ByteTree Market Health Update; Issue 43

ByteTree was conceived in 2013, which by Bitcoin standards, makes it part of the crypto furniture. Our first multi-coin index came about in April 2014, and by July 2014, we successfully downloaded Bitcoin’s blockchain data into excel. Back then, I spent days pouring over those numbers and came up with many of the metrics you see around you today, not just at ByteTree but across the crypto space.

The most fruitful early observations were “general spend” (now referred to as transaction value in the crypto space), and fees; both of which have proved to be essential in fundamental network analysis. That led to the Network to Spend Ratio (NSR) which took the active network value and divided it by the general spend, a sort of price to sales ratio for Bitcoin. Not only did it prove useful for forecasting, as in September 2019, but it gave the best estimate of fair value, which is essential for an asset’s credibility. Institutional investors won’t buy something without a valuation framework. These observations were described in Atlas Pulse back in 2014/5.

The bad news for ByteTree came in 2017 when the NSR was “invented” all over again; this time becoming known as the Network Value to Transactions Ratio (NVT). Having seen the NSR renamed the NVT, I now fully understand the controversies behind past inventions such as the telephone, calculus, the radio, and the lightbulb. I look forward to someone reinventing other ByteTree metrics such as the Miners’ Rolling Inventory (MRI).

MRI is another forward-looking indicator that has turned out to be a useful gauge of market liquidity. All of the metrics which we publish are analytically useful, and those that aren’t such as MVRV, SOPR and stock to flow, we don’t publish.

ByteTree’s new and improved metrics

The recent boom in privacy coins has prompted us to up our game.

Our methodology for spend, devised in August 2014, has worked exceptionally well and until this summer, we felt there was no need to change it (no pun intended). We first identified the problem in late July as our spend data started to rise, which ought to mean a pickup in network traffic, which I first reported here. But it soon became clear this was too good to be true, and my suspicions were confirmed by stripping out the large transactions.

Blockchain analysis is our bread and butter, and these things are simple for us to do behind the scenes, but take time to track back through time and publish on the site. We initially suspected it was CoinJoin transactions that were causing the headache in the spending boom. Still, on closer investigation, it turned out to be the CoinMixers that were having a more significant impact. Both of these relate to privacy transactions.

In the short time that the ByteTree Terminal has been offline, these have been identified and fixed. Not only have we captured “complex spend” and “batched spend” (multiple transactions), we have eliminated noise to give a cleaner, more accurate and up to date spend calculation.

Spend is complex and subjective to calculate because you need to make assumptions. Bitcoin transactions see “pieces” of bitcoin handed over and receive change, but the blockchain doesn’t clearly state how much change was received. Different transaction types with multiple inputs and outputs have flourished, and any methodology that looks to calculate the economic value transferred will only ever be an estimate. Our goal has never been to capture the exact spend number because it doesn’t actually matter for financial analysis. What does matter is a consistent methodology that will move in sync with the precise spend number, which can then be accurately calibrated for financial analysis.

We have completed this work in the background and look forward to sharing our new and improved data with you soon — hat tip to my co-founder Mark Griffiths, who lost a weekend for the cause.

Network Demand Model

Like all good companies, ByteTree has several backup procedures and a disaster recovery plan. What if ByteTree data shut down, or what if a privacy boom skewed some of our metrics? Sadly, I can’t share our secret sauce due to the rampant plagiarism in this space, but I can say we have separate and independent systems, and “manual” means of collecting data from the blockchain. At no point has the network demand strategy given a false buy or sell signal. The dip in early September was macro related as the Nasdaq took a knock and was little to do with the Bitcoin Network.

That said, this exercise has made us think harder, which is always a good thing. There’s much we want to improve, and this will be an excellent opportunity to revisit velocity following the improved metrics. I will report back to you on that. It is currently on the low side, below 600%, which keeps the network demand score at 4 out of 6. A 4 is bullish, and the price action seemingly wants to please.


As you know, ByteTree associates bull markets with buoyant velocity and vice versa. We don’t just think that we know, because it is a statement that is backed up by data and results. In a recent interview with a senior blockchain strategist on a leading crypto news channel, the discussion suggested that low velocity implied high HODL and therefore justified high valuations. That is 100% nonsense.

High HODL means lots of coins are locked up, which means less network activity, which means less value creation, and less potential for price appreciation. Be careful what you read; there’s no helping some.

This article has been cross-posted from the ByteTree blog, originally published on 16th September 2020.

Applied data for digital asset investors. ByteTree provides real-time data, fundamentals, technical and deep blockchain market analysis for Bitcoin, and more.