Re-examining one of the most important on-chain critical indicators of Bitcoin

Author: Matt Crosby, Director of Research and Analysis at Bitcoin Magazine Pro; Translation: Shaw, Golden Finance

For years, the on-chain analysis field has relied on a fixed value to distinguish long-term Bitcoin holders from short-term traders: 155 days. Holding Bitcoin longer than this duration classifies someone as a long-term holder. This is one of the most widely used threshold values in the industry, and many analysts rely on this metric daily. I have spent a lot of time researching whether this fixed threshold still makes sense, and I have built a more dynamic analysis model that I believe offers greater accuracy and practicality.

Key Points

• The traditional 155-day long-term holder threshold originates from research in 2020 and is no longer suitable for the changes in Bitcoin holding behavior since then;

• The dynamically recalibrated threshold value today is 114 days, which is 27% lower than the fixed standard;

• After removing unspent transaction outputs (UTXOs) that lack actual economic transaction attributes, the realized price data becomes cleaner and more accurate;

• Using this new set of thresholds to reconstruct various indicators, it has historically been possible to accurately identify market cycle tops and bottoms;

• Based on these new indicators, Bitcoin is currently in a deep value zone.

The Fixed Threshold Is No Longer Applicable

The original logic behind setting the 155-day dividing line was sound: beyond this duration, the probability of holders selling Bitcoin drops sharply and stabilizes. Statistically, users holding for more than 155 days tend to be long-term holders. But the problem is that the behavior patterns of Bitcoin holders have changed dramatically. Today, very few people use Bitcoin for peer-to-peer payments. An increasing number of users hold Bitcoin indefinitely as a store of value, driven mainly by Bitcoin ETFs and corporate treasury allocations.

When user behavior on the entire Bitcoin network changes so drastically, the fixed threshold set six years ago can easily misjudge the true on-chain market situation. Therefore, this article no longer uses a static fixed value but instead adopts a set of dynamic thresholds: based on all Bitcoin transaction history data, recalibrated daily. Currently, this dynamic threshold is 114 days.

Lowering the “long-term holder” threshold may seem counterintuitive, but it is actually logical: today, more people than before are using Bitcoin as a store of value, and the point at which holders are much less likely to sell has moved forward. You don’t need to hold for a long time; the willingness to sell drops significantly at earlier points.

More Purely Realized Price

The second core indicator is the realized price, which is the average cost basis of all Bitcoin holdings across the network and is considered one of the most important on-chain metrics. However, the current algorithm that calculates “all Bitcoin on the network” includes a large amount of token balances that lack actual economic transaction attributes.

These include Bitcoin dust balances: transfer fees far exceed the token’s value, making them essentially unusable; completely unspendable transaction outputs; inscribed assets, which have surged in issuance since 2023 and do not represent meaningful value transfer transactions; newly mined block rewards, which are produced every ten minutes and are not immediately transferred. Counting these at the time of issuance or again during subsequent transfers can lead to double counting.

By removing all these invalid data points, we obtain a realized price that more accurately reflects the real market. This value is slightly higher than traditional measures because it no longer includes the costs of token holdings that are almost never circulated in reality.

More Practical and Informative Signals

The key criterion for evaluating a new analysis framework is not whether its design is elegant but whether it produces more practical and actionable insights. To this end, I have reconstructed mainstream on-chain indicators based on this new underlying logic.

Long-term holder realized profit/loss ratio (SOPR) measures whether long-term holders are selling at a profit or a loss. When calculated with the dynamic threshold and analyzed across historical quantiles, this indicator’s signal recognition improves significantly, having previously accurately predicted major market cycle tops and bottoms. Currently, its value is close to the lows seen during the $60k stage, indicating a historically very low level.

Applying this analysis method to short-term holder market value / realized value ratio (MVRV), the indicator’s effectiveness is similarly enhanced. When the indicator peaks sharply, it corresponds to major market rallies; when it dips deeply, it signals key market turning points.

Current Market Position

Using this quantile-based model on cleaned, filtered realized price data, we can generate upper and lower bands; these bands, relying solely on real on-chain fundamentals, can roughly predict Bitcoin’s price trend without the need for backtested historical data adjustments. Compared to fixed threshold models, this dynamic, adaptive analysis framework provides highly practical signals.

Summary

Bear markets are a time for focused research and refinement. Rather than simply enduring downturns, this period is an ideal window for meaningful research. Recalibrating the core on-chain analysis thresholds is a highly significant starting point.

This study is only an initial result, and all content is fully open. All formulas and data filtering rules used in this paper have been published, allowing anyone to reproduce, verify, and further optimize this model. If this analysis framework can be continuously improved, the entire crypto industry will benefit—this is the original intention of this research. Indicators built on this new dynamic threshold are more accurate, respond faster to market changes, and are more practical. Based on current readings, Bitcoin is in a deep value zone.