X influencers like to point to a rising M2 chart or a weakening dollar as evidence that Bitcoin is about to explode.
These overlays provide great engagement, but flatten a much more complex relationship. They’re important, but they’re not sold in the typical simple, linear way.
It is said that printing money, which increases the world money supply M2, precedes Bitcoin price movements by about 12 weeks. The idea is that as more liquidity becomes available, it will take some time for it to be introduced into Bitcoin.
I have confirmed that the closest correlation is actually over 84 days. Therefore, the chart below uses that window as the basis for its analysis.
Liquidity and the Dollar – Two Watches, One Alarm
Bitcoin runs on two clocks: liquidity and the dollar. However, they rarely attack together.
We compiled daily price data over the past 12 months to map the interaction between Bitcoin, global M2 supply (84 days staggered), and the DXY Dollar Index.
However, this situation does not match any single rule.
Liquidity matches prices with slower changes, the dollar puts pressure on more quickly, and all three relationships are strengthened or dissolved by market regimes.
The relationship at the whole-period level is clear. The price of Bitcoin is linked to a liquidity gauge and moves in the opposite direction to the dollar.
For the entire year, the correlation between Bitcoin and M2 (84 days back) is 0.78 and the 84 day forward version (showing future prices) is 0.77, while the correlation between Bitcoin and DXY is -0.58. M2 and DXY themselves are inversely correlated at -0.71.
The series is a multi-month trend, so these numbers provide context rather than day-to-day action. On daily tape, they barely line up.
Using log returns rather than levels, the same-day correlation is 0.02 for Bitcoin vs. M2 and 0.04 for Bitcoin vs. DXY, meaning that the common maxim of dollar appreciation and Bitcoin depreciation is not a one-day phenomenon in this window. Timing exists in lag.
The daily return lag test shows two time scales. To avoid false fits, use at least 120 overlapping observations. Bitcoin returns are most correlated with past movements in the liquidity series about 6 weeks agoand Most inversely correlated with DXY’s previous movement about a month ago.
The best values within these constraints are a correlation of 0.16 when M2 leads by 42 days and -0.20 when DXY leads by 33 days.
To put it simply, Liquidity acts like slow gravity and the dollar acts like a throttleAnd both push forward with modest but measurable intensity only if the urge persists for several weeks.
The relationship between bull and bear markets
The split in the regime over Bitcoin’s 2025 high price is decisive. Before the October 6th peak, the level correlation between Bitcoin and M2 was 0.89, the correlation with forward-shifted M2 was 0.87, and the correlation with DXY was -0.58.
In the post-peak slice through November 20, the sign of liquidity reverses and the correlation for both M2 series is around -0.49, while the backlink to the dollar remains around -0.60. This pattern matches the visual overlay traders see on the chart.
On the rise, the 84-day ahead M2 line tracks the price path.
During the downswing, M2 continues to rise even as the price diverges.
Pressure on the dollar will persist in both situations.
We also created a 180-day rolling correlation panel defined as Bitcoin and M2 lagged 84 days. This captures the same sales figures in one row.
It peaks at 0.94 on December 26, 2024, then fades through the first quarter, passes near zero, and records a low of -0.16 on September 30, 2025.
The reading on November 20th is -0.12. This arc coincides with a bullish leg respecting M2’s lead, followed by a late-cycle period where dollar consolidation and positioning compress the link.
As a result, no single variable “explains” Bitcoin. Data show that relationships are conditional and change over time.
Liquidity adds a late impulse that often builds up multi-month advances when the dollar isn’t rising. That’s why the forward-shifted overlay looks exactly right before and after the turn.
If the dollar’s own trend is solid, the dollar adds a faster impulse to track Bitcoin’s drawdowns or hesitations.
When M2 and DXY match, this tendency becomes stronger and the trajectory becomes smoother.
When they conflict, the correlation breaks down, and a lag that worked one season won’t work the next.
M2 liquidity will cause a slow rise over several months, but only if the dollar is not rising.
A strong dollar puts rapid pressure on Bitcoin, cooling the rally and deepening the pullback.
So, in simple terms, this means:
To focus on timing rather than narrative, here are the core numbers from the data.
| measurement | series | window | value | Precautions |
|---|---|---|---|---|
| Abnormal level | BTC vs M2 (84 days shift) | full sample | 0.78 | 203 days |
| Abnormal level | BTC vs M2 (84 days forward) | anterior sample | 0.77 | 203 days |
| Abnormal level | BTC vs DXY | full sample | −0.58 | 203 days |
| Error in return | BTC vs M2 (same day) | full sample | 0.02 | 162 days |
| Error in return | BTC vs DXY (same day) | full sample | 0.04 | 162 days |
| best lag correction | M2 leads BTC | 42 days delay | 0.16 | n=120 |
| best lag correction | DXY leads BTC | 33 days delay | −0.20 | n=129 |
| Pre-peak level correction | BTC vs M2 (84 days shift) | Until October 6th | 0.89 | forward |
| Level correction after peak | BTC vs M2 (84 days shift) | After October 6th | −0.49 | drawdown slice |
| rolling core panel | BTC vs M2 (84 days shift) | maximum value | 0.94 | December 26, 2024 |
| rolling core panel | BTC vs M2 (84 days shift) | minimum value | −0.16 | September 30, 2025 |
| rolling core panel | BTC vs M2 (84 days shift) | latest | −0.12 | November 20, 2025 |
These numbers are in line with what chart readers would guess by eye, but with one improvement. The best rugs are not fixed.
The 84 days I chose perform well during the uptrend, but the performance declines in the second half of 2025 as the dollar strengthens.
In this sample return data, tThe strongest M2 relationship is near 6 weeks, while the dollar relationship is around 1 month. The forward overlay adds further value as a directional anchor, while the lugs are resilient.
How to interpret the data
The practical idea is to treat M2 as a compass for slow trends and DXY as a gatekeeper that can block or accelerate the path.
If the compass points north and the gate is open, the correlation will be high.
When the compass points north and the gate closes, the track will bend or stall.
For those who want to monitor these trends, two basic checks can cover most of what the sample shows.
- You should monitor the slope of the liquidity series and the slope of the dollar over 1-3 months, with returns rather than levels, and adjust before relying on the M2 overlay.
- The dominant lead around the holiday period in 2024 is not the same as the optimal lead in late 2025, so we vary the lag within a band rather than fixing it to a single number.
Both steps can be implemented using a rolling correlation on weekly returns and a simple lag search.
The key is not the slogan, but the framework.
When the dollar settles to weaken, liquidity dominates turns and multi-month trends.
When the dollar is trending upward, it tends to dominate short-term fluctuations.
Last year, both states had results and the correlation changed.