Fidelity Digital Assets: UTXO Model Leads to Overvaluation of Bitcoin's Real Trading Volume

Digital asset data is complex and should not be calculated at surface value.

Original author: Fidelity Digital Asset Research Director Ria Bhutoria

Digital asset data can be misleading in its raw and unfiltered form.

When calculated at surface value, some indicators exaggerate certain activities, while others may underestimate usage.

For example, the value of transactions on the Bitcoin network is often overvalued, at least economically, because the Bitcoin network uses an unspent transaction output (UTXO) system. Although the ledger does not distinguish between economic and non-economic transactions, we believe that it is important to exclude economic transactions where feasible.

Data providers (such as Coin Metrics) have made great strides in separating economic value from non-economic value. Several filters they apply to the raw data do not include:

  1. Known change output, or the transaction input part sent back to the original address, and
  2. Transactions that may be counted multiple times due to continuous spending and transmission (caused by transaction processes, mixers, etc.);

Therefore, the difference between adjusted and unadjusted transaction values ​​is significant.

For example, Bitcoin's total adjusted transaction value from year to date is about $ 2.2 trillion, while the number of unadjusted transactions is about $ 7 trillion. In other words, estimating the economic activity of Bitcoin based on accumulated raw data is actually about 3.3 times higher than the actual activity.


Figure 1 From: Coin Metrics, Fidelity Digital Assets (May 12, 2019)

Figure 1 shows the difference between the accumulated original transaction value and the adjusted transaction value. The adjusted value is adjusted based on the change in output of other factors.


Figure 2 From: Coin Metrics, Fidelity Digital Assets (May 12, 2019)

Figure 2 shows the ratio between Bitcoin's original transaction value and adjusted value.

Using unfiltered transaction value metrics also has a significant impact on the ratios that use this number as input, such as commonly used network value to daily transaction value (NVT) ratios and NVTS ratios.


Figure 3 From: Coin Metrics, Fidelity Digital Assets (May 12, 2019)

Figure 3 shows the difference between the use of adjusted transaction value and the use of unadjusted transaction value in the NVT ratio denominator.

Given the transaction value as the denominator, the original, overvalued number will result in an undervalued ratio (other things being constant).

Based on the NVT / NVTS ratio, this may be particularly misleading when comparing Bitcoin to other digital assets.

The concept of NVT ratio was proposed by Willy Woo more than two years ago. It is a price-earnings ratio of Bitcoin, which was later named by Chris Burniske. The NVTS ratio was proposed by Dmitry Kalichkin. It is the heir to the NVT ratio. These two indicators are the most commonly used by analysts to measure the Bitcoin cycle.

While some factors need to be adjusted downwards to clarify the value of economic transactions, certain other factors, such as transactions occurring on Layer 2 solutions such as Lightning Network, and the volume of transactions on centralized exchanges can The economic value of the transaction is underestimated.

Although the digital asset network provides a large amount of real-time, fine-grained data, in actual analysis, we should not base on surface value, but need to conduct deeper analysis to filter out noise and form correct conclusions as far as possible. .

Thanks to the Coin Metrics and Avon Ventures teams for their feedback and suggestions for this analysis.