Correlation, Beta & Portfolio Risk

Advanced Track. Extends single-trade risk from Risk Management 101 and Kelly sizing to the full book. Builds on portfolio basics and relative strength from Momentum & RS. Hub: stock courses.

Five Tickers, One Bet

A trader risks 1% on each of five “independent” longs: five different tech names, five clean charts, five separate thesis sentences. On a risk-off tape they all gap down together. Account damage is not 1% — it is a correlated 4–5% event wearing a diversification costume. Portfolio risk is the risk that remains after you stop pretending names are independent.

This course defines correlation and beta for operators (not pure statisticians), shows how to budget open risk across the book, identifies crowded trades, and connects single-name process from Tracks 2–3 to book-level survival. Free stock trading calculators still size each leg; this lesson caps how many legs you may open.

1. Correlation in Operator Language

Correlation measures how two return series move together, typically from −1 (perfect opposite) to +1 (perfect lockstep). For traders:

• High positive correlation → losses stack on the same days
• Low / negative correlation → smoother equity curve if edges exist in both legs
• Correlation is unstable — stress regimes push many assets toward +1

You do not need a PhD matrix every morning. You need a practical map: same sector, same factor (growth/value), same macro sensitivity (rates, oil), same beta to SPY/QQQ. Those proxies catch most retail false diversification. Deepen single-name structure with market structure and chart literacy from Reading a Stock Chart.

2. Beta to SPY and QQQ

Beta estimates how much a stock tends to move when the benchmark moves. A beta of 1.5 to SPY roughly means the stock has historically moved ~1.5× the index on average (with noise). High-beta names amplify index risk; low-beta names dampen it — until they do not in idiosyncratic disasters.

Practical uses:

• Size high-beta longs smaller if your true thesis is not “max index leverage”
• When SPY is extended, high-beta RS leaders can still be late chases — measure extension with the percentage change calculator
• Pairs of high-beta names are not two bets; they are a levered index sleeve

Relative strength from Course 26 asks “who leads?” Beta asks “how much index risk am I already carrying?” Both matter for portfolio risk.

3. Sector and Factor Crowding

Sectors co-move. Five semiconductor longs on breakouts from breakout playbooks are one semiconductor book. Factor crowding (everyone long the same mega-cap growth cohort) produces synchronized drawdowns when the factor rotates — the “crowded trade cascade” problem.

Operator rules:

• Cap sector open risk (e.g. max 2% equity risk in one sector)
• Cap number of correlated active names (e.g. max 2–3 high-beta tech)
• Prefer true diversifiers only when they have their own edge (not random opposite charts)

Trend systems from trend following and momentum from RS are especially prone to stacking the same factor. Mean-reversion books from mean reversion can also correlate if every fade is on the same index dump.

4. Risk Budgets: Per Trade vs Per Book

From Risk 101 you already risk a fraction of equity per trade (e.g. 0.5–1%). Portfolio risk adds layers:

• Max total open risk: sum of planned stop risks across open trades (e.g. 3%)
• Max correlated open risk: sum within a cluster (sector/factor)
• Daily / weekly drawdown caps: reduce size after breaches
• Overnight risk cap: smaller than pure day-trade risk if holding (swing trading, gaps)

Worked example. Equity $50,000. Policy: 1% max per trade ($500), 3% max open ($1,500), 2% max in one sector. Already long two semis with $400 and $350 planned risk (correlated). Remaining sector budget = $1,000 − $750 = $250. A third semi with a $500 structural stop is not allowed at full plan — cut size so dollar risk ≤ $250, or skip. Size with the risk / position size calculator; exits with the SL/TP calculator.

Kelly from Course 31 never applies independently to three correlated bets as if they were separate coin flips. Use the Kelly calculator for strategy-level fractions; use portfolio caps for concurrent exposure.

5. Gross, Net, and Directional Exposure

Gross exposure sums absolute notionals (long + short). Net exposure is long notional minus short notional. A book can be “hedged” on paper (low net) while gross is enormous — or fully long net with high beta. Day traders from day trading fundamentals often run high turnover with low overnight net; swing books carry multi-day net beta.

Know your default posture: are you running a long-only momentum book or a market-neutral experiment? Mixing styles without labeling them creates accidental net beta. Foundations of what you own: What Is Stock Trading?

6. Stress Days and Correlation Breakdown

In calm markets, correlations look moderate. In crashes and liquidity shocks, many “diversifiers” fall together. That is why portfolio stress tests ask: “If SPY is −3% and my sector is −5%, what does my open book lose if stops slip?” Microstructure and halt risk from order flow / microstructure and market mechanics matter when exits gap.

Practical habit: on high-VIX / risk-off days, cut max open risk and new correlated entries. Do not “double down for diversification” by adding a sixth tech long.

7. Building a Simple Portfolio Risk Dashboard

Before each session, list open positions with:

1. Ticker, direction, thesis one-liner
2. Planned dollar risk to stop
3. Sector / factor tag
4. Beta bucket (high / med / low to SPY)
5. Overnight? (Y/N)

Sum dollar risk by sector and total. If over policy, reduce before opening new risk. Journal net P&L with the P&L calculator. Track whether multi-name days improve or destroy expectancy with the win rate calculator and break-even calculator.

Optional rebalancing intuition for longer holds: portfolio rebalancer. Broker constraints: broker course. Venues: exchanges hub.

8. Multi-Timeframe and Strategy Correlation

Even different strategies can correlate: a day-trade OR breakout book and a swing breakout book both long risk-on names will bleed together on a red open. Multi-timeframe process from Course 20 should include: “Does this new trade increase the same beta I already hold on a higher timeframe?” Structure and S/R from support/resistance and patterns from chart patterns do not remove factor risk.

9. Common Portfolio Risk Mistakes

• Counting tickers instead of counting risks
• Ignoring sector ETFs when sizing single names
• Applying full Kelly to each correlated position
• Adding “hedges” that are actually more of the same beta
• No overnight risk distinction vs pure day trades
• Revenge re-entry into the same crowded cohort after a factor flush (beginner mistakes)

Optional reading: DennTech blog. Full tool stack: tools hub. TA probability framing: Intro to TA.

10. Checklist, Drill, Narrative

Before new risk: per-trade size OK? total open risk OK? sector bucket OK? overnight bucket OK? thesis not duplicate of existing factor?

Drill: For two weeks, hard-cap open risk at 2% and max two names per sector. Journal whether P&L volatility fell without destroying expectancy. Then carefully test 3% if process is clean.

Narrative: Trader holds three AI-related longs, each “1% risk,” into a sector downgrade day. Stops gap; combined damage exceeds 4%. Correct process would have tagged all three as one sector cluster and refused the third entry — or cut size across the board. Single charts looked fine; the book was a single concentrated bet.

11. Designing Hard Portfolio Rules You Will Actually Obey

Soft intentions (“I’ll try not to stack tech”) fail under FOMO. Convert portfolio risk into hard numbers written in your plan: maximum open risk as a percent of equity; maximum risk in any one sector; maximum number of concurrent high-beta names; maximum overnight risk; maximum new risk after a daily loss threshold. Put the numbers where you see them before each order. If a proposed trade violates a number, the answer is automatic — cut size or pass — without a debate about how clean the chart is. Charts do not get a vote against arithmetic that protects the book.

Review the rules monthly. If you never hit a sector cap, the cap may be cosmetic. If you hit it daily and always override it, the rule is theater. Either tighten the process that generates too many correlated signals or admit you are running a concentrated factor book and size the entire book as one risk unit. Honesty beats performative diversification. A concentrated book can be viable if sized like concentration; it is lethal when sized like five independent coin flips.

12. Scenario Planning: Index Shock and Idiosyncratic Shock

Run two mental (or spreadsheet) scenarios each week on your open book. Scenario A: index shock — SPY −2% to −3% with your sector ETF worse. Estimate mark-to-market loss if stops are late by a realistic slippage factor. Scenario B: idiosyncratic shock — one name gaps −8% on news while others are unchanged. If Scenario A already exceeds your weekly drawdown tolerance, you are over-allocated to beta regardless of how pretty each setup was at entry. If Scenario B ruins the month from a single name, that name’s overnight size is too large for your account goals.

These scenarios are not prophecy. They are budget tests. Airlines do not board without weight limits; traders should not open risk without loss limits under plausible co-movement. When volatility regimes shift upward, re-run the scenarios with wider assumed gaps. Yesterday’s comfortable three-position book can become an uncomfortable single factor bomb when correlation rises. Reduce before the stress day if the test fails — not after.

13. Correlation Across Strategies and Timeframes

Portfolio risk is not only about multiple tickers. It is also about multiple playbooks that secretly buy the same economic bet. A morning gap-and-go long in a high-beta name, an afternoon breakout long in a related name, and a swing pullback long in the sector ETF can triple the same directional exposure while your journal lists three “different” strategies. Tag every trade with strategy, timeframe, and factor. Once a week, aggregate open and closed risk by factor. If “long duration growth” is 80% of your risk budget, you do not have a multi-strategy operation — you have a theme with costumes.

The fix is allocation, not more indicators. Assign maximum risk percentages to strategy sleeves (for example, momentum swing 40% of risk budget, mean reversion 20%, event-driven 20%, discretionary 20%). When the momentum sleeve is full, new momentum ideas wait even if they are A+ on a standalone basis. That is how institutions prevent every desk from buying the same factor on the same day. Retail traders can copy the governance without copying the bureaucracy: a one-page sleeve budget updated weekly is enough if you obey it.

Finally, remember that portfolio risk management is a daily operating system, not a quarterly philosophy. The best rule set fails if you only open the dashboard after a bad day. Before the open, after the open, and before the close: recompute open risk, re-tag sectors, and refuse new correlated exposure when budgets are full. That three-check rhythm is how advanced traders keep correlation from silently rewriting their carefully sized single-name plans into an accidental macro bet.

Key Takeaways

Tools for This Course

• Risk & Position Size Calculator — compute each leg, then sum into portfolio budgets.
• Kelly Calculator — strategy-level fractions only; still apply correlation caps.
• Portfolio Rebalancer — optional aid for multi-position weight drift on longer holds.
• Stock Courses Hub — Track 4 continues with sentiment, journals, and tax records.