1) The Setup

Two facts sit uneasily together. Global liquidity, on the broadest available aggregate, expanded last week to a record $192.4 trillion, with three-month annualised growth steady at 7.8 per cent and annual growth at 7.0 per cent [1]. Over the same window, Bitcoin was trading in the high-$70,000s, around forty per cent below its October 2025 cycle high near $126,000 and, under the CF Benchmarks framework earlier in 2026, roughly 1.3 standard deviations below liquidity-implied fair value [6].

The problem is not whether liquidity matters. The problem is how to use it without pretending that every change in Global M2 mechanically causes the next Bitcoin move. Some investors hold a long-run liquidity argument that treats Bitcoin as a global money proxy but says little about cycle timing. Others treat the drawdown as a technical event largely separate from liquidity. Neither is sufficient for an allocator deciding when to size a long-duration position.

This memo sets out a narrower claim. The indicator described here is a timing overlay, not a causal proof. It translates a recurring lead-lag observation in the cited literature into a weekly monitoring rule, then asks whether the rule helps identify inflection points inside the longer-cycle thesis developed in the April note [7]. The tool is useful because it either confirms the existing liquidity-cycle bias or challenges it early enough to keep implementation disciplined. That longer-cycle thesis continues to hold: a Q3-to-Q4 capitulation risk, a structural ETF bid emerging from weakness, and allocation-driven flow eventually dominating arbitrage-driven flow as post-MSBT market infrastructure matures [7].

At the time of writing, the series reads 0.33, the first higher print since the February 2026 peak. We do not read it as confirmation. The historical evidence is small: five selected inflection observations across one cycle, plus one live read. That is enough to justify a disciplined watch-list; it is not enough to publish a false-positive rate, a confidence interval, or a deterministic price path. Confirmation and invalidation criteria are set out below, and both outcomes remain live.

2) The Liquidity Backdrop

In the week ending 19 May 2026, global liquidity expanded to $192.4 trillion, with three-month annualised growth at 7.8 per cent and annual growth at 7.0 per cent [1]. Beneath that record headline, the expansion is concentrated in two sources: Federal Reserve liquidity injections and suppressed bond-market volatility [1]. The Shadow Monetary Base, which combines central-bank liquidity with bond collateral and is the bedrock of broad-based liquidity growth, fell back to $110.1 trillion in the same week, contracting at a 6.3 per cent three-month annualised rate against an annual rate of plus 2.9 per cent [1]. The gap between the headline aggregate and the Shadow Monetary Base points to a narrowly driven upturn rather than a system-wide expansion.

The collateral multiplier sits between those two series. It converts central-bank liquidity into Shadow Monetary Base and is sensitive to bond-market volatility through the haircuts mechanism. The MOVE index, averaging near 70 in recent weeks and close to four-year lows, has been supportive. The latest print at 86 is the first reading in some time that pressures the channel [1]. Dollar strength against most majors bar the yuan is a drag on non-US liquidity in dollar terms; PBoC liquidity remains weak; the ECB, Bank of England and Bank of Japan continue, on net, to subtract from global liquidity at the margin [1].

These data are consistent with a mature liquidity cycle in which the headline aggregate is sustained by discretionary policy levers rather than broad-based money creation [1, 2]. The simultaneous May sell-off in long bonds and weakness in gold is the kind of joint move that, in prior cycles, has been associated with deteriorating liquidity transmission rather than orderly repricing [3]. With drivers concentrated in Fed injections and bond-volatility suppression, BTC remains exposed if either lever weakens.

3) The Signal

In simple terms, the indicator takes the Global M2 series, shifts it forward by 13 weeks, and charts the 13-week rate of change of that forwarded series against BTC. The purpose is not to predict a precise price level, but to assess whether liquidity momentum is improving or deteriorating before BTC confirms the turn.

The signal is simple enough to monitor weekly, but the definitions matter.

The choice of variable matters. Liquidity can sit at a record level while its underlying drivers deteriorate, as the live divergence between headline liquidity and the Shadow Monetary Base demonstrates [1]. The level alone does not carry the directional information an allocator needs. A simple weekly change improves on the level but can smooth through turning points. ROC(13) captures the acceleration or deceleration of the forwarded liquidity series over a 13-week window. In this small exploratory sample, it fits the selected inflection observations better than the tested alternatives: the level led none of the five, the twelve-week change led three, and ROC(13) led all five if the 17-bar near-band observation is included as a close exception. This should be read as a working specification, not as a proven structural law. The indicator is useful because it creates a repeatable weekly discipline: wait for confirmation, define failure in advance, and separate timing utility from a broader liquidity story.

4) The Inflection Record

Across the 2023-2026 sample, inflections in the ROC series have led BTC directional turns by roughly 13 to 17 weekly bars. The target operating window remains 13 to 16 bars; the November 2023 observation is retained as a near-band 17-bar exception rather than hidden. The table below is a working summary from the live model and the dates already stated in the memo and chart annotations. It is not presented as a tick-level audit trail.

The terminology is straightforward. A ROC trough is the low point in liquidity momentum; when the ROC stops falling and turns higher, it suggests liquidity momentum is improving. A ROC peak is the high point in liquidity momentum; when the ROC stops rising and turns lower, it suggests liquidity momentum is deteriorating. The table should be read from left to right: first identify the liquidity-momentum turn, then look forward by the historical lag window to assess whether BTC subsequently turned in the same direction.

The logic is directional and sequential, not a precise BTC price-targeting exercise. In this sample, troughs in forwarded-liquidity momentum were followed by constructive BTC windows, while peaks were followed by weaker BTC windows. The February 2026 read matters because it was identified in real time and BTC weakness arrived inside the specified window; one live observation, however, still does not validate the model by itself.

5) The February Read

The February 2026 inflection is the concrete live case.

ROC of forwarded Global M2 peaked in the first week of February and rolled over. On a 13-to-16 bar lead, the implied projection was for BTC directional weakness between mid-May and mid-June 2026. By the time that ROC peak was visible, Bitcoin was already well below its early-October 2025 cycle high near $126,000; the value of the indicator was not in calling that high, but in keeping the mid-May to mid-June weakness window in view. BTC traded heavily through February
as the U.S. dollar strengthened, weakened further through April as the equity correlation identified in last month's memo [7] tightened against fragile dollar funding conditions, and broke lower in May. A spot print cited in Howell's 19 May note placed BTC around $76,452, consistent with a renewed leg lower in the high-$70,000s range [1].

The first live read was consistent with the indicator's projected timing window. That is the right level of confidence. The February peak was visible at the time, the 13-bar lead was the stated forward read, and BTC selling arrived in the window. The §7 bear case explains why one live observation should not change priors. It does, however, justify taking the current print seriously and defining the next decision rule before the data arrive.

6) Where the Chart Sits Today

From the February peak, the indicator declined through the early spring and registered its first higher weekly print at 0.33 in the week ending 19 May 2026 [1]. This is the reading visible on the live model.

The print is not a confirmed bottom. It is one higher tick after a sustained decline, and from a single tick a potential base and a continuing down-leg are indistinguishable. Confirmation requires two to three consecutive higher weekly readings with a steepening slope; invalidation requires the indicator to break below its recent low or roll lower for two or more weeks. Between those outcomes sits the present state, consistent with either.

Figure 2. The Three States Framework. Current state and conditional branches for the forwarded-liquidity ROC tool.

The relevant portfolio response is therefore conditional, not pre-committed. If the next two or three weekly prints confirm, the August-September inflection trade would warrant consideration under the existing staged-entry plan [7]. If the series rolls over, the same framework would argue for patience, continued defensive posture, and preparation for a deeper Q4 entry framework. Implementation remains subject to eligibility, mandate, liquidity, and risk limits.

7) The Other Side

The strongest reading against treating 0.33 as a turn is straightforward.

The headline liquidity reading of $192.4 trillion is real, but its drivers are concentrated in Federal Reserve liquidity injections and the suppression of bond-market volatility [1]. Both are discretionary policy levers rather than broad-based expansion, and both can reverse on a timescale shorter than the lag the indicator monitors. The Shadow Monetary Base contracted at a 6.3 per cent three-month annualised rate in the latest week [1]. The MOVE index, after averaging near four-year lows around 70, moved to 86 in the latest data; a sustained break higher would pressure the collateral-multiplier tailwind supporting the headline aggregate [1]. Outside the U.S., PBoC liquidity is weak, the BoJ remains on QT, and the ECB and Bank of England continue to subtract at the margin [1].

The divergence between the headline aggregate and the Shadow Monetary Base is the sharpest challenge to the constructive read. When the headline is at a record while the Shadow Monetary Base is contracting, the system is signalling that the marginal source of liquidity is fragile. Two drivers that depend on continued Treasury issuance behaviour and continued suppression of bond volatility cannot be assumed to persist. That is the setup in which forwarded liquidity momentum can roll lower and the ROC indicator can invalidate before it confirms.

The indicator itself also carries risk the macro frame does not capture. A false positive would look like one or two higher ROC prints followed by a re-acceleration to a new low, with BTC failing to follow the implied lag. On the in-sample evidence we have, that has not happened across the selected observations; on a sample of five, it cannot be ruled out. We do not attach a number to the false-positive probability. The sample does not support one.

If the bear case plays out, present BTC weakness could extend through Q3 and into a deeper Q4 2026 downside window. That outcome is consistent with the longer-cycle thesis developed in prior memos [7]. The structural ETF bid identified there continues to absorb, but at lower marginal prices, against a holder cohort whose cost basis sits in the high-$70,000s and which was carrying an aggregate mark-to-market loss of about seven per cent in early 2026 [8]. If Q4 delivers that deeper entry window, the framework would support reassessing long-duration exposure at a lower cost basis, subject to the relevant mandate and risk limits.

On sample size: five inflection observations across one cycle is the strongest data we have, and it is not strong enough to publish a false-positive rate or a confidence interval on the 13-to-16 bar lead. The CrossBorder Capital methodology cited in [1] extends further back in usable form, and we intend to test ROC(13) on the 2014-2022 window in subsequent work. Until that is complete, the 13-to-16 bar rule should be read as a working estimate that has held inside one cycle, not a parameter with stated error bars.

8) Why the Framework Remains Useful Either Way

The framework remains useful because both outcomes are decision-relevant, not because it removes uncertainty. It either confirms the existing liquidity-cycle bias or challenges it early enough to keep implementation disciplined.

On the constructive path, the 0.33 reading is followed over the next two to three weeks by higher prints and a steepening slope. That tells us forwarded liquidity momentum has turned. On a 13-to-16 bar lead, the BTC inflection window falls between mid-August and mid-September 2026. Subject to eligibility, mandate, liquidity, and risk limits, confirmation would warrant consideration of staged exposure under the existing plan [7]. Further scaling would depend on BTC following the indicator on the expected lag.

On the alternative path, the reading rolls back down within two to three weeks, takes out the recent low, and re-accelerates negative. Forwarded liquidity remains in a contraction regime, the late-stage picture set out in §7 holds, and BTC weakness extends through Q3 into a potential deeper Q4 downside window. Invalidation would argue for patience, continued defensive posture, and preparation for the deeper-discount entry framework outlined for Q4. The cycle thesis is preserved; only entry timing shifts.

Either outcome informs the same long-duration thesis but points to a different prospective entry window. Confirmation moves the relevant window to August-September. Invalidation moves it to Q4. The discipline is to wait for two to three more weekly prints before changing the implementation read.

9) Four Studies, Three Types of Evidence

The 13-to-16 bar lead is the central operational claim, and the fair question is whether it is more than an in-sample fit. The cited work helps, but it does not all validate the same object. Howell and TimeXer are lag and timing evidence. Onramp is regime and transmission evidence. CF Benchmarks is valuation and fair-value context. Together they make the multi-month window plausible; they do not prove the exact 13-to-16 bar ROC rule.

• Lag / timing — Howell, Capital Wars [1, 2]. Practitioner work on the CrossBorder Capital GL Indexes weekly liquidity series supports a rule of thumb near three months. For the monitoring framework, it makes a 13-week forward shift a plausible operating assumption. The limitation is that it is a practitioner framework, not a formal test of the exact ROC(13) rule.

• Lag / timing — TimeXer, Karthick & Moharir [4]. The TimeXer work uses a transformer model with Global M2 from 18 major economies and daily BTC, with a 12-week lag and long-horizon forecast improvement. For the monitoring framework, it places independent timing work in the same neighbourhood as the 13-to-16 bar window. The limitation is that it uses a different data frequency, model class, sample, and forecast target.

• Regime / transmission — Onramp Institutional [5]. Onramp's regime work shows the BTC-liquidity relationship strengthening at 6-to-24-month horizons and being weaker at one month. For the monitoring framework, it supports the idea that liquidity effects operate through regimes and lags, not same-week causality. The limitation is that this is regime classification, not a 13-to-16 bar inflection model.

• Valuation / fair value — CF Benchmarks [6]. CF Benchmarks provides the M2-implied fair-value methodology and Z-score context for BTC relative to liquidity. For the monitoring framework, it frames how far spot sits from liquidity-implied fair value. The limitation is that this is valuation context, not direct validation of the timing signal.

This is the bridge we should rely on. The exact 13-to-16 bar rule remains an in-sample operational specification. The broader multi-month lag is supported by independent work, and that is enough to make the indicator worth monitoring with defined confirmation and invalidation rules.

10) What We're Watching

Table 2 sets out the decision weights for the next four to eight weeks. ROC is primary. Shadow Monetary Base and MOVE can veto or condition a constructive ROC read because they speak directly to the breadth and durability of liquidity transmission. DXY, real yields, PBoC liquidity, and the weekly Global M2 headline are context variables. The cadence is weekly unless the underlying source updates at a different frequency.

Primary and veto variables

Context variables

11) Where We Stand

The memo's purpose is to define the indicator and the decision discipline around it. The longer-cycle thesis articulated since early 2026 continues to hold [7]. The forwarded-liquidity chart is how we track that thesis in motion. Its components are drawn from the cited literature; the contribution is the synthesis, the weekly rule, and the discipline not to treat one tick as a turn.

The 0.33 print is the first data point of the next test. It is not a confirmed turn. Confirmation requires two to three higher weekly prints with a steepening slope. Invalidation requires the series to roll lower or break its recent low. Either path is decision-relevant, but neither should change the implementation read before the signal resolves.

The February projection was directionally consistent with the live window. The current print decides nothing yet.

Disclaimer

This material is provided for information purposes only and does not constitute investment advice, a recommendation, or an offer or solicitation to buy or sell any financial product or service in any jurisdiction.

The views expressed are those of the author as at the date of publication and are subject to change without notice. While reasonable care has been taken, no representation or warranty, express or implied, is made as to the accuracy, completeness, or reliability of the information contained herein, and no reliance should be placed upon it.

This material has not been reviewed or approved by the Dubai Financial Services Authority, the British Virgin Islands Financial Services Commission, or the Cayman Islands Monetary Authority, or any other regulatory authority.

This material is not intended for distribution to, or use by, any person in any jurisdiction where such distribution or use would be contrary to local law or regulation. In particular, this material does not constitute a public offer of securities in the British Virgin Islands or the Cayman Islands.

Any investment or service referred to herein will be available only to persons who meet the relevant eligibility criteria under applicable laws and regulations, including, where applicable, Professional Clients (as defined by the Dubai Financial Services Authority) or other equivalent categories of sophisticated or professional investors under applicable law.

Investments involve risk, including the possible loss of capital. Past performance is not a reliable indicator of future results.

“Axys” is a trading name used by a network of independent businesses. Axys does not operate as, and is not, a separate legal entity in any jurisdiction and has no distinct legal personality. Each business operating under the Axys name is independently owned and operated.

Axys Capital  Ltd is authorised and regulated by the Dubai Financial Services Authority. Axys Investment Management Ltd is authorised and regulated by the British Virgin Islands Financial Services Commission.

References

[1] Michael Howell, "Global Liquidity Watch: Weekly Update — Liquidity Expands, But Drivers Remain Concentrated," Capital Wars Substack, 19 May 2026.
URL: https://capitalwars.substack.com/p/global-liquidity-watch-weekly-update-008
[2] Michael Howell, "Are We Still 'Late' Cycle?" Capital Wars Substack, 10 May 2026.
URL: https://capitalwars.substack.com/p/are-we-still-late-cycle-crypto-bonds
[3] Michael Howell, "Are The Bond Sell-Off And The Drop In Gold Warning Signs?" Capital Wars Substack, 17 May
2026.
URL: https://capitalwars.substack.com/p/are-the-bond-sell-off-and-the-drop
[4] M. Karthick and A. Moharir, "Expert System for Bitcoin Forecasting: Integrating Global Liquidity via TimeXer
Transformers," arXiv:2512.22326, December 2025.
URL: https://arxiv.org/abs/2512.22326
[5] Onramp Institutional, "Bitcoin's Macro Liquidity Cycle," Q4 2025 / January 2026.
URL: https://onrampbitcoin.com/research/bitcoins-macro-liquidity-cycle
[6] CF Benchmarks Research, "The M2-Bitcoin Relationship: What the Data Actually Shows," March 2026.
URL: https://www.cfbenchmarks.com/blog/the-m2-bitcoin-relationship-what-the-data-actually-shows
[7] Axys Research, "A Different Bitcoin: The Post-ETF Structural Bid," April 2026 Monthly Research Note. Internal reference; prior memo in this series.
[8] Bianco Research, aggregate mark-to-market analysis of the Bitcoin holder cohort, early 2026; figures as cited in
the Axys April 2026 internal memo [7].