The "Sleep at Night" Sizing Rule

Every position-sizing rule—maximum position size, sector limits, volatility adjustments—is mathematically sound. But they're all secondary to one more fundamental question: if this position crashes 50%, can you hold it, or will you panic and sell?

If a 10% position falling 50% causes you to lose sleep and consider selling, then 10% is the wrong size for you. Period. Your psychological tolerance for loss is the ceiling on position size.

This is the "sleep at night" test, and it's often more important than any mathematical framework.

Quick definition: The sleep-at-night test asks: "If this position crashed 50% tomorrow, would I lose sleep over it or consider selling?" If yes, the position is too large. The maximum size you're comfortable holding through a crash is your true maximum position size, regardless of what formulas recommend.

Key takeaways

• Mathematical sizing rules are guidelines, not commandments; psychology trumps math when they conflict.
• A position that causes you to panic-sell at the worst time destroys returns more than any "suboptimal" sizing does.
• The sleep test reveals your true risk tolerance; it's often smaller than you think.
• Different investors have different sleep thresholds; there's no universal "right" size.
• Knowing your sleep threshold prevents over-sizing and forced sales during crashes.
• If a position is large enough to cost you sleep, it's large enough to cause emotional mistakes.

Why psychology matters more than math

The mathematical case for a 12% position in a high-conviction stock is ironclad. Higher conviction justifies larger position. But if a 12% position crashes 50% and you panic-sell, the mathematical optimality is meaningless.

You've now realized a 6% portfolio loss at exactly the wrong time. A slower, smaller position that you could have held through the crash might have recovered. The psychological break cost you more than the "suboptimal" sizing would have.

This is why the sleep test is foundational. It's not an aspirational framework; it's a reality check.

Finding your sleep threshold

Sit with the following scenarios:

Scenario 1: A position is 5% of your portfolio, crashes 50%.

• Portfolio loss: 2.5%
• Portfolio impact: Mild
• Emotional reaction: Annoyed but holding
• Verdict: ✓ Comfortable size

Scenario 2: A position is 10% of your portfolio, crashes 50%.

• Portfolio loss: 5%
• Portfolio impact: Moderate
• Emotional reaction: Stressed, considering selling
• Verdict: ? Maybe comfortable, maybe not

Scenario 3: A position is 15% of your portfolio, crashes 50%.

• Portfolio loss: 7.5%
• Portfolio impact: Significant
• Emotional reaction: Panicking, seriously considering selling
• Verdict: ✗ Too large

Your sleep threshold is somewhere in there. For most investors, it's between 5-10% per position. Some exceptionally patient investors can handle 15%+. Some risk-averse investors sleep well only with 3-4% per position.

The right answer is the one where you can actually hold during crashes.

Different sleep thresholds for different people

Sleep tolerance varies based on:

Age and time horizon:

• 25 years old with 40 years to retirement: Can handle 12-15% positions because they have time to recover
• 55 years old with 10 years to retirement: Should keep positions to 5-6% because recovery time is limited
• 65 and retired: Maybe 3-4% maximum because portfolio is now funding withdrawals

Income stability:

• Stable high income: Can handle larger positions because job loss won't force selling
• Variable income (commission, business owner): Smaller positions are prudent because job downturns coincide with market downturns
• Low income or already retired: Smaller positions due to lack of new capital to deploy

Overall portfolio size:

• $50k portfolio: 5% = $2,500 loss from 50% crash. Emotionally manageable.
• $500k portfolio: 5% = $25,000 loss. Can you really hold this without panic?
• $5M portfolio: 5% = $250,000 loss. Many investors sleep fine; others panic.

Life circumstances:

• Stable life, no major expenses planned: Can handle larger positions
• Planning to retire, buy a home, or make major changes: Smaller positions are prudent

The mermaid decision tree: finding your sleep threshold

The interaction with overall portfolio volatility

A single position's size affects how much sleep you lose, but so does overall portfolio volatility. A 10% position in a 18% volatility stock creates less portfolio turbulence than a 10% position in a 50% volatility stock.

Some investors can handle a 10% position if portfolio volatility stays 12-15% (because the rest of the portfolio is stable). But a 10% position that pushes portfolio volatility to 18-20% is unbearable.

The sleep test accounts for this implicitly. When you imagine a crash, you're imagining it in the context of your total portfolio. A 10% position that doubles portfolio volatility will keep you awake; the same 10% position that marginally increases volatility won't.

The margin-call angle

If you use margin (borrowed money), the sleep threshold must be lower. You're not just worried about losses; you're worried about margin calls that force selling.

If 10% of your portfolio on 2x leverage falls 50%, that's a forced liquidation scenario. The sleep threshold for leveraged positions must be tight enough that even 60-70% declines don't trigger margin calls.

Most long-term investors should avoid leverage entirely. If you use it, cap individual positions at 3-4% maximum and use 1.5x leverage or less to keep margin buffer.

Real-world examples of sleep tests

The concentrated Berkshire holder (2008): An investor held 20% of their portfolio in Berkshire Hathaway. In 2008, Berkshire fell 45%, costing them 9% of their portfolio. They panicked and sold at the bottom, locking in a 9% loss. The sleep test would have suggested a 5-7% position maximum.

The Amazon believer (2000): An investor held 15% in Amazon through the dot-com crash. Amazon fell 95%, costing them 14.25% of their portfolio. They held (most slept poorly), and 20 years later, it was their largest winner. Sleep threshold mattered less than conviction, but they lost sleep for two years.

The Tesla fan (2022): Tesla fell 65% in 2022. An investor with 12% in Tesla lost 7.8%, and while they held, they considered selling for two months. An 8% position would have cost 5.2%—still painful, but psychologically more bearable.

Knowing yourself: the hard part

The hardest part of the sleep test is honest self-assessment. Most investors overestimate their tolerance. They say, "I can handle 15% positions," then panic at the first crash and sell at 50% losses.

A better approach: experiment with smaller sizes and observe your behavior. If you genuinely don't care when a 5% position crashes 30%, try 7%. If 7% causes you to obsessively check prices, you know your threshold is below 7%.

Over years, you'll learn your true tolerance. And it might be smaller than you think—which is fine. Knowing your threshold prevents expensive mistakes.

The relationship with conviction

High conviction and high position size reinforce each other. A position you're deeply convinced about can be larger because conviction sustains you through crashes. A position you're modestly convinced about should be smaller because conviction won't carry you.

The sleep test thus combines both: size such that you can hold through crashes and your conviction is strong enough that you want to.

If a position is large enough to cause sleep loss, but you don't have strong conviction, that's a red flag. Consider:

• Reducing size, or
• Building more conviction through research, or
• Passing on the idea entirely

A position where you're losing sleep but not deeply convinced about is the worst scenario.

Common mistakes with the sleep test

Mistake 1: Testing your sleep threshold with imagined crashes, not real ones. It's easy to say, "I can handle a 50% crash." Until a position actually crashes 40%, you don't know. Real crashes are emotionally different from hypothetical ones.

Mistake 2: Changing your answer based on recent performance. During bull markets, you feel invincible and claim high tolerance. During bear markets, fear dominates. Use the sleep test during calm markets to set reasonable limits, then stick to them during crashes.

Mistake 3: Ignoring the time cost of monitoring. A large position doesn't just cost sleep; it costs mental bandwidth. If you're checking the price daily and obsessing over news, the position is too large.

Mistake 4: Confusing sleep threshold with expected returns. A position might have great expected returns (justifying a large size mathematically) but still be emotionally too large. Both factors matter.

Mistake 5: Having different sleep thresholds for different types of positions. You might say, "I can handle 10% in index funds but only 5% in individual stocks." This is fine, but be explicit. Don't pretend they're the same risk level.

FAQ

Q: Should I size based on sleep threshold or mathematical risk? Both. Let mathematical frameworks (position size, volatility adjustment, sector limits) suggest a size. Then run the sleep test. If the suggested size fails the sleep test, use the smaller size from your sleep threshold.

Q: Can my sleep threshold change over time? Yes. As you age, your threshold might decrease (retirement approaching). As you gain experience, it might increase (you've held through crashes and recovered). Review your threshold every 5 years.

Q: Is it weak to have a small sleep threshold? No. Self-knowledge is strength. Knowing you can only handle 5% positions and sticking to them is better than claiming 10% and panic-selling at 40% losses.

Q: What if my sleep threshold is smaller than my conviction suggests? You have options: (1) Size at sleep threshold and accept modesty; (2) Build sleep threshold over time through experience; (3) Find a different thesis you're equally convicted about but less emotionally attached to. Option 1 is most prudent.

Q: Should I tell my advisor my sleep threshold? Yes. If you're working with a financial advisor, clearly state your sleep threshold. They can then size your portfolio to respect it.

Q: How do I know if my sleep threshold is realistic? Test it during market corrections. The next 10-20% market decline is your opportunity to observe your actual behavior. If you panic during a 15% market decline, your sleep threshold for 10% positions might be lower than you think.

Related concepts

• Risk tolerance: Your mathematical capacity to handle losses
• Risk comfort: Your psychological capacity to handle losses (often smaller than risk tolerance)
• Emotional investing: Letting emotions drive decisions (what the sleep test helps prevent)
• Anchoring bias: Being unable to sell because of the original purchase price (a sleep threshold problem)
• Buy-and-hold psychology: The emotional difficulty of truly holding for decades

Summary

All position-sizing frameworks are mathematical and sound. But they're all secondary to one reality: if a position is so large that you panic-sell during a crash, no sizing framework prevents disaster.

The sleep-at-night test reveals your true maximum position size. It's the position where you can intellectually hold through a 50% decline without emotional breaks. For most investors, this is 5-8%. For some, it's 3-4%. For a few with exceptional patience and conviction, it's 10-15%.

The correct maximum position size is the one you can actually stick to during crashes. Everything else is a theoretical exercise.

Next

With individual position sizes determined, you can now document all your sizing rules into a formal investment policy statement that governs your entire portfolio construction and rebalancing.