I. The Invisible Tax on the Future
Why does a dollar today feel more valuable than a dollar next year? This question, at the heart of time preference theory, has occupied economists for over a century. Eugen von Böhm-Bawerk’s Positive Theory of Capital (1889) framed it as an agio — a premium that present goods naturally command over future goods simply because humans systematically undervalue future satisfaction. Irving Fisher’s The Theory of Interest (1930) later mathematized this as the “rate of impatience”: a discount rate δ applied to all future claims.
In traditional finance, the time discount rate is embedded in interest rates, bond yields, and discount cash flow models. But in cryptocurrency — a domain of provable scarcity, on-chain timestamps, and radical individual sovereignty — time preference theory finds its purest laboratory. Every HODL decision is a marginal vote against Fisher’s rate of impatience. Every vintage coin that remains unmoved for a decade is a lived experiment in deferred gratification.
“Present goods are, in general, worth more than future goods of like kind and number.” — Eugen von Böhm-Bawerk, Positive Theory of Capital, 1889
Ludwig von Mises integrated this into his action theory: all human action is aimed at removing future uneasiness, and the time preference rate is the premium placed on present over future satisfaction. In crypto, this manifests as the choice between spending today and HODLing for tomorrow — a choice recorded immutably on-chain, timestamped, and quantifiable.
II. HODL Waves — A Living Time Preference Distribution
Perhaps the clearest visualization of time preference in crypto is Glassnode’s HODL Waves chart — a 2D surface that tracks what percentage of circulating supply last moved in each age band, from 1-day-old coins to 10+ year veterans.
| Age Band | % of BTC Supply (Apr 2025) | Interpretation |
|---|---|---|
| 1–7 days | 4.2% | Exchange flows — high time preference (spending now) |
| 1–3 months | 5.1% | Short-term speculation |
| 3–6 months | 4.8% | Mid-term traders |
| 6–12 months | 8.5% | Accumulators — lower time preference emerging |
| 1–2 years | 12.1% | Core holders |
| 2–3 years | 10.3% | Bear market survivors |
| 3–5 years | 14.7% | 2019-2020 vintage — low time preference conviction |
| 5–7 years | 9.2% | 2018 bear cohort — tested by volatility |
| 7–10 years | 9.8% | 2015–2017 vintage — deeply patient capital |
| 10+ years | 13.1% | Ultra-low time preference — Satoshi era and early adopters |
Data source: Glassnode UTXO Age Distribution, April 2025
The 10+ year band — coins from the Satoshi era (2010–2015) — has grown from approximately 2% of supply in 2020 to over 13% in 2025. This represents a secular shift toward lower time preference across the entire Bitcoin network. These coins have survived:
- The Mt. Gox collapse (2014)
- The Chinese ban (2017)
- The COVID crash (2020)
- The FTX contagion (2022)
- Multiple 70%+ drawdowns
Each survived volatility event represents a positive time preference update — a reaffirmation that future value exceeds present liquidity.
III. Coin Days Destroyed — The Velocity of Conviction
Coin Days Destroyed (CDD), a metric introduced by Bitcoin Magazine in 2012, measures the economic weight of moving old coins. CDD = sum of (number of coins × days since each coin last moved). When old coins move, CDD spikes; when they stay still, CDD compresses.
| Period | Avg. Daily CDD (BTC) | Interpretation |
|---|---|---|
| 2013 (peak speculation) | 50M+ | Old coins were being spent — high time preference |
| 2017 (bull run) | 25–35M | Moderate old-coin spending |
| 2021 (peak) | 18–25M | Lower — HODL culture maturing |
| 2024–2025 | 12–18M | Compressed — historic lows in old-coin velocity |
Source: Glassnode Coin Days Destroyed, 2025
The compression of CDD over time tells a clear story: network-wide time preference is falling. The velocity of old coins — the rate at which long-term holders capitulate — has reached historic lows. This is consistent with Fisher’s framework: as the perceived future value of an asset rises, the “rate of impatience” falls because the penalty for spending early becomes steeper.
This is not merely a Bitcoin phenomenon. CDD compression across Litecoin and Dogecoin (on a per-unit basis adjusted for supply size) shows a similar, if less dramatic, trend.
IV. The Time Discount Rate of Crypto vs. Traditional Assets
In Irving Fisher’s framework: Present Value = Future Value × e^(−δt) where δ is the time discount rate. For vintage coins, we can reverse this: if we know the acquisition price and the peak price, we can calculate the implicit δ that a HODLer effectively ignored.
| Vintage | Coin | Acquisition Price | Peak Price | Holding Period | Annualized Return | Implied δ (Discounted) |
|---|---|---|---|---|---|---|
| 2010-07 | BTC | ~$0.008 | ~$73,000 (2024) | 14 years | ~180% | Very negative — effectively ~−180% |
| 2011-08 | BTC | ~$8 | ~$73,000 | 13 years | ~120% | ~−120% |
| 2011-12 | LTC | ~$0.30 | ~$380 (2021) | 10 years | ~106% | ~−106% |
| 2013-12 | DOGE | ~$0.0002 | ~$0.73 (2021) | 8 years | ~175% | ~−175% |
| 2013-12 | BTC | ~$800 | ~$73,000 | 11 years | ~51% | ~−51% |
| 2015-01 | BTC | ~$200 | ~$73,000 | 9 years | ~91% | ~−91% |
| 2017-12 | BTC | ~$19,000 | ~$73,000 | 6 years | ~23% | ~−23% |
What does this mean? In Fisher’s framework, a positive δ represents impatience — discounting the future. A negative δ means the future has outperformed any reasonable discount rate. A 2010 Bitcoiner who held for 14 years achieved a negative discount of approximately 180% per year — in other words, they were rewarded far beyond any traditional time preference model.
Compare this with traditional assets: the S&P 500 has delivered ~10% annualized over the same period. A 30-year US Treasury yields ~4-5%. The time discount that crypto HODLers have been willing to accept (effectively zero, or even negative) would be irrational in any traditional finance framework — yet the data shows it has been the dominant strategy.
This suggests that crypto time preference is endogenous: the act of HODLing itself creates upward pressure on price, which in turn lowers time preference, creating a self-reinforcing cycle. This aligns with the Austrian concept of time preference learning — agents update their time discount rates based on observed outcomes.
V. Monetary Policy Shapes Time Preference: BTC vs. LTC vs. DOGE
Different networks show different time preference distributions, reflecting their monetary policies:
| Asset | Inflation Rate | Supply Cap | % Supply Unmoved 5+ Years | Velocity Trend |
|---|---|---|---|---|
| Bitcoin | ~0.8% (2025) | 21M | ~30% | Declining (lower time pref) |
| Litecoin | ~3.5% (2025) | 84M | ~12% | Stable |
| Dogecoin | ~3.9% (2025) | ∞ (Inflationary) | ~5-8% | High (higher time pref) |
Data sources: CoinMetrics, Messari, Glassnode (2025)
Key insight: Bitcoin’s disinflationary schedule produces the lowest time preference — 30% of supply has not moved in 5+ years. Litecoin, with a similar halving structure but larger supply, shows moderate HODL behavior. Dogecoin, with its fixed yearly inflation of ~5 billion DOGE, creates structural pressure toward higher time preference — holding DOGE means accepting dilution from new supply.
This is a natural experiment in time preference formation: same timestamp technology, same proof-of-work security model, but different monetary policies produce radically different time preference distributions across their holder bases.
VI. The Realized Cap HODL Ratio — A Conviction Thermometer
The Realized Cap HODL Ratio (RHR), developed by analyst Woonomic in 2021, divides the realized capitalization of HODLers (UTXOs aged 6+ months) by the realized capitalization of traders (UTXOs under 6 months).
$$ RHR = \frac{\text{Realized Cap of HODLers}}{\text{Realized Cap of Traders}} $$
| RHR Range | Interpretation |
|---|---|
| < 5 | Speculative frenzy — traders dominate |
| 5–10 | Mixed sentiment |
| 10–20 | HODL conviction building |
| 18–22 (2024–2025) | Deep conviction — HODLers value 18–22× traders’ unrealized value |
| > 30 | Extreme conviction (possible cycle top in some models) |
Source: Look Into Bitcoin, Realized Cap HODL Ratio, 2025
An RHR above 18 means the cumulative cost basis of long-term holders is 18 times larger than that of short-term traders. This is not a measure of how many people hold, but of the total capital weighted by time preference — a direct economic measure of how much value the market places on patience.
VII. The Vintage Premium as a Time Preference Signal
We can now define the vintage premium — the price premium that older coins command over newer coins — as a direct function of time preference:
Vintage Premium = f(Time Preference, Holding Duration, Network Monetary Policy)
Where holding duration is the independent variable with the largest explanatory power. Coins that have survived more market cycles carry:
- Survivorship proof — they have passed through multiple drawdowns without being spent
- Supply irreversibility — those coins can never be recreated at the same age
- Timestamp scarcity — the timestamp itself is a non-fungible coordinate in time-space
This is why a 2010 Bitcoin and a 2024 Bitcoin, though identical in protocol terms, are not identical economic goods. The 2010 coin carries 14 years of time preference proof — an economic signal that cannot be replicated.
VIII. Conclusion: HODLing as a Time Preference Arbitrage
Time preference theory, forged in the 19th century by Böhm-Bawerk and refined by Fisher and Mises, finds its most vibrant 21st-century expression in the blockchain timestamp. Every UTXO age band is a revealed preference — a market participant voting with their time horizon.
The data is unequivocal:
- 10+ year Bitcoin coins are growing as a share of supply
- CDD is compressing to historic lows
- The RHR remains in deep conviction territory
- Vintage coins have outperformed every other asset class on a time-adjusted basis
The most patient participants are rewarded with the highest time-discount arbitrage. This is not speculation — it is a structural property of a system with fixed supply, provable timestamps, and rational agents who learn to optimize their time preference.
The final lesson from time preference theory applied to crypto is simple: time is the scarcest input in the system. The coins that have absorbed the most of it carry the highest economic weight.
— Encryption Archive · TimeB.news