We introduce a new methodology to classify Bitcoin supply owned by long-term investors and short-term traders, and show when these two investor types are in a state of profit or loss.

To understand investor behaviour from an on-chain perspective, it is crucial to differentiate between Long-Term Holders (LTH, “investors”) and Short-Term Holders (STH, “traders”).

While traders aim to “beat the market” and exploit price fluctuations on short time scales, long-term investors have a low time preference and are in for the long haul — remaining of the conviction that BTC will see future price appreciation. These market participants HODL over long time frames ( A True HODLer Does Not Sell Their Coin), or only temporarily decrease their bitcoin positions in bull markets for profit-taking (“Swing HODLers”).

But how can such different…

Effectively mapping UTXO-based on-chain metrics to improved account-based versions for Bitcoin and Ethereum.

Abstract. Numerous on-chain metrics for crypto assets based on the Unspent Transaction Output (UTXO) scheme, such as Bitcoin, suffer from undesired contributions due to change volume or internal transfers, which overshadow the true market behavior signal. In the present work we use the concept of entities, to map UTXO-based blockchains to an effective account-based model and enable the computation of metrics without the aforementioned limitations. In addition, we show that many metrics, originally designed for UTXO chains, can be straightforwardly realized for real account-based systems like Ethereum. …

Introducing SOPR and MVRV for Long and Short Term Holders

Bitcoin’s UTXO-based system allows for analyses of on-chain data based on the “age” of bitcoins in the network, i.e. the categorization of bitcoins depending on the last time they moved.

This allows for instance to assess information on investors’ hodling behaviour, and to gauge whether coins that have been dormant for a long time are currently being moved.

Popular example of on-chain metrics incorporating coin age are, amongst others, HODL waves, Spent Output Age Bands, MSOL/ASOL, Coin Days Destroyed, and Bitcoin Dormancy.

Identifying Industry Stakeholders: Short and Long Term Holders

In the present work, we make use of coin age information with the goal of classifying market players…

Assessing Bitcoin’s True Transfer Volume

Even though blockchain data is publicly accessible, it is a non-trivial challenge to make sense of it in a meaningful way.

On-chain data is without a doubt highly valuable — but in its raw form it’s just not good enough.

It contains a substantial amount of noise, and careful preprocessing and contextualisation is required in order to distil useful information from it.

Consider on-chain transaction volume: Figure 1 shows Bitcoin’s raw daily on-chain transaction volume (USD value) for 2019.

Introducing a New Generation of Entity–based On–chain Metrics Using Clustering and Advanced Data Science

The Problem with Quantifying the Number of Bitcoin Users

A major question amongst Bitcoin researchers and investors has been that of knowing how many people actually own and use Bitcoin.

And even though Bitcoin’s entire transactional history is publicly accessible through its open ledger, assessing the number of users in the Bitcoin network is a non–trivial task.

Down to the present day, it is most often still the number of addresses in the Bitcoin network that is being used as a proxy to the number of Bitcoin users/holders.

However, it is well established that this approach is fallacious, mainly because there is no one–to–one mapping between users and Bitcoin…

Introducing a New Set of Metrics to Help Time Bitcoin Market Cycles

Understanding the state of the Bitcoin blockchain with respect to the price at which BTC is being valued by the market is indispensable for any investor in this new asset.

In fact, Bitcoin on–chain data has been used in conjunction with off–chain market data in order to create metrics and indicators that help investors gain fundamental insights into the current state of the market, understand investor sentiment, or model the value of Bitcoin.

Popular examples are MVRV and SOPR.

These metrics have in common that they make use of Bitcoin’s underlying UTXO structure by comparing the value of Bitcoin at…

On-Chain Metrics Show Highly Uneven Token Distributions Across ERC20s

TL;DR We integrated four different on–chain distribution metrics for ERC20 tokens that are now available on studio.glassnode.com. These metrics are Percent Supply in Smart Contracts, 1% Richlist Balance, Gini Coefficient, and Herfindahl Index. An analysis of these metrics across 73 ERC20 tokens shows that, generally speaking, tokens are unevenly distributed across network addresses. For instance, on average the richest 1% addresses hold 83.2% of funds held by externally owned addresses.

You can access all distribution metrics presented here on Glassnode Studio.

On–chain data gives insight into the underlying state and activity of decentralised networks. One interesting aspect is analysing and…

Using Litecoin when Sending Bitcoin Becomes Expensive

This article was originally published on Glassnode Insights.

If you’re no stranger to crypto, this is probably old news to you: The cryptocurrency market moves largely together, prices of digital assets correlate with each other.

Take a look at the following chart illustrating the correlation of price changes over time for BTC, ETH, and LTC.

Whale Accumulation at ATH: The Richest 1% Hold 87% of Funds

At Glassnode we delve into the world of blockchain data, in order to provide contextualised insights that help foster market maturity and support investors and traders make informed decisions.

In this post we are taking a look at the token holder distribution of Basic Attention Token (BAT).

BAT is a utility token on the Ethereum blockchain that is “used to obtain a variety of advertising and attention-based services on the BAT platform”.

The present analysis is based on block 7504775 on April 4th 2019, at which point ~104,600 addresses were holding BAT, whose total supply is 1.5 billion tokens.

Special Addresses

BAT tokens in contracts


Why you should be cautious with neural networks for trading

So I built a Deep Neural Network to predict the price of Bitcoin — and it’s astonishingly accurate.


See the prediction results for yourself.

Looks pretty accurate, doesn’t it?

And before you ask: Yes, the above evaluation was performed on unseen test data — only prior data was used to train the model (more details later).

So this is a money-making machine I can use to get rich!


In fact, I am giving you the code for the above model so that you can use it yourself…

Ok, stop right there. Don’t do it.

I repeat: Don’t do it! Do not use it for trading.

Rafael Schultze-Kraft

Data Science, Machine Learning & Crypto | Twitter: @n3ocortex | Building glassnode.com

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store