Smart money flows in! Decoding the three main drivers behind BTC's rebound

Author: Yuan Shan Insights

Data sources: Farside Investors, SoSoValue, Federal Reserve H.4.1 Report, CryptoQuant

On the first trading day of 2026, BTC ETF net inflow was $471 million.

What does this number mean?

In November and December, the total net outflow of spot BTC ETFs was approximately $4.57 billion; of which December alone saw a net outflow of about $1.09 billion.

Many people are frantically cutting losses above 93K, while institutions bought back about one-tenth in a single day on January 2.

At the same time, the following occurred:

• The Federal Reserve’s balance sheet increased by approximately $59.4 billion week-over-week (WALCL: as of 12/31, $6.6406 trillion, an increase of about $59.4 billion from 12/24)

• Newborn whale holdings surpassed 100,000 BTC (worth about $12 billion)

• BTC rebounded from 87.5K to 93K (+6.8%)

These three data points appearing simultaneously indicate a shift in capital flow.

The rise in 2025 relies on “narratives” (halving, ETF launch), while the rise in 2026 depends on “real money” (Fed easing, institutional subscriptions, whale accumulation).

This is the second phase of the market: shifting from emotion-driven to capital-driven.

01｜What happened: Three simultaneous signals

Signal 1: ETF reverses selling pressure

In November and December, the total net outflow of spot BTC ETFs was about $4.57 billion; December alone saw a net outflow of approximately $1.09 billion. Retail investors frantically cut losses in the 90—93K range, spreading panic.

But on January 2, BTC ETF saw a single-day net inflow of $471 million, the highest since November 11, 2025.

What does this mean? Institutions are stepping in to buy the dip created by retail selling.

More intuitive data:

BlackRock’s IBIT is currently the largest single BTC spot ETF; in terms of trading activity, IBIT is often estimated to account for nearly 70% of trading volume.

The total net assets of spot BTC ETFs are in the hundreds of billions of dollars.

US crypto ETFs have a cumulative trading volume surpassing $2 trillion.

Signal 2: Fed shifts to balance sheet expansion

In March 2022, the Fed launched QT (quantitative tightening), which lasted nearly three years. The essence of QT is to withdraw liquidity from the market, which was the fundamental reason for the plunge in all risk assets in 2022–2023.

But according to authoritative sources (Reuters, Fed reports, etc.), QT will cease/end balance sheet reduction on December 1, 2025.

Starting January, the Fed not only stops draining liquidity but begins injecting it.

The Fed’s balance sheet increased by about $59.4 billion week-over-week (WALCL: as of 12/31, $6.6406 trillion, an increase of about $59.4 billion from 12/24).

Since December, the Fed has been automatically buying short-term government bonds from the market to replenish reserves (RMP), about $40 billion in the first week; subsequent market expectations still maintain a “slow balance sheet expansion to replenish reserves” pace, but with more controllable scale.

In other words, the key turning point is from “withdrawing liquidity” to “injecting liquidity.”

Signal 3: Newborn whales accelerate accumulation

On-chain data shows that newborn whales are accumulating BTC at record speeds:

New addresses holding over 100,000 BTC, worth about $12 billion.

Tether bought 8,888 BTC (about $780 million) on New Year’s Eve 2025, with total holdings exceeding 96,000 BTC.

Long-term holders have shifted to a “net accumulation” state over the past 30 days.

But there is an important controversy: CryptoQuant’s research director points out that some “whale accumulation” data may be misleading due to exchange internal wallet consolidation. After filtering out exchange factors, truly large whale addresses (holding 100–1,000 BTC) are actually slightly reducing holdings.

The main sources of genuine buying are: newborn whales (small, dispersed addresses) + ETF institutional subscriptions.

The common point among these three signals: money is flowing in, and it’s “smart money.”

02｜Why are institutions entering when retail is cutting losses

First layer: Fed easing creates a liquidity floor

Since March 2022, the Fed has started QT, reducing its balance sheet from $9 trillion to $6.6 trillion, withdrawing a total of $2.4 trillion in liquidity.

What happened during QT?

2022: Nasdaq down 33%, BTC down 65%

2023: interest rates raised to 5.5%, FTX bankruptcy, Luna zeroed out

All risk assets are under pressure.

But by December 2025, QT officially ends. Starting January, the Fed shifts to “reserve management purchases.” This is not QE (quantitative easing), but at least liquidity is no longer flowing out; it begins to flow in modestly.

What does this mean for BTC?

Historical reference: In March 2020, the Fed launched unlimited QE, and BTC surged from $3,800 to $69,000 (+1,715%). This scale was much larger than now, but the direction has changed.

More dollars entering the market will seek high-yield assets. BTC, as “digital gold,” is a natural recipient of liquidity.

Second layer: ETF becomes the “highway” for institutional allocation

In January 2024, BTC spot ETF launched, greatly lowering the threshold for institutional BTC allocation.

• No need to learn private keys, cold wallets, on-chain transfers
• Regulatory channels, can be included in pension funds, hedge funds, family offices’ asset allocations
• Good liquidity, tradable at any time, no withdrawal restrictions

Why did outflows happen in December? Retail FOMO chasing higher prices, taking over above 93K.

Why did inflows happen in January? Institutions rationally allocating, buying on dips between 87–90K.

Key data:

• BlackRock’s IBIT holds 770,800 BTC, the largest single BTC ETF
• ETF’s cumulative trading volume exceeds $2 trillion

Before ETF launch, institutions wanting to allocate BTC had to set up cold wallets, train teams, and face regulatory risks. After ETF launch, they only need to click a few buttons in their brokerage accounts.

Third layer: The “intergenerational replacement” of newborn whales

Traditional whales (entered during 2013–2017) may have taken profits at high levels. Their costs are extremely low (a few hundred to a few thousand dollars), and 90K is an astronomical return.

But newborn whales (entered during 2023–2026) are taking over. Their costs are in the $50K–$70K range, with 90K just the starting point.

Tether’s logic is typical: since May 2023, buying BTC with 15% of profits each quarter. Regardless of whether BTC is at $60K or $40K, they keep buying. This has been executed for 10 consecutive quarters, never interrupted.

Average cost: $51,117; current price: 93K; unrealized profit exceeds $3.5 billion.

This is not luck; it’s discipline.

This is “intergenerational shift of chips,” from “early believers” to “institutional allocators.” Old whales take profits, new whales take over. The market structure becomes healthier, and holders are more dispersed.

03｜Three risks that cannot be ignored

Risk 1: Controversy over “newborn whale” data

CryptoQuant’s research director points out that recent “whale accumulation” data may be misleading: exchange internal wallet consolidation can be mistaken for “whale buying.”

After filtering out exchange factors, truly large whale addresses (holding 100–1,000 BTC) are actually slightly reducing holdings.

The main sources of genuine buying are: newborn whales (small, dispersed addresses) + ETF institutional subscriptions.

What does this mean?

Data must be discerned for authenticity; don’t believe blindly. Real buying still exists, but not as exaggerated as surface data suggests. The market’s upward movement relies more on “continuous small purchases” than “large buys.”

This is actually a good sign. It indicates a more dispersed market, less dependent on a few big players.

Risk 2: The “limited” nature of Fed balance sheet expansion

Balance sheet expansion is “technical buying to replenish reserves,” different from QE, with more controllable scale. If the market overinterprets it as a 2020-style QE expectation, disappointment will follow.

Currently, RMP is technical buying, not an active injection of liquidity into the market, and the scale is far smaller than the 2020 QE (monthly expansion over $100 billion).

In other words, liquidity improvement is limited. BTC will not “recklessly surge” like 2020–2021 (from $4K to $69K). It requires waiting for clearer monetary policy shifts (such as rate cuts or restarting QE).

2026 may be a “slow bull.”

Risk 3: The “time lag trap” between retail and institutions

Institutions buy at 87–90K, retail chases at 93K. If BTC retraces to 88K:

Institutions still profit, continue holding; retail gets trapped, panic sells.

Result: institutions buy again at the low.

This is an eternal cycle:

• Institutions view a 4-year cycle; retail focuses on 4-week fluctuations
• Institutions have discipline; retail relies on feelings
• Institutions buy against the trend; retail chase and panic sell

Data from November–December is the best proof: retail cut losses at 93K (total net outflow about $4.57 billion over two months), while institutions bought at 87K (net inflow of $471 million on January 2). Institutions profit from retail panic selling.

04｜How does this rally differ from 2025

2025 rally: Narrative-driven

Core logic: halving + ETF launch + supply shock after halving

Funding sources: retail FOMO, institutional tentative allocation

Price performance: from $25K to $73K (+192%)

Risks: after narrative realization, capital retreats (November–December net outflow of about $4.57 billion over two months)

2026 rally: Capital-driven

Core logic: Fed easing + continuous ETF inflows + newborn whale accumulation

Funding sources: long-term institutional allocation, sovereign funds, family offices

Price performance: from $87K to $93K (+6.8%, just beginning)

Advantages: capital-driven is more sustainable than narrative-driven

Key differences:

The driving force has changed: 2025 relies on “expectation,” 2026 on “real money.” Narratives can change overnight (e.g., SEC attitude shift, regulatory policy adjustments), but capital inflows are real buying.

The sustainability has changed: narratives fade (halving effect diminishes, ETF novelty wears off), but capital remains (institutional allocation is long-term, not frequent entry and exit).

Volatility differs: the capital-driven phase has smaller fluctuations. Institutions do not chase prices like retail; they have clear allocation plans and discipline.

This means 2026 may not experience the “boom and bust” like 2021, but rather a “slow bull”: climbing step by step with small pullbacks.

Retailers need to adapt to the new rhythm, avoid expecting “double overnight,” and be patient.

Looking at gold from 2019–2024, it rose from $1,300 to $2,700 (+107%) over five years. No explosive surge, but no crash either. This is characteristic of a market dominated by institutions.

05｜Three insights for us

First, learn to interpret “smart money” movements.

Don’t follow the candlestick charts blindly; follow the capital flow:

• ETF inflow = institutions are buying
• Fed balance sheet expansion = liquidity improving
• Newborn whale accumulation = long-term signal

These three indicators are more important than any technical analysis. Candlesticks can deceive (fake breakouts, washouts, false signals), but capital flow does not lie.

Second, understand the “time lag” trap.

Institutions buy when retail panics, sell when retail FOMO. If you always chase highs and sell lows, you are the one being harvested.

Learn to:

• Buy when institutions buy (even if very panicked)
• Sell when institutions sell (even if very excited)
• Use ETF flow data for judgment, not feelings

Third, 2026 may be a “slow bull,” so patience is needed.

The “explosive rise” of 2021 will not repeat. This bull market is more like:

• 5–10% monthly gains
• Lasting 12–18 months
• Ultimately reaching new highs, but with a more winding path

If you expect “get rich overnight,” you will be disappointed. But if you are patient, you may find this bull market more “comfortable” than the last. Smaller pullbacks, less daily anxiety.

Previously, BTC dropped from 69K to 15K, a 78% decline. Many cut losses at 60K, 50K, 40K, ending in total despair at 15K.

If 2026 is a slow bull, the retracement might only be 15–20%. From 90K to 75K, not from 90K to 20K. In such an environment, holding is easier, mindset more stable.

Final words: understanding institutional capital dynamics is more important than predicting prices. When you understand the flow of funds, you won’t panic when it’s time to buy, nor greed when it’s time to sell.

Retail cut losses at 93K in December; institutions added positions at 87K in January. That’s the gap.