Glassnode: How does coin age affect the buying and selling patterns of BTC?

Author: Mario Schröck, Glassnode, Glassnode; Translation: Tao Zhu, Golden Finance

Preface

The transparent blockchain of Bitcoin allows for detailed analysis of token changes and holder behavior. By examining the age and spending probability of Unspent Transaction Outputs (UTXOs), we can gain insights into the dynamics of the Bitcoin ecosystem. This article explores the power-law relationship between UTXO age and buying/selling probability, revealing predictable patterns of token holding and trading over time.

Why is this analysis important

Understanding the UTXO spending behavior of Bitcoin provides powerful insights for traders, investors, analysts, etc. By revealing predictable patterns that control dormant coins, you can:

• Enhanced Investment Strategy: Predict potential liquidity changes and better quantify market sentiment.
• Improved on-chain analysis: Using mathematical framework to complement traditional LTH/STH indicators.
• Predicting holder behavior: Determine when the token may re-enter circulation and inform the time of trading or decision-making.

Whether you are optimizing trading algorithms, analyzing market trends, or refining investment methods, this framework can provide you with a clear, data-driven advantage in the Bitcoin ecosystem.

What are UTXO and spending probability?

The core of the Bitcoin blockchain is the UTXO model. UTXO stands for Unspent Transaction Output, which essentially refers to Bitcoin blocks that have been received but not yet spent. Each Bitcoin transaction consumes existing UTXOs as inputs and creates new UTXOs as outputs. These UTXOs can be considered as tokens stored in specific addresses, waiting to be used in future transactions.

By analyzing the age (in days since creation) of these UTXOs, we can infer the behavior patterns of holders in the network. A fundamental concept in this analysis is the spending probability, which measures the likelihood of a UTXO at any given date being spent. This metric quantifies the movement of Bitcoin within the ecosystem and the evolution of holder behavior.

Methodology

Data set and UTXO count

Our analysis is based on Bitcoin UTXO data from 2015 to November 2024. For each day during this period, we calculate the number of UTXOs for each possible coin age, ranging from one day to 10 years (approximately 3,650 days). We limit the maximum coin age to 10 years to avoid inherent noise in extremely old UTXO data.

Expenditure Rate Calculation

To determine the spending probability, we compare the number of UTXOs with a specific coin age on one day with the number of UTXOs with a higher coin age on the next day. The calculation for the consumed part is as follows:

Spent Score = 1 - (Number of UTXOs with T-day Coin Age N) / (Number of UTXOs with T-1-day Coin Age N-1)

The formula represents the proportion of UTXOs with coin age N-1 that did not appear as UTXOs with coin age N the next day, meaning they have been spent.

Then, we calculate the average expenditure rate for each age group in the entire dataset, as well as the standard error of the mean. Figure 1 intuitively shows the average expenditure rate divided by coin age.

Power-law dynamics in logarithmic-logarithmic space

To better understand the relationship between UTXO coin age and spending rate, we plotted the data in logarithmic space. This transformation is beneficial because power-law relationships appear as straight lines in logarithmic space, making them easier to identify and analyze. Figure 2 shows the double logarithmic graph of the spending rate.

Power-law Fitting

We perform linear regression on double logarithmic data to quantify power-law relationships. We use weighted least squares regression, where the weights are proportional to the square of UTXO count divided by the square of the standard error of the mean. This weighting takes into account the variability in data point reliability due to differences in sample size and variance.

The slope of the regression line corresponds to the power-law exponent, indicating how quickly the consumption probability decreases with age. Figure 3 shows the fitted regression.

Analyze residuals to assess the quality of fit

In order to assess the fitting quality of power law in different coin age groups, we analyzed the residuals, which are the differences between the observed average spending rate and our model’s predicted values. Plotting the residuals helps us identify patterns or systematic biases in the model. Figure 4 shows the relationship between residuals and UTXO coin age.

We observe that the residual of UTXO around 200 days is very small, indicating high predictability of the queue. This is consistent with the gradual transition from short-term holders (STH) to long-term holders (LTH). The S-shaped function models this transition to achieve a smooth transition of holder behavior. The central point of this transition is marked at 155 days, representing a 50-50 ratio between STH and LTH classification. At approximately 200 days, the completion rate of the transition from STH to LTH is 99%.

Our analysis indicates that the power-law model fits the STH token almost perfectly until they fully transition to LTH. For LTH tokens with a coin age of 3-4 years (the second transition zone), the model still remains in good shape (small deviation). These deviations suggest that the spending probability of the mid-term LTH group is slightly higher than the probability predicted by the model.

However, for ultra-long-term holders (ULTH) - tokens held for more than about one halving cycle - we observe more significant deviations from the model. Specifically, the observed expenditure probability is lower than the power law predicted probability. This suggests a greater tendency to hold these tokens, possibly due to strong holding beliefs or the possibility of some tokens being lost.

Power law arranged by time

We study whether the power-law dynamics of token expenditure probability change over time from another perspective. Instead of averaging the UTXO count for each coin age on all dates, we track groups of UTXOs born on the same day. Based on these date groups, we can analyze how the expenditure rate of tokens in different periods of Bitcoin’s history evolves.

For each group, we calculate the consumption rate on a daily basis as the age of the group’s currency increases. Then, we conduct linear regression on the double logarithmic expenditure probability of each group separately. Ignoring data groups with survival time less than 10 days in the most recent records will result in approximately 3600 remaining groups and the corresponding linear regression.

The coefficient of determination (R2) of each regression indicates the fit of the power-law model to the queue data. The slope of each line allows us to understand the rate of decline in consumption rate as the age of the coin increases. Figure 5 plots the R2 values and line slopes of each date group over time.

Overall, power laws are very applicable on different dates, confirming the consistency of this dynamic over time. However, specific periods show lower fitting quality, despite the lack of significant correlation with price changes during these periods. We observed that the expenditure probability for the entire year of 2019 (smaller slope value) was extended in advance. One possible explanation is that investors who bought in during the -80% drop from the 2017 ATH were for long-term investment, so their expenditure rate was higher than usual.

The impact of on-chain analysis

These findings provide a continuous perspective on the coin age and expenditure probability, complementing the existing LTH/STH framework. The power-law relationship reflects the gradual transition from active trading to long-term holding.

It is worth noting that the model fits younger tokens almost perfectly, and still performs well for tokens with a coin age of about four years (only a very small deviation). Beyond this coin age, the deviation of the model becomes more significant, indicating that other factors may affect the spending behavior of long-term holders.

A power law with a slope close to 1 provides a clear and intuitive empirical rule: as the lifespan of a token increases by a factor of ten, the probability that it will be spent decreases by roughly a factor of ten. The approximate model values in the table below illustrate this.

! [VOaUg7lHvQD57LKkaWv4c2hRH8LmKkGUbbHpcVQg.jpeg] (https://img.jinse.cn/7334389_watermarknone.png “7334389”)

The decaying probability of predictable expenditures highlights a behavioral pattern: younger tokens are actively traded or speculated, while older tokens become more dormant over time. By adopting this persistent perspective, analysts and investors gain a deeper understanding of the declining expenditure activity as tokens age, thereby enhancing the interpretation of on-chain data and investor behavior.

Quantitative heat supply assumption

Based on our data, we evaluated a simple predictive heuristic:

If UTXO is less than 7 days, assume that the UTXO will be spent on the same day. Otherwise, assume it will not be spent.

Using historical data, this heuristic method has an accuracy of 98%, which indicates that it can correctly predict whether UTXO will be spent in the vast majority of cases. However, due to the imbalance of the dataset, high-precision numbers may be somewhat misleading - there are a large number of unspent UTXOs on any given day.

Summary

Our analysis shows that **Bitcoin UTXO spending behavior is strongly controlled by power-law dynamics, with decreasing likelihood of spending older tokens. The power-law relationship fits younger tokens almost perfectly and remains good for tokens up to four years old (with only small deviations). For long-term holders with coin ages exceeding this, deviations from the model become more apparent, indicating even lower spending probabilities than predicted by the model. This suggests that other factors, such as strong holding beliefs or lost tokens, influence the spending behavior of these oldest UTXOs.

This finding enhances the existing LTH/STH framework by providing a continuous mathematical perspective on the transition from active trading to long-term holding. Power laws provide a precise empirical rule: the probability of a token being spent decreases by about ten times for every tenfold increase in its lifespan. This predictable decay in spending probability provides valuable insights into investor behavior and token dormancy over time.

With the continuous development of Bitcoin, the power law model provides a mathematically based framework for on-chain analysis, enabling people to gain a deeper understanding of the life cycle dynamics of UTXO.