# Cross-Asset Lead-Lag Research: BTC → Altcoin Cascade

## Executive Summary

**Key Finding:** Bitcoin leads altcoins by **16-118 seconds** (not milliseconds), with an average of **57 seconds**. This window is exploitable without HFT infrastructure.

---

## 1. Beta Coefficients (Volatility Multiplier)

When BTC moves X%, altcoins move β × X%:

| Asset | Beta (vs BTC) | Interpretation | Source |
|-------|---------------|----------------|--------|
| **BNB** | 1.34 | Moves 34% more than BTC | JamesBachini |
| **SOL** | 1.98 | Moves ~2x BTC | JamesBachini |
| **MATIC** | 2.07 | Moves ~2x BTC | JamesBachini |
| **AVAX** | 2.38 | Moves ~2.4x BTC | JamesBachini |
| **FTM** | 2.68 | Moves ~2.7x BTC | JamesBachini |
| **MANA** | 3.79 | Moves ~3.8x BTC | JamesBachini |
| **SAND** | 3.45 | Moves ~3.5x BTC | JamesBachini |
| **SHIB** | 3.57 | Moves ~3.6x BTC | JamesBachini |
| **GALA** | 4.66 | Moves ~4.7x BTC | JamesBachini |

### Tiered Beta Categories

```
Tier 1 (Low Beta, 1.0-1.5):  BNB, ETH
Tier 2 (Medium Beta, 1.5-2.5): SOL, MATIC, AVAX, LUNA
Tier 3 (High Beta, 2.5-4.0):   FTM, MANA, SAND, SHIB
Tier 4 (Extreme Beta, 4.0+):   GALA, ENS, SPELL
```

**Implication:** If BTC drops 2%, SOL drops ~4%, GALA drops ~9%.

---

## 2. Correlation Coefficients

How tightly altcoins track BTC direction:

| Pair | Correlation | Timeframe | Source |
|------|-------------|-----------|--------|
| BTC-SOL | 0.99 | Recent 2024 | DeFiLlama |
| BTC-XRP | 0.86-0.92 | 3-month | BitInfoCharts, Coinbase |
| BTC-DOGE | 0.38-0.87 | Varies | BitInfoCharts, DeFiLlama |
| BTC-ETH | 0.61-0.85 | 90-day 2024 | Coinbase Research |
| BTC-LTC | 0.30 | 3-month | BitInfoCharts |
| BTC-BCH | 0.05 | 3-month | BitInfoCharts |

### Correlation Variability

- **High volatility periods:** Correlations spike toward 1.0 (everything moves together)
- **Low volatility periods:** Correlations diverge (more independent movement)
- **Trend:** Correlations have increased over time as market matures

---

## 3. Lead-Lag Time Measurements

### Empirical Research: BTC → ADA (Tick-by-Tick Study)

Source: [Business Perspectives - Lead-Lag Duration Study](https://www.businessperspectives.org/index.php/journals/investment-management-and-financial-innovations/issue-421/a-tick-by-tick-level-measurement-of-the-lead-lag-duration-between-cryptocurrencies-the-case-of-bitcoin-versus-cardano)

| Metric | Value |
|--------|-------|
| **Minimum lag** | 16 seconds |
| **Maximum lag** | 118 seconds |
| **Average lag** | 57 seconds |
| **Trend** | Declining over time (markets getting efficient) |
| **Period studied** | Jan 2019 - May 2021 |

### Estimated Lag Times by Asset (Extrapolated)

Based on liquidity, market cap, and trading volume:

| Asset | Estimated Lag | Confidence | Reasoning |
|-------|---------------|------------|-----------|
| **ETH** | 1-5 seconds | High | 2nd largest, highest liquidity after BTC |
| **BNB** | 5-15 seconds | Medium | Large cap, Binance native advantage |
| **SOL** | 10-30 seconds | Medium | High volume, but less than ETH |
| **XRP** | 15-45 seconds | Medium | Large cap but variable liquidity |
| **DOGE** | 20-60 seconds | Medium | Retail-driven, slower price discovery |
| **ADA** | 16-118 seconds | High | Empirically measured |
| **Small caps** | 60-300 seconds | Low | Low liquidity = slow adjustment |

### Comparison to Traditional Markets

| Market | Lead-Lag Time | Source |
|--------|---------------|--------|
| Crypto (BTC→ADA) | 16-118 seconds | Academic research |
| US Equities (2000-2005) | Few seconds | Market microstructure research |
| US Equities (2021-2022) | <10 milliseconds | Market microstructure research |

**Key Insight:** Crypto markets are ~10,000x slower than equity markets for price discovery.

---

## 4. Cascade Model

### Theoretical Price Propagation

```
T+0ms:     BTC drops 1% on Binance
T+500ms:   BTC drop confirmed across exchanges
T+1-5s:    ETH begins adjusting (β=0.85 → drops 0.85%)
T+5-15s:   BNB, SOL begin adjusting
T+15-45s:  XRP, DOGE, mid-caps adjusting
T+45-120s: Small caps, low-liquidity tokens adjusting
```

### Factors Affecting Lag Duration

1. **Liquidity depth** - More liquid = faster adjustment
2. **Market cap** - Larger = faster (more eyes on it)
3. **Exchange concentration** - If traded mainly on one exchange = slower
4. **Time of day** - Asian hours may have different dynamics
5. **Volatility regime** - High vol = faster adjustment (more attention)

---

## 5. Practical Strategy Parameters

### For Your System (SOL, XRP, DOGE, ETH focus)

| Parameter | BTC→ETH | BTC→SOL | BTC→XRP | BTC→DOGE |
|-----------|---------|---------|---------|----------|
| Beta | 0.85 | 1.98 | ~1.0 | ~1.5 |
| Correlation | 0.85 | 0.99 | 0.90 | 0.70 |
| Est. Lag | 2-5s | 10-30s | 15-45s | 20-60s |
| Confidence | High | Medium | Medium | Low |
| Opportunity | Small | Good | Good | Risky |

### Recommended Thresholds

```python
# Minimum BTC move to trigger (avoids noise)
MIN_BTC_MOVE_PCT = 0.3  # 30 bps

# Maximum lag window to hold position
MAX_HOLD_TIME_S = {
    "ETH": 10,
    "SOL": 45,
    "XRP": 60,
    "DOGE": 90,
}

# Expected altcoin response
EXPECTED_RESPONSE = {
    "ETH": lambda btc_move: btc_move * 0.85,
    "SOL": lambda btc_move: btc_move * 1.98,
    "XRP": lambda btc_move: btc_move * 1.0,
    "DOGE": lambda btc_move: btc_move * 1.5,
}
```

---

## 6. Risk Factors

### Why This Might Not Work

1. **Lag is declining** - Study showed statistically significant trend toward shorter lags
2. **HFT competition** - More firms entering crypto HFT
3. **Correlation breakdown** - In stress events, correlations can invert
4. **Execution risk** - You might not get filled at expected price
5. **Wrong direction** - Beta and correlation aren't 100%

### When It Works Best

1. **Flash crashes** - Sudden BTC moves create longer lag windows
2. **News events** - Fed, CPI, ETF announcements
3. **Liquidation cascades** - Leveraged traders getting liquidated
4. **Low liquidity hours** - Asian session overlap gaps

---

## 7. Data Sources for Live Implementation

| Data Point | Source | Endpoint |
|------------|--------|----------|
| BTC price leader | Binance | `@bookTicker` (already built) |
| Altcoin prices | Binance | Same endpoint, add symbols |
| Historical correlation | Coin Metrics | API available |
| Real-time correlation | Calculate from tick stream | Rolling window |

---

## References

1. [Coinbase Institutional - Crypto Correlations August 2024](https://www.coinbase.com/institutional/research-insights/research/monthly-outlook/monthly-outlook-august-2024)
2. [James Bachini - 100 Altcoin Betas](https://jamesbachini.com/crypto-market-volatility/)
3. [BitInfoCharts - Correlation Matrix](https://bitinfocharts.com/correlation.html)
4. [Lead-Lag Duration BTC-ADA Study](https://www.businessperspectives.org/index.php/journals/investment-management-and-financial-innovations/issue-421/a-tick-by-tick-level-measurement-of-the-lead-lag-duration-between-cryptocurrencies-the-case-of-bitcoin-versus-cardano)
5. [DeFiLlama - BTC-SOL 0.99 Correlation](https://cryptopotato.com/defillama-crypto-correlations-hit-record-highs-as-btc-sol-reaches-0-99/)
6. [Springer - Profiting from Crypto Pair Correlations](https://link.springer.com/chapter/10.1007/978-3-031-68974-1_16)
7. [SSRN - Latency Arbitrage in Crypto 2025](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5143158)