4.4 Correlations

Understand how individual stocks correlate with the broader market, indices, sectors, currencies, and bonds. Use correlations to identify true edges and avoid hidden risks.

Layer 4: Market Structure & Context — Chapter 4 Goal: Understand how individual stocks correlate with the broader market, indices, sectors, currencies, and bonds. Use correlations to identify true edges and avoid hidden risks.

The Core Idea

Most stocks move with the broader market most of the time. When SPY goes up, most stocks go up. When SPY drops, most stocks drop. The "tide" lifts or sinks the ships.

Understanding correlations is essential because:

Your individual stock trade is largely a bet on the broader market
"Diversification" across correlated stocks isn't really diversification
Some moves are sector or macro driven, not stock-specific
Cross-asset correlations (DXY, TLT) provide context

What Is Correlation?

A statistical measure of how two variables move together, ranging from -1 to +1.

+1.0: Perfectly correlated (move identically)
0: No correlation (independent)
-1.0: Perfectly inversely correlated (opposite moves)

Practical Categories

Strong positive: > +0.7
Moderate positive: +0.3 to +0.7
Weak: -0.3 to +0.3
Moderate negative: -0.3 to -0.7
Strong negative: < -0.7

Stock-to-SPY Correlations

This is the most important correlation to understand for any stock you trade.

Typical Correlations

Tech mega-caps (AAPL, MSFT): ~0.7-0.85 with SPY
Small-cap stocks (in IWM): ~0.7-0.8 with SPY
Defensive names (KO, PG): ~0.5-0.7 with SPY
Utilities (XLU stocks): ~0.4-0.6 with SPY
Gold miners (GDX): -0.2 to +0.2 with SPY (low)
Bonds (TLT): -0.2 to -0.4 with SPY (often inverse)

Implication

If you're long AMD, ~70% of your daily return comes from SPY's direction, not AMD-specific events.

When you "buy AMD" you're really buying:

70% SPY exposure
20% XLK (tech) exposure
10% AMD-specific exposure

Beta: A Refined Correlation Measure

Beta measures how much a stock moves relative to the market.

Formula

β = Cov(Stock, Market) / Var(Market)

In plain terms: if SPY moves 1%, the stock typically moves β%.

Beta Values

Beta	Meaning
0	No relation to market
0.5	Moves half as much as market
1.0	Moves with the market
1.5	Moves 1.5× the market
2.0	Moves 2× the market
Negative	Moves opposite to market

Examples

NVDA: β ≈ 1.7-2.0 (high-beta, amplified moves)
AAPL: β ≈ 1.2 (slightly higher than market)
WMT: β ≈ 0.5 (defensive, lower volatility)
GLD: β ≈ 0.0-0.2 (gold is largely uncorrelated)
TLT: β ≈ -0.3 (bonds often move opposite stocks)

Trading Implications

High-Beta Stocks

Amplify market moves
Bigger gains in rallies
Bigger losses in declines
Need wider stops (more volatile)
Smaller position sizes

Low-Beta Stocks

Steadier, less volatile
Smaller gains and smaller losses
Tighter stops possible
Can hold larger positions

Why Beta Matters

Your account-level risk depends on the betas of your positions. If you hold 5 high-beta tech stocks, you don't have 5 different bets — you have one giant high-beta bet on tech.

Sector-to-Sector Correlations

Sectors have predictable correlation patterns.

Highly Correlated Sectors

XLK and XLC (tech + comm services) — high overlap
XLY and XLI (discretionary + industrials) — both cyclical
XLF and XLI (financials + industrials) — both cyclical

Less Correlated / Inverse

XLK (growth tech) and XLU (defensive utilities) — often inverse
XLE (energy) and XLP (staples) — different drivers
XLF (banks) and TLT (bonds) — inverse (banks profit from higher rates)

500 largest US stocks
~30% tech weight
The most-watched index
Liquidity king

QQQ (Nasdaq-100)

Top 100 non-financial Nasdaq stocks
~50%+ tech weight
More growth-tilted
Often higher beta than SPY

IWM (Russell 2000)

2000 small-cap stocks
More cyclical
Higher beta during risk-on periods
Sentiment indicator

DIA (Dow Jones Industrial Average)

30 large-cap stocks
More industrial-tilted
Older, less widely used by traders
Mostly a media benchmark

How They Relate

SPY and QQQ correlation: ~0.85 (high, but QQQ more volatile)
SPY and IWM: ~0.75 (related but small-caps diverge)
IWM leading SPY = risk-on signal
IWM lagging SPY = risk-off signal

Practical Application

Trading tech? Use QQQ as your context, not just SPY
Trading small caps? IWM is your tide
General market direction? SPY rules

Bond Correlations: TLT

TLT (20+ year US Treasury bonds) is the most-watched bond ETF.

Inverse Correlation with Stocks (Usually)

Bonds up = "risk-off" (investors fleeing stocks)
Bonds down = "risk-on" (investors leaving safety for growth)

When TLT Decouples

Inflation regimes (both stocks AND bonds fall as rates rise)
Stagflation
Periods of stress on bonds specifically

What to Watch

TLT making lower lows → rates rising → growth stocks at risk
TLT rallying hard → flight to safety → general stock weakness ahead
TLT and SPY both falling = bad sign (no safe haven)

Hurts multinational earnings (overseas revenue translates to fewer dollars)
Hurts commodities (priced in dollars)
Pressures emerging markets

Weaker Dollar

Benefits multinationals
Boosts commodities and energy
Supports emerging markets

Sectors Most Affected by DXY

Tech: mostly negative correlation (multinationals)
Energy: negative (oil priced in dollars)
Consumer Staples: negative (multinationals like KO, PG)
Banks: mixed (depends on rate context)
Domestic small caps: less impacted

Trading Application

DXY rallying hard → cautious on tech, energy, staples
DXY falling → tailwind for cyclicals and commodities

VIX: The Fear Index

Covered in detail in Chapter 4.5, but worth mentioning in correlations.

VIX-Stocks Correlation

VIX rising = stocks usually falling (negative correlation)
VIX falling = stocks usually rising
Spikes in VIX often mark short-term bottoms in stocks

Bitcoin Correlation

BTC has correlated more with risk assets since ~2020.

Recent Patterns

BTC and QQQ: moderate positive correlation (~0.3-0.6)
BTC behaves as a risk asset during risk-off events
Sometimes decouples (gold-like behavior)

TLT rallying + VIX rising + SPY weak = clear risk-off → be defensive
Dollar weakening + commodities rallying + small caps strong = clear risk-on

Calculating Correlations (Practical)

You don't have to compute beta or correlation by hand. Tools available:

Free Sources

Yahoo Finance: "Beta" listed on stock summary page
StockCharts: Correlation analysis tools
Portfolio Visualizer: Multi-asset correlation matrix

Default Periods

1-year beta is standard
3-month correlations for current market regime
Both views matter

When Correlations Break (And Why)

Correlations are statistical — they describe the past, not guarantee the future.

When Correlations Increase

Crises (everything falls together)
"Risk-off" sells everything indiscriminately
Sector ETF rebalances

When Correlations Decrease

Stock-specific news (earnings, M&A, scandals)
Sector-specific catalysts
Regime changes

Most boats move with the current (high correlation)
A few boats use motors and can move against it (low correlation, alpha)
During storms, all boats get pushed the same way (correlations spike)
During calm, boats can go their own ways more

If you're trading individual stocks without understanding the current, you're flying blind.

Practical Setup

Daily Tracking

SPY, QQQ, IWM moves
VIX level and direction
DXY, TLT major moves
Sector ETF performance

Pre-Trade Check

Before any individual stock trade:

What's SPY doing this week?
What's the sector doing this week?
Is my stock moving with these or against them?
Is the macro environment risk-on or risk-off?

This 30-second check prevents many bad trades.

Practical Takeaways

Individual stocks are largely macro and sector bets. Recognize that.
Beta tells you how much amplification you have. Adjust position size accordingly.
SPY, QQQ, IWM, VIX, TLT, DXY are your daily context. Watch them.
Correlation ≠ diversification. Mixing tech names isn't diversifying.
Strong stocks outperform their sector AND the market.
Correlations spike in crises. Don't rely on past correlations during stress.
Cross-asset signals (bonds, dollar, VIX) give context stocks alone don't provide.

Quick Self-Check

Before moving to 4.5, you should be able to answer:

What does beta measure?
What's the typical correlation of a tech stock to SPY?
What does it mean when TLT rallies sharply?
How does a stronger dollar affect tech stocks?
Why isn't holding 5 tech stocks "diversified"?
What's the difference between SPY, QQQ, and IWM?
When do correlations typically spike?

Previous: 4.3 Sector Analysis and Rotation Next: 4.5 Volatility Regimes (VIX)