March.

  • Adoption: With nearing end of live for the Babylon testnet, dev teams are slowly migrating over to Carthagenet. Both testnets combined we see a continued growth in developer activity (+10% calls) and a strong increase in contract diversity (+34%). Clusters of activity on testnets appear to be around new token designs and STO product demos which is very encouraging.

    Smart contract activity on mainnet is still stagnant in terms of traffic and gas. However, real value in blockchains does not necessarily hide behind lots of traffic. StakerDAO actually passed their 3rd on-chain vote, and the first STO token backed by a real-world asset (a french music pavilion) was released by Equisafe in early March. Despite its value and fame, calls to such a contract are extremely rare so real impact for Tezos and the blockchain industry is hard to quantify using on-chain data alone.

📌 Cosmos

Contributor: Chjango Unchained, Director of Community at Cosmos

  • Inter-blockchain Communication (IBC): IBC is the flagship digital asset and data transfer protocol for communicating value across bespoke blockchains. When IBC is launched in a few months' time, it will usher in the era of an Internet of Blockchains—a watershed moment that will benefit the entire industry because it would bring about what I liken to the Industrial Revolution for the macro crypto economy. Today, IBC is nearing production-readiness and is being stress-tested in a testnet environment.

    Source: https://ibc-relay.herokuapp.com/

    This visualizer shows you the number of established IBC connections there are between any two IBC-enabled Cosmos SDK blockchain which, once connected, start relaying IBC packets between one another.

    There are currently 8 separate Cosmos SDK-based testnets (e.g. Dokia, Ping IBC, Iris, etc.) that have established IBC connections to each other's testnets and are transporting IBC messages to one another across chains. Image 2 isolates Dokia's testnet chain and shows that it has 25 connections established to allow it to "trade" data/value/etc. with the economies of the other chains.

    Metaphorically, it helps to think of each IBC connection as a highway or even a trade route, a relayer as any untrusted vehicle (you don't need to trust the relayer in order to be sure that the contents in the package were not tampered with) that delivers payloads (e.g. smart contract calls, tokens, location data, etc.) from chain A to chain B, and finally, a bespoke chain as a sovereign economy that wishes to embark in trade with another chain's economy. Put it all together and what do you get? Globalization as the end-state for token economies.

  • Game of Zones (GoZ): Incentivized testnets were a thing pioneered by Cosmos in 2018. Game of Stakes (GoS) was the first of its kind which subsequently spurred the creation of at least 11 other incentivized testnets since its inception, including that of Cardano, Solana, Near, Oasis, Irisnet, Celo, Matic, Icon, Lino, Coda, and Nucypher, to name a few.

    Game of Stakes was necessary for preparing for the Cosmos Hub launch to ensure that no critical vulnerability would break the mainnet. This year, GoZ is the sequel to that for stress-testing IBC before we dive, head first, into a high-stakes connected multi-chain world. This year, participants are going to be spinning up relayers and racing to drive the most IBC packets across federated testnets for GoZ.

    The GoZ topology will look something like this:

    Source: https://goz.p2p.org/

    Participate in the Game of Zones: https://goz.cosmosnetwork.dev/

  • Cosmos Governance: The Cosmos Hub community tax pool had been slowly accumulating ATOMs since the launch of the mainnet to having over 340 thousand ATOMs. At the end of January 2020, the very first community spend vote was passed: Prop 23 (source)

    Prop 23 stated that it would pay 5250 ATOMs to instantiate a Cosmos Governance Working Group. It passed with ~91% Yes in which 63% of the voters were validators while the other 37% of the voters were non-validating ATOM holders who overrode their validators' vote.

    In Cosmos, there is a semi-liquid democracy whereby votes are by default cast by validators and individuals with voting power inherit the votes of the validators they're staked with. However, if individuals choose to cast their own votes, they can override their validators' votes, which is what we have seen happen with this proposal. Because of this governance feature, Cosmos governance differentiates itself from representative democracies as seen in DPOS protocols like EOS.

  • Staking Amidst Market Crash: March was a tough month for everyone. If we rewind to February 2, 2020, we saw peak staking in the network at 187 million ATOMs. It fell by 4% six weeks later, just two days after the entire crypto market took a nosedive on March 16th following Bitcoin's lowest price point since 2017. Bitcoin, Gold, and the DJIA were all of a sudden correlated. But because Bitcoin movement informs all the rest of the token economy's movements, ATOM, like comparable POS protocols (e.g. EOS, Tezos), took 69% haircuts ubiquitously. Network staking levels:

    Source: http://cosmos.fan/#/data

    Because it takes 21 days, or 3 weeks, to unbond ATOMs, you can see the first wave of people who undelegated at the first trough on January 28th. The second lower trough came exactly 3 weeks later on March 17th, where the ATOM price hit its year-to-date all-time-low to $1.55—a 69% dump. One can only infer that those first wave of undelegators were the ones who sold off to the bottom of the market.

📌 Decred

Contributor: Checkmate, Decred contractor

  • The past month of price action in the cryptocurrency markets has been extreme, with Decred price action being no exception. This week we look at a number of key metrics which have reached extreme values, often associated with the formation of price bottoms, reflexivity, and mean reversion.

    The first metric is the Stock-to-Flow model which was developed by the author. This model considers a log-log regression fit between daily values of Market Cap and the Stock-to-flow ratio of the DCR supply. The S2F Multiple is also shown which functions as an oscillator, indicating when network valuation has become over/undervalued relative to the S2F 'fair value' model. Following the price drop on 12/Mar, the DCR S2F multiple has entered the historical low zone last seen in Jan 2017.

  • The next chart shows the statistical distance between the Decred Market Cap and the predicted S2F model valuation, measured in standard deviations. For reference, an equivalent S2F model for Bitcoin is shown, with some interesting similarities in the fractals playing out in Decred's price discovery.

    It can be seen for both networks, that as network value approaches ~2x standard deviations from the prediction, price tends to snap back towards the mean. For Bitcoin, this generally coincides with halving events, a shock to S2F value and scarcity. For Decred, this is more closely associated with oversold conditions since the smooth issuance curve is less variable than Bitcoin's.

  • An on-chain metric developed by @permabullnino is the 142-day sum of all USD value bound in Decred tickets. DCR coins bound in tickets are indicative of strong demand for holding DCR long term. This metric (red line) has shown to act similar to an upper bound Bollinger Band as resistance during price discovery.

    By taking Fibonacci multiples (23.6%, 38.2% and 61.8%) of the 142-day ticket sum, additional trading ranges and boundaries have been identified. In particular, the 23.6% Fibonacci multiple (green line) has provided lower bound support throughout bull and bear cycles. In the 12/Mar market sell-off, price pierced below this level before rapidly bouncing back into the range.

  • Decred ASIC miners have endured very challenging market conditions after ASIC hardware was first released in Jan 2018, at the peak of the alt-coin market cycle. Given miners are long term thinkers and investors, the Puell Multiple provides insight into whether income streams are profitable or not and the level of stress in the hash-power network.

    The Puell Multiple takes the ratio of daily PoW USD income to its 365day average. This provides a view of today's income relative to the past year. Similar to the metrics shown above, the Puell Multiple is approaching an extreme value commonly associated with the proverbial event where 'miners put the bottom in'.

  • The Decred DEX is currently under development and is aiming to provide trustless exchange between crypto-assets via atomic swap technology. On Wednesday this week, Decred DEX server client successfully coordinated its first trustless exchange between DCR and BTC on test net.

    The DEX swapped 42 DCR for 0.42 BTC with an output from bitcoin-core testnet below showing successful receipt of the coins. Full transaction details of the atomic swap are found here for those interested in the inner workings (https://gist.github.com/chappjc/6c5bc6d9244e02249b867e8fe76e4762).

Subscribe to Our Network for free crypto insights every week:

About the editor: Spencer Noon leads investments for DTC Capital, a fundamentals-focused crypto fund. He actively tweets about on-chain metrics.

Our Network: Issue #14

Updates on MakerDAO, Compound, Aave, and Set.

Our Network is a weekly newsletter where top blockchain projects and communities share data-driven insights and advanced metrics.

Editor’s Note

Before we dive into this week’s issue, I wanted to share that Our Network is officially live on Gitcoin Grants!

Please consider donating if you believe in our mission of delivering free community-driven crypto analytics. All funds generated will go towards supporting three new strategic initiatives, which you can read more about on our grants page.

Click here to view the Our Network grant page.

Network Coverage

This week our contributors cover the following DeFi projects:

  • MakerDAO

  • Compound

  • Aave

  • Set Protocol

📌 MakerDAO

Contributor: Primož Kordež, Founder of BlockAnalitica

  • Liquidation Period P&L Analysis. Starting on March 12th, the price of ETH crashed more than 50% in less than 24 hours and caused massive liquidations on MakerDAO. Another severe ETH price drop came 3 days later, causing an additional $4.3m collateral was liquidated. The first wave of liquidations ended poorly for MakerDAO and Vault owners, as auction keepers severely underbid liquidated collateral due to reasons such as liquidity issues, low bid duration, and network congestion.

    Stats during this period of extreme volatility from March 12-16:

    • During the 3 days of liquidations, Maker lost $5.8m and Vaults lost $7.1m (or 90% of remaining collateral value), while Keepers earned $10.7m (83% ROI). Note: the difference between the higher overall loss vs. total profits is due to falling prices during the auction duration as well as an oracle price delay. Also, Vault losses include penalty fees and price exposure loss during auctions for their remaining collateral.

    • The loss for Maker equals 2.3x its yearly burn rate if we extrapolate the first 3 month of MCD net revenue cashflow ($624k). The Maker system also lost more than it has earned since inception, as total MKR supply after debt auction has now surpassed 1m MKR.

    • If auctions would have worked as planned (i.e. bid from keepers would be at least 13% above all debt that was liquidated), Maker would have instead earned $2.4m in penalty fees and incurred zero loss. That said, Maker still earned almost $1m in penalty fees on some loans.

    • While most of the losses came from ETH-collateralized Vaults from March 12-13, there was also an insignificant amount of losses generated on March 16 from BAT-collateralized loans. On that day, keepers bid ~50% below debt value as BAT’s price and liquidity worsened heavily. Losses from BAT loans amounted to $340k for Maker and $410k for Vault owners.

  • Dai Peg & Liquidity Issues. A lack of Dai liquidity was one of the main reasons that the liquidation auctions ended poorly. Even if everything else worked as planned, it is questionable whether keepers would have been able to source 22m DAI in less than 24 hours to protect Maker and Vault owners from losses.

    This was also one of the reasons Dai experienced heavy price appreciation and traded above $1.1 even before the auctions were triggered March 12th, as Vault owners were searching for Dai to close out their CDPs. On the first day of liquidations, Vaults repaid a net 12.6m Dai, which was the largest unwind of positions ever seen in MCD. This significantly reduced the number of future liquidations, and had this not happened the sheer amount of total debt could have been catastrophic for the protocol. DeFi Saver alone managed to protect 4m of debt from getting liquidated.

    The chart above shows how Dai price was appreciating during the time when Vaults were deleveraging and auction keepers were acquiring inventory to bid in auctions. The other reason that the price of Dai increased during this time was because DeFi investors were flocking to it as a safe haven. It is commonly known that Dai price is negatively correlated with ETH price, but this negative correlation is even more severe when ETH price experiences large drops. To illustrate, Pearson correlation coefficient between ETH and DAI price (both 24h rolling VWAP) for last month measures -0.82.

  • USDC-Collateralized Dai Analysis. The lack of Dai liquidity led Maker governance to perform an executive vote to onboard USDC as a third collateral asset type in advance of MKR debt auctions. This would allow auction keepers and market makers to provide sufficient liquidity to the system by minting new Dai with USDC (an asset that has a large liquidity base and does not carry ETH price exposure risk). Vault owners minted about 6m Dai in the first 3 days after it was introduced:

    At the time of writing, there are 95 USDC Vaults open with at least 100 USDC in collateral. The total amount of USDC collateral is 11.725m, and the total debt from this is 6.838M Dai (171% collateralization ratio). It’s also worth noting that whales dominate these Vaults:

    • The largest Vault represents 41% of total USDC Dai debt

    • The 5 largest Vaults represent 78% of total USDC Dai debt

    • The 10 largest Vaults represent 88%. of total USDC Dai debt

    On-chain analysis suggests that Vaults with higher collateral ratios are accumulating Dai and also hold some of the largest Dai debt positions in Compound. These entities might be keepers accumulating liquidity for potential upcoming collateral auctions and past debt auctions. On the other hand, many of the lower collateralized vaults opened leveraged DAI short positions by using iLeverage contracts from iearn or sold Dai on various decentralized exchanges. These entities are essentially providing liquidity and downward pressure on Dai price, which was one of the purposes of introducing USDC as a new collateral type.

    As of March 24th, Dai is currently still trading above the peg at around $1.03 but the situation has improved by at least $0.02 since USDC collateral was introduced and market makers arbitraged the price. If we aggregate Dai orderbook bids above $1 across all DEX and CEX, there still seems to be 8m Dai in demand at prices above $1. However, there is also additional Dai demand not seen in order books, so the real demand might well be 2x-3x higher. This also somehow coincides with a 33m Dai supply drop since 12nd of March. It is therefore reasonable to assume that previous Dai holders were short-term holders and sold Dai because of a premium (although it is worth noting Dai demand is probably also lower because of 0% DSR and more stable ETH price.)

  • MKR Debt Auctions. As of Wednesday this week, 86 different MKR debt auctions successfully finished, recapitalizing the MakerDAO system by $4.3m. There are still 20 auctions needed (~1m Dai debt remaining) to recapitalize the system fully. Below are stats for MKR debt auctions during the period 19th till 23rd of March:

    • 31 unique auction participants

    • 14 unique bidders won auctions

    • 86 auctions finished

    • 4.3m raised / 17.630 MKR minted

    • Average DAI price per MKR: 243.9

    • Total MKR supply: 1,002,804

    The average auction price of $243.9 implies a MKR market capitalization of $244.6m. In order to calculate at what forward P/E ratio auction participants recapitalized the system, assume some inputs for the next 365 days:

    • Stability Fee = 8%

    • DSR = 6% (DSR spread = 2%)

    • DSR utilization = 75%

    • Average DAI debt throughout the year = 100m

    • Bad debt = 0

    • No fees from SCD

    • Debt liquidated each month = 1.7% (based on 2019 comparable data)

    • 100% of penalty fees successfully collected

    Considering the inputs above, net revenue for Maker in one year amounts to $6.1m. This means that the forward P/E ratio at which MKR auction buyers made recapitalization measures 40.1. Sensitivity table on different debt and DSR spread inputs can be seen below.

📌 Compound

Contributor: Lucas Campbell, Growth at DeFi Rate

  • Compound’s weekly borrowing volume has increased steadily since the beginning of the year. The lending protocol reached new yearly highs earlier this month after hitting $25.54M in borrowing activity across all supported assets. In March, the average weekly volume was $14.7M, up +42.37% from February’s average of $10.3M in loans originated.

  • Compound users are mostly taking out loans in either DAI or USDC - a common occurrence throughout lending markets as a whole. In the last 8 weeks, Dai borrowing volumes have substantially outpaced the rest of the field with over $61M in total volume. The second highest asset is USDC which only processed $22.56M in loans over the same time period. This is largely due to the two stablecoins offering the only attractive yields on the market. To highlight this briefly, Compound’s 30-day average lending rate for Dai is 7.66% while USDC’s sits at 2.27%. Comparatively, lending rates on other supported assets like ETH or ZRX are <0.02%.

  • Naturally with the rise in borrowing activity, there’s also been a moderate increase in Compound’s supply volume over the past 8 weeks. The protocol reached yearly highs of $113M in supply volume on the week of February 17th. The nine-figure record pushed February’s average weekly deposits to just over $69M, beating out March’s average of $66M by 4.34%.

  • There’s also been a surge in Ether deposits in recent weeks, making it the most liquid lending pool on Compound with over $41.61M in gross supply as of writing. The biggest spike was in February where the protocol saw $83M in ETH supplied to the lending pool in less than a week. With that, total ETH locked on Compound peaked at 517,000 ETH in early March as average weekly ETH supply volume reached $26.85M. Other assets with notable deposit volumes include DAI and USDC which have both remained relatively steady over the past few months. Since late January, DAI has averaged $16M in weekly supply volume while USDC averaged $10.2M.

  • While lending rates have fallen off due to Black Thursday and the DSR drop to 0%, USDC and DAI are the clear leaders as income-generating assets for its holders. DAI accounts for over 68% of total interest accrued this month on Compound, generating over $127,000 for cDAI holders. The other stablecoin, USDC, represents 28% of the protocol’s accrued interest in March, earning around $52,000 for cUSDC suppliers. The rest of the assets (BAT, ZRX, ETH, REP, SAI, and WBTC) account for less than 4% of the remaining interest with SAI having the most notable earnings out of the group with $4,323 in accrued interest.

📌 Aave

Contributor: Isa Kivlighan, Aave team

  • Flash Loans Volume. Recently, one of the most interesting statistics is the huge increase in the volume of Flash Loans made on Aave Protocol the past 2 weeks. This spike occurred on 3/12, the day now known as Crypto Black Thursday. This increase was driven by the liquidations on Maker Dao. It is worth noting that DeFi Saver played a major role in driving up the Flash Loan volume on Aave on Crypto Black Thursday (source).

  • After the Crypto Black Thursday liquidation events, Aave Protocol was not largely affected by the overall reduction in LTV. Aave passed from 5th to 4th on DeFi Pulse as a result.

  • The market size of Aave Protocol has significantly increased (values are reflected in USD) recently. It now stands at $37.2M, with $29.9M total value locked and $7.3M total borrowed value currently (source).

  • According to Aave Watch, 266.76 ETH have been collected for protocol fees so far out of a total 363.23 ETH (the remaining pending fees will be collected in the future). Roughly 80% of fees go toward burning the LEND token.

  • Aave Watch shows that the volume of deposits and borrows (denominated in ETH) has steadily increased since the launch of Aave Protocol in January.

📌 Set Protocol

Contributor: Anthony Sassano, Head of Marketing at Set Protocol

  • Over the last few weeks, Set has been working closely with imToken on exposing TokenSets to their users. Their team generously moved TokenSets to the Recommended Section of their app which resulted in imToken becoming a top 5 traffic source for TokenSets.

  • Most Sets were positioned in USD during crypto's Black Thursday which resulted in Set Protocol's AUV being mostly unaffected and some Sets were even up 100%+ against ETH. The below image is from DeFi Pulse one day after Black Thursday.

  • Set Protocol handled almost $3.4 million worth of buy, sell and rebalance volume on February 26th which was a new all time high for the protocol. This volume mainly came from rebalances with the largest being the ETH20SMACO Set ($1.2mil rebalance) and the yield version of the same Set ($842k rebalance).

  • Set Social Traders had a shakeup over the last month with Aaron Kruger falling from first to third in the rankings while Adam Haeems and Sanz Prophet rocketed to the first and second place respectively. This happened as a result of both of Adam's Sets being the 2 best performing social trading Sets since inception (up 83.8% and 65.5% against ETH respectively) and Sanz Prophets Set being the third most profitable Set as it's up 62.3% against ETH.

  • Social Traders have accumulated over $24,000 in fees since we launched the platform a little over 2 months ago. It's important to keep in mind that these fees are just one time buy-in fees - we'll be deploying a more advanced fee structure very soon that'll allow traders to take fees based on performance.

Subscribe to Our Network for free crypto insights every week:

About the editor: Spencer Noon leads investments for DTC Capital, a fundamentals-focused crypto fund. He actively tweets about on-chain metrics.

Our Network: Issue #13

Updates on emerging networks.

Welcome to Issue #13 of Our Network, a weekly newsletter where top blockchain projects and communities share data-driven insights and advanced metrics.

This week our contributors deliver updates on emerging networks:

  • Opyn 🆕

  • Handshake

  • Coda Protocol 🆕

  • Kadena

  • Polkadot

Network Coverage

📌 Opyn

Contributors: Zubin Koticha, Co-Founder & CEO of Opyn

  • Opyn provides DeFi insurance and risk management built using protective put options. Opyn launched one month ago and has 1851 ETH ($234k at time of writing) collateral backing $130k of insurance. This is across three insurance markets: ocUSDC, protecting USDC deposits on Compound; ocDAI, protecting DAI deposits on Compound; oCRV, protecting y.curve.fi. The graph below shows the collateral locked up over time. (Source - Opyn dashboard by Tom Schmidt)

  • In the past month there have been 166 unique Ethereum addresses holding oTokens, protecting themselves against DeFi risks. (Source - Opyn dashboard by Tom Schmidt)

  • To provide protection, insurance providers put down ETH into a vault to mint oTokens. Insurance providers must maintain a minimum collateralization ratio of 140%. Given recent ETH price volatility, a decentralized response has formed with community members liquidating on the platform. Decentralized liquidators and the liquidation process is necessary for a robust protocol. The graph below shows the amount of collateral liquidated over time. In total $81,474.21 has been liquidated. (Source - Query by Matteo Leibowitz)

  • Right now insurance buyers and insurance providers buy / sell oTokens through Uniswap. This means that the price of oTokens is also determined by demand / supply on Uniswap. Uniswap liquidity is crucial to allow for buying and selling oTokens. The graph below shows Uniswap pool liquidity and activity for each oToken over time. (Source - Opyn dashboard by Tom Schmidt)

📌 Handshake

Contributor: Chjango Unchained, Handshake Enthusiast

  • Supply: The total circulating supply of HNS was hampered during the period of transfer lockups, reflecting only mining rewards, that is, until transfers were unlocked on February 13th, when the blockchain surpassed 2016 blocks. The subsequent and sudden increase in total circulatory supply jumped by 63%, increasing the total circulating supply from 3.7M to 237.8M, over the course of 24 hours.

    The majority of those transaction activities are characterized by airdrop redemptions. While airdrop amounts were pre-allocated to recipients, the allocated amounts did not count toward circulating supply until airdrops were redeemed.

    Fun fact: As to be expected anytime there is financial incentive involved, people will find ways to capitalize on those incentives. Knowing that a majority of the Handshake supply was pre-allocated to unique GitHub SSH and PGP key holders, certain individuals went so far as to try to buy SSH/PGP keys from GitHub users, causing a stir in GitHub land.

  • Transaction Volume: Since launch, Handshake has seen over 170 thousand transactions executed on-chain, a transaction volume far exceeding that of comparable decentralized naming systems.

    The spike you're seeing in the chart indicates over 13 thousand transactions executed in one day. It so happens that, on that day, there was a spike in the number of name redemptions on the Alexa Top 100,000 that were expired. We think that this is attributed to a single person who scraped the top 100,000 list for expired names, bought them all at deep discounts on a regular domain name registrar, and collected the 503 HNS bounty per name redeemed on the list, making a ~33% profit per name if we assume 1 HNS = $0.20 and $3/name.

    On Handshake, if you own a name belonging to the the Alexa Top 101-100,000, then you'd receive 503 HNS as an incentive for redeeming your name. If you own a name belonging to the Alexa Top 100, then your incentive is disproportionately higher. No pre-reserved Top 100 name has yet been redeemed.

  • Coins Burned: The total number of HNS that has been burned since mainnet launch is over 1 million. This represents the total number of HNS that's been spent on name auctions to procure the names that have become available so far (over 13,000 names have been opened for auction to date).

    Name purchasing comprises of 0.4% of the total transaction volume, based on number of coins burned. The activity that comprises the majority of transaction activity would, again, be airdrop redemptions.

  • Hashrate: The network hashrate has risen to over 50.96 TH/s since we last visited Handshake on Our Network. That's a 162% increase in hashrate since a month ago. If the network hashrate continues on this trajectory, it would only need to grow 3 times more in order to catch up to the network hashrate on Ethereum.

  • Profitability: Handshake is currently the most profitable coin to mine with GPUs. It's currently 87 times more profitable to mine than Ethereum and 10 times for profitable to mine than Bitcoin.

    Estimated Daily Revenue for 1 GH/s ≈ 5.53459012 HNS ≈ $ 0.7973 according to F2pool's mining profitability calculator.

📌 Coda Protocol

Contributors: Claire Kart, Head of Marketing & Community

  • Coda’s become one of the largest layer 1 testnets by number of unique users since launch in July 2019. Currently, there are 400+ users, growing 1.5-3X with each testnet release. 20% of network participants are running a node for the first time, growing the overall ecosystem of node operators. The easy of use & maintenance, quality of technical documentation and community-based support for new users have been the main drivers of both overall growth and first-time operators. This broad participation will increase decentralization at launch.

  • The most recent testnet release, 3.1, saw 300+ users stake and become block producers. 3x the size of the prior stable release. This influx of new users uncovered critical bugs, broke the testnet and aided the team in identifying where to focus future development. The testnet is currently down and the team is addressing the identified issues.

  • The last stable testnet, “Winter Special”, was conducted from Dec 2019 - Jan 2020 and achieved several milestones that advanced the stability and health of the network. These include:

    • 10+ week live network with over 12,000 blocks on the longest chain

    • 150,000+ SNARKs were accepted in blocks. Most successful SNARK worker sold over 20,000 SNARKs to block producers on the SNARKetplace

    We started to see users specializing their skills sets between block production and SNARK work, suggesting that users are exploring the different network roles on Coda, the incentive mechanisms, and finding ways to optimize their performance and results. Only one participant was able to become a top performer in both categories. As we progress towards mainnet, more of this type of behavior will be measured and analyzed.

  • The team recently published a new mental model, ScaDe (Scalability-per-unit-of-Decentralization) to guide choices during the design & optimization of blockchains. From the analysis, it appears that the industry has hit a frontier in how powerful we can design these trust machines, and a new model of optimization is needed. The hypothetical mapping of the mental model matches fairly closely with real world data. The team uses this model to consider how certain choices will impact the performance and stability of our network.

📌 Kadena

Contributor: Tony Pham, Head of Marketing at Kadena

  • Kadena's hash rate has grown considerably by 3x to consistently ~100 TH / s. A couple weeks ago, the hash rate was averaging 30-40 TH / s.

  • Since our last update about 6 weeks ago, over four million blocks have been mined on Kadena's public network, a 33%+ increase.

  • There have been 75,000 successful token transfers on the Kadena network. That is a 20% increase in the number of transactions taking place since the last update. Below is a visualization of the transactions:

  • Block time for KDA remains consistent at ~3 seconds, making Kadena among the fastest public chains in the market. (For context, Ripple runs one block about every 3.5 seconds.)

📌 Polkadot

Contributor: Bill Laboon, Technical Education Lead at Web3 Foundation

  • Although Polkadot mainnet has not been released yet, our canary network, Kusama, has now been running since August 2019. A canary is an early and unaudited release of software; it is not a testnet. Kusama currently has 378 nodes running and registered with our telemetry server, with 180 validators producing and finalizing blocks. The number of active validators was recently increased from 160, by vote through on-chain governance. Additionally, 520 nodes are running other Substrate-based chains, which can be parachains on Polkadot or Kusama. The latest statistics for all chains can be found on the Polkadot telemetry server (link).

  • As a nominated proof-of-stake network, the security of the network can be measured in terms of the amount of economic power backing the validators. Currently, the validators on Kusama are backed by a total 3.201 million KSM, or approximately US $4.7 million. This has more than doubled from January, when less than US $2 million worth of KSM was staked.

  • One of the more interesting aspects of Polkadot (and Kusama) is the ability to do an on-chain upgrade by propagating a new runtime over the network. This is because the entire runtime is a blob of Wasm code, which is executed by the Polkadot Host software running on the node. After on-chain governance votes on new runtime code, the Wasm blob can then be sent to running nodes. Twenty-two on-chain runtime upgrades have now occurred on the Kusama network, showing the feasibility of this approach.

  • During Kusama's run, mean block production time has, with one noticeable exception (detailed below), stayed remarkably close to the target of 6 seconds per block. In fact, over the last two weeks, the most the daily mean has strayed from 6 seconds was 6.22 seconds on 9 March. Block production has been steady even during higher-than-expected network usage and runtime upgrades.

    You will notice a spike in block production time in early January. This was an intentional blip where the timestamps were artificially sped up to fix a problem with the network Gav's post for details. (Data courtesy of Polkascan.)

  • Web3 Foundation's grants program has continued to review grant proposals and has accepted almost 100 different projects, in fields ranging from new development tools and runtime modules, to wallets and bridges to external blockchains, and even basic cryptography. A full list of accepted grant applications can be found here.

Subscribe to Our Network for free crypto insights every week:

About the editor: Spencer Noon leads investments for DTC Capital, a fundamentals-focused crypto fund. He actively tweets about on-chain metrics.

Our Network: Issue #12

Mega-update on #DeFi.

Welcome to Issue #12 of Our Network, a weekly newsletter where top blockchain projects and communities share data-driven insights and advanced metrics.

This week our contributors deliver a DeFi mega-update:

  • Chainlink 🆕

  • Lending rates 🆕

  • Nexus Mutual

  • Stablecoins

  • PoolTogether

  • Kyber Network

Network Coverage

📌 Chainlink

Contributors: ChainLinkGod and CryptoSponge

  • Economic activity of Chainlink oracle networks in aggregate has been steadily increasing since mainnet launch on May 30th, 2019. This is due to a combination of fluctuations in the price of the ERC677 token LINK (used to request data from nodes), an increase in active nodes, and an expanding array of oracle networks.

    In total, node payments have increased from $400 of LINK per day to approximately $7,000 of LINK per day, a growth of 17.5x. Active node count has increased from three to 31, a growth of 10.33x. Payments per day have become more volatile as oracle networks can update on a static schedule (e.g. hourly, daily) and/or now a deviation based schedule (e.g. 1% deviation since last update). This chart shows the LINK payments to each node per day, multiplied by the LINK/USD close price each day.

  • The below chart shows the cumulative number of contracts that have interacted with the LINK token and starts at the token sale in September 2017. From then, the number of Chainlink-related contracts grew until January 2018 when the overall crypto markets turned over. Growth in contracts stagnated between 400 and 700 contracts for 15 months (increase of only 1.75x) until just before the launch of mainnet in May 2019 when contract growth hit an inflection point and began to turn vertical.

    Since then, the number of contracts has increased from 700 to over 5,300 cumulative contracts, a growth of 7.57x in nine months. Recent growth in this metric can largely be attributed to smart contract developers integrating Chainlink oracle functionality in their decentralized applications.

  • The diversity of price feeds live on mainnet and their respective sponsors have expanded from 10 price feeds supported by two projects to now 29 price feeds supported by 10 projects, a growth of 2.9x in price feeds and 5x in sponsors in the past three months. The BTC/USD and ETH/USD price feeds consume such a high percentage of Chainlink’s bandwidth due to its higher relative decentralization (20-21 nodes vs 7-9) and more sensitive update frequency (every 20 minutes and 0.5% deviation).

    The ETH/USD price feed has shifted from a single sponsor to three independent sponsors. This reduced the costs experienced by the original sponsor by 66% with no reduction in node payments. As more projects support a common oracle network, the costs for each individual sponsor goes down, creating a public good model where security can be higher than what a single project could reasonably afford on its own.

  • Tracking Ethereum addresses that hold LINK reveals a similar trend to the cumulative contracts graph. After the token sale in September 2017, there were 3,867 addresses that held at least one LINK. After a few months, this grew to 15,262 addresses (increase of 3.94x). In February 2018, active address count hit 20,000 after which growth slowed substantially with only an increase of 1.37x in 15 months to 27,500 addresses.

    Growth picked up just before mainnet in May 2019 with an additional boost in June when Coinbase listed LINK. After hitting an inflection point, active addresses have since increased 2.68x in the past nine months. At the time of writing, there are 73,900 active addresses that hold at least 1 LINK, with a total of 268,320 addresses that have ever held the LINK token. From token sale to today, active addresses have increased 19.11x and total addresses increased 69.38x.

📌 Lending Rates

Contributor: Lucas Campbell, Growth at DeFi Rate

  • USDC Lending Rates: In the past 3 months, USDC lending rates have seen a relatively steady increase across the board. Nuo is leading the pack, currently offering 14.25% APR on USDC loans with a 90-Day average of 6.48% in annualized returns. Coinbase, the stablecoin’s issuing entity, offers the lowest rates on the market with a mere 1.25% APR for all Coinbase users. With that, we also see DeFi protocols like Compound and dYdX offering substantially higher rates than Coinbase with a 90-day average of 4.08% and 4.24%, respectively.

  • USDC Risk Premium: In traditional finance, the risk premium is calculated by subtracting the expected return by the risk free rate. In the lens of crypto, the “risk free rate” can essentially be seen as the rate of return offered by the stablecoin’s issuing entity while the expected return is the rate offered by a third-party lending platform/protocol. The risk premium is a good measure for the underlying risks associated with the loan as well as the additional incentive for making a deposit into the respective third party lending platform/protocol.

    For USDC, the risk free rate is Coinbase’s 1.25%. With that, Nuo has the highest risk premium of 5.23% followed by Fulcrum (2.77%), dYdX (2.99%) and Compound (2.83%). Lastly, CoinList, a centralized crypto bank, offers the second lowest risk premium of only 0.70%.

  • Dai Lending Rates: Since the launch of the Dai Savings Rate (DSR), Dai interest rates have steadily increased as well. The DSR started off at 4.00% back in November and has incrementally increased to today’s rate of 8.00%. With that, dYdX is offering the highest rate of 11.53% despite the recent drop off in lending rates earlier this month. Interest rates for Fulcrum are 9.54% followed by Compound at 8.09% and Nuo at 7.44%. Finally, CoinList offers the lowest interest rate in the field for Dai of 2.4%.

  • Dai Risk Premiums (90-Day): The introduction of the DSR established the “risk-free rate” for the Dai lending markets. By looking at the 90 Day average risk premiums, we can actually see that all other platforms offer a negative risk premium, showing the nascency of the DSR and the potential risks associated with the savings contract. However, we can expect that over time the DSR will act as the risk-free rate or the base rate for the market as the DSR is battle-hardened and confidence in the system grows.

  • Dai Risk Premiums (Today): The notion of the DSR as the risk-free rate is becoming more apparent once we take a snapshot of the risk premiums today. Compound, dYdX, and Fulcrum have all factored in a small risk premium of 0.09%, 3.53% and 1.54% respectively. Interestingly, Nuo has a slightly negative risk premium as it only offers 7.44% APR on Dai loans. The only other platform to offer a negative risk premium today is CoinList, showing that users may be more comfortable trusting centralized crypto banks over nascent money protocols.

📌 Nexus Mutual

Contributor: Richard Chen, Partner at 1confirmation

  • After the bZx attack, Nexus Mutual paid out its first claims of 4 ETH, 30000 DAI, and 2600 DAI. The claims voting was not contentious (almost unanimous in fact!), showing that the mutual members agreed what was covered by Nexus and what wasn't.

  • The bZx attack woke up the DeFi community to the smart contract risk and the need for insurance. If anything, bZx ended up being good marketing for Nexus. After the hack, $1.3M worth of cover was taken out across a variety of smart contracts: 0x, 1inch, Aave, Compound, Curve, DDEX, dYdX, Flexa, Gnosis, HEX, Idle, iearn, InstaDApp, MakerDAO, Nuo Network, Set Protocol, and Uniswap.

  • Nexus Mutual became undercollateralized for the first time. The peer-to-pool (as opposed to peer-to-peer) liquidity model works well for insurance because there can be several times more cover than the capital pool size, so long as claim events aren't correlated. Currently the active cover amount to capital size ratio is 125%. By comparison AIG has a ratio in the order of 6000%.

  • The NXM token return can be thought of as 1) a leveraged bet on ETH and 2) an index bet on DeFi. Since Nexus Mutual launched last May, NXM has outperformed ETH by 43%.

  • There are now 531 unique addresses that hold NXM, a large number joining the mutual after the bZx attack.

📌 Stablecoins

Contributor: Ankit Chiplunkar, Research Lead at TokenAnalyst

  • Monthly Transactions: A total of 26M stablecoin transactions have occurred in the last 15 months. 76% of those transactions are of USDT (ERC20). The stablecoin number of transactions peaked at 4M transactions in Sep 2019.

  • Average Transaction Value: Over the last 15 months, the average transaction value is highest for USDC ($22k) and lowest for DAI ($5k). As you can see, the average value of transactions has been decreasing over time. Notably, in Jan 2019 the average transaction value of USDC was $100k, which has dropped to $16k in Mar 2020.

  • Exchange Inbound Transactions: A total of 835k transactions were the result of stablecoins being sent into exchanges over the last 15 months. Of those transactions, 70% were USDT (ERC20) while 18% were USDC. This figure peaked at 140M transactions in Feb 2020, and the numbers continue to climb.

  • Exchange Monthly Inflows: Since Jan 2019, USDT has posted the highest average monthly inflows among stablecoins at $27k/month while DAI has posted the lowest at just $1k/month. The total monthly inflow value of stablecoins has been trending downward over time, likely because there was a mass onboarding of ERC20 stablecoin liquidity to exchanges in Jan 2019.

📌 PoolTogether

Contributor: Leighton Cusack, CEO of PoolTogether

  • PoolTogether is a prize savings protocol. The Dai pool launched 11 weeks ago has now distributed over $10,000 Dai in prizes. The following chart shows what percent of total winnings have been paid out to ticket holders in each category. 30% of all prizes have been paid out to addresses holding less than 100 Dai (1 Dai = 1 ticket). The luckiest winner had deposited 5 Dai and won 1,448.

  • A major trend of 2020 is less tickets owned by whales and more tickets owned by smaller addresses. As of March 10th, 17% of all tickets are owned by accounts with less than 100 tickets. In contrast on January 31st only 5.44% of total tickets were owned by accounts with less than 100. During this period, the total pool size continued to increase.

  • Unique Ethereum addresses in the pool has continued to grow currently averaging 30.42% week over week for the last 11 weeks.

  • One way to measure the overall health of the protocol is relative to other parts of the ecosystem. PoolTogether is built on top of Compound Finance. Compound Finance currently has 5,489 unique Dai suppliers (addresses) through their interface and PoolTogether has 5,296, meaning there is almost more people supplying Dai through PoolTogether to Compound than to Compound directly.

📌 Kyber Network

Contributor: Deniz Omer, Head of Ecosystem Growth at Kyber

  • Kyber had its best month on record in February as it reached new monthly and daily volume records and achieved its all time highest number of users and number of new users in a month since launch. USD volume for February was $139,590,000 while on the 13th of the month it reached an ATH of $11M traded on a single day (subsequently broken again 10th of March with $14M traded in a day). Kyber has also averaged $6.9M daily volume over the last 7 days (vs $5.4M Uniswap, $4.3M 0x, $3.4M Oasis).

  • DeFi dapps are showing continuing growth going into 2020 and many Kyber-integrated dapps have achieved double digit volume growth (for transactions passing through Kyber) from January to February. Fulcrum continues to be the largest DeFi integration and saw its February volume reach an impressive $15M despite some of the known difficulties faced by bZx.

  • Looking at the Kyber reserves, we see a very healthy liquidity market that is scaling up as volumes scale up with 576,943 ETH traded over 85,024 transactions in February. Kyber’s own reserve makes up 38% of all volume by providing liquidity to over 44 tokens, while the Oasis Dex Bridge and Uniswap Dex Bridge accounted for a quarter and a tenth of total volumes respectively.

  • The most popular token traded on Kyber continues to be DAI by a significant margin. Oasis shows its DAI price competitiveness by fulfilling 62% of all DAI volume ($32M) vs 22% fulfilled by the Kyber reserve ($11M) and 10% for Uniswap ($5M). Out of all tokens available on Kyber Network, KNC has seen the largest increase in volume by jumping from $3.6M volume in January, to $14.7M in February, followed closely by WBTC with $14.3M volume.

Subscribe to Our Network for crypto insights every week:

Our Network: Issue #11

Updates: ETH on exchanges, 0x, and Uniswap.

Image 2020-01-02 at 3.25.48 PM

Welcome to Issue #11 of Our Network, a weekly newsletter where top blockchain projects and communities share data-driven insights and advanced metrics.

Editor’s Note

One common criticism of crypto metrics is that they are too easily gameable.

This is true to some extent. Because there are huge incentives for crypto companies to juice their numbers, it happens a lot. For an example of this in the wild, look no further than the list of top exchanges by volume on CoinMarketCap:

When you look at the full list, you will notice a number of “top” exchanges that you’ve likely never heard of before. These are not sleeping giants on the other side of the world, beating out heavyweights like Coinbase and Huobi. They are simply small exchanges that are gaming their volume numbers so they appear on this list. It makes sense — CMC is the #534 website in the entire world (source), which means it’s a great source of referral traffic.

How do you game trading volumes? There are a few methods with varying degrees of legitimacy that these exchanges employ:

Metric gaming also happens frequently on public blockchains in order to give the appearance of a healthy network, which is a positive indicator for price.

Gaming is obviously problematic and something we should be aware of, but it doesn’t mean we should write off all crypto data as poor. It just means we need to approach metrics and lists like these with a healthy dose of skepticism. We also need to be rigorous with our methodologies in order to stay ahead of the gamers.

There is much signal when you sift through the noise.

Network Updates

This week our contributors provide data-driven coverage on the world of exchanges:

  • ETH on Exchanges

  • 0x

  • Uniswap

📌 ETH on Exchanges

Contributor: Alex Svanevik, co-founder at D5

  • More than 31M ETH is held in exchange wallets. This includes centralized exchanges such as Coinbase, plus decentralized exchanges such as Uniswap. For context, the commonly reported "ETH Locked in DeFi" is currently sitting at 2.9M ETH. In other words, there's more than 10X the amount of ETH "locked" in exchange wallets than in DeFi.

  • The vast majority of this ETH is in centralized exchanges, with Coinbase, Huobi, Bitfinex, Binance, and Kraken holding the most (above [fig. 1]). Drilling down on the full history of Binance (below [fig. 2]), we see how the exchange quickly hit its first 1M ETH right at the crypto peak of January 2018.

  • The millions of deposit wallets associated with exchanges can serve as proxies for their growth metrics. Binance will normally provide one single address per retail customer, which means that new deposit wallets (below [fig. 3]) is likely to correlate with the number of new exchange customers. Again we see that the largest influx of customers was around the peak in January 2018. Conclusion: Binance's growth in the last 18 months has likely come from 1) existing customers, and 2) non-Ethereum cryptoassets.

📌 Uniswap

Contributor: Caleb Sheridan, co-founder of Blocklytics

  • February's volatility has been great for Uniswap's trading volumes. The 30d average daily volume is now over $5m.

  • Liquidity reserves saw a sharp drop in both USD and ETH. The drop is due to recent ETH price action and liquidity being removed from the MKR market. This appeared to be a defensive tactic against flash loans. The liquidity has not returned despite a Maker upgrade that introduces new defense against flash loans.

  • As reported previously, other contracts drive a large amount of the volume on Uniswap. Notably, DEX aggregator 1inch is driving slightly more of the network traffic than the last time we looked at this chart.

  • Timing matters. Uniswap LPs win when trading fees surpass the opportunity cost of not holding ETH. This chart shows a tale of two theoretical positions. One where the price of ETH grew 75% and the LP misses out. And a second where the price of ETH grows 20% with high volatility.

📌 0x

Contributor: Alex Kroeger, Data Scientist at 0x

  • It’s been a great month for DEXs. Overall DEX volume is up, and 0x is no exception. Over the past month, 0x weekly volume increased roughly 6x, up from ~$5M per week to ~$30M per week. Overall trade count has largely been driven by NFT trading on TokenTrove, but excluding TokenTrove there has been an increase in the number of trades taking place on the protocol as well.

  • In the last update, there was a discussion the new economic incentive structure of the 0x ecosystem launched in version 3 of the protocol. An integral component of the new token economics is the introduction of protocol fees with each 0x trade, designed to align the incentives of market makers and ZRX holders with the best interest of the ecosystem.

    As adoption of v3 has grown, there has been a steady increase in protocol fees collected, which are paid out to market makers and ZRX stakers at the end of 10-day “epochs.”

  • The last update had a discussion of how protocol fees in 0x v3 share arbitrage profits with market makers. Shortly after that update, a large (likely errant) trade on Uniswap caused a massive DEX arbitrage opportunity between 0x and Uniswap’s ETH-DAI pool. After a gas price competition, the winning arbitrage bot paid a whopping 6 ETH (~$1,400) in gas costs and an additional 6 ETH in 0x protocol fees, showing the new token economics in action. Check out the transaction and Will Warren’s breakdown.

Subscribe to the newsletter and learn about crypto analytics every week:

About the editor: Spencer Noon leads investments for DTC Capital, a fundamentals-focused crypto fund. He actively tweets about on-chain metrics.

Loading more posts…