---
name: drawdown-psychology
description: >
  Deep expertise in drawdown management — the psychology of losses, when to cut vs when to hold,
  stop-loss discipline, drawdown budgets, and the behavioral frameworks of Howard Marks, Tudor Jones,
  and Buffett. Use when navigating portfolio losses or planning pre-trade risk discipline.
metadata:
  author: nirav
  version: "1.0"
compatibility: Designed for Claude Code
---

# Drawdown Psychology — The Human Element of Risk

Everything in position sizing, tail risk, and correlation analysis is worthless if you panic at -30% and sell everything. The hardest part of risk management isn't calculating the optimal position size — it's honoring your stop when the loss is real, or holding through a drawdown when every instinct screams to sell. This skill covers the behavioral dimension of drawdowns: when to be aggressive, when to be defensive, how to plan for losses before they happen, and how to make decisions during losses.

The core truth: **You will experience drawdowns. The question is whether you've planned for them or whether they plan you.**

## Howard Marks: When to Be Aggressive vs Defensive

### The Marks Cycle Framework

Howard Marks's investment philosophy centers on one idea: **the pendulum of market psychology swings between euphoria and despair, and the best returns come from leaning against it.** The goal is not to predict the cycle — it's to recognize where you are in it and adjust aggressiveness accordingly.

### When to Be Aggressive

Increase risk exposure when ALL or most of the following conditions hold:

| Condition | How to Measure | Why It Matters |
|-----------|---------------|----------------|
| Prices below intrinsic value | Valuation metrics (CAPE, EV/EBITDA, credit spreads) at historical highs for risk compensation | You're being paid to take risk |
| Pessimism is widespread | Sentiment surveys, fund flows showing outflows, media tone overwhelmingly negative | Fear creates opportunity; assets are being sold without regard to value |
| Risk premiums are high | Credit spreads wide, equity risk premium elevated, volatility elevated | The market is pricing in bad outcomes, creating asymmetric upside |
| Capital is scarce | Few buyers, bank lending tightening, fundraising is difficult | Less competition for bargains; sellers are desperate |
| Recently experienced losses | Market has drawn down significantly from highs | Marks: "The best opportunities come from buying from forced sellers" |

**The Marks test for aggressiveness**: "If most people are terrified and selling, and you can buy what they're selling at prices that compensate you well for the risk, you should be aggressive. Not because you're smarter, but because you're doing the opposite of what fear compels."

### When to Be Defensive

Reduce risk exposure when ALL or most of the following conditions hold:

| Condition | How to Measure | Why It Matters |
|-----------|---------------|----------------|
| Prices above intrinsic value | Valuation metrics at historical lows for risk compensation (expensive) | You're not being paid enough for the risk |
| Optimism is widespread | Sentiment bullish, fund flows showing inflows, "new era" narratives | Greed inflates prices; the market is fragile |
| Risk premiums are compressed | Credit spreads tight, VIX low, equity risk premium thin | The market is pricing in good outcomes with no margin for error |
| Capital is abundant | Too many buyers chasing too few deals, easy lending, record fundraising | Competition compresses returns and inflates prices |
| Recently experienced gains | Market near all-time highs, strategies have performed well | Complacency masks rising risk |

**The Marks test for defensiveness**: "If everyone is enthusiastic and buying, and prices have been bid up to the point where the risk-reward is poor, get defensive. Not because a crash is imminent — you can't know that — but because the margin of safety is thin."

### The Marks Asymmetry Principle

Marks doesn't try to be right about the direction of markets. Instead, he adjusts the **ratio of aggressiveness to defensiveness** based on where the cycle stands:

- **At cycle lows**: Aggressiveness/defensiveness ratio of 70/30 — heavily weighted toward taking risk
- **At mid-cycle**: Ratio of 50/50 — balanced, no strong lean
- **At cycle peaks**: Ratio of 30/70 — heavily weighted toward preserving capital

The beauty of this framework: you don't need to call the top or bottom. You just need to recognize the general neighborhood and adjust accordingly.

## The Psychology of Drawdowns

### Prospect Theory and Loss Aversion

Daniel Kahneman and Amos Tversky's prospect theory explains why drawdowns are psychologically devastating:

1. **Loss aversion**: Losses hurt roughly 2-2.5x more than equivalent gains feel good. A $50,000 loss creates 2-2.5x more psychological pain than a $50,000 gain creates pleasure.

2. **Reference point dependence**: People evaluate outcomes relative to a reference point, not in absolute terms. If your portfolio was $1M and dropped to $700K, you feel the $300K loss — not the fact that you still have $700K.

3. **Diminishing sensitivity**: The difference between a -10% and -20% drawdown feels larger than the difference between -40% and -50%, even though the dollar amount may be the same. This creates a dangerous dynamic: by the time the drawdown gets severe, you're numb to incremental losses and may hold far too long.

4. **The disposition effect**: Investors are more likely to sell winners (to "lock in" gains) and hold losers (to avoid "realizing" a loss). This is exactly backwards — it keeps you in losing positions and cuts winning positions short.

### The Emotional Sequence of a Drawdown

| Phase | Drawdown | Emotional State | Common Mistake |
|-------|----------|-----------------|----------------|
| 1 | -5% to -10% | Denial: "It's just a dip, it'll come back" | Ignoring stop-loss levels |
| 2 | -10% to -20% | Anxiety: "Maybe I should sell some" | Selling randomly instead of systematically |
| 3 | -20% to -30% | Fear: "This could get much worse" | Panic selling at the worst time |
| 4 | -30% to -50% | Capitulation: "I can't take it anymore, sell everything" | Maximum loss crystallized at the bottom |
| 5 | Recovery begins | Regret: "I sold at the bottom" | Refusing to re-enter, missing the recovery |
| 6 | New highs | FOMO: "I need to get back in now" | Buying at the top to avoid missing more |

This emotional sequence is nearly universal. Every investor who has lived through a major drawdown recognizes it. The purpose of pre-trade planning is to short-circuit this sequence by making decisions when you're rational (before the drawdown) rather than emotional (during it).

## Maximum Drawdown Analysis

### Historical Drawdowns by Asset Class

Understanding the typical and extreme drawdowns for each asset class calibrates expectations and prevents the shock of "this shouldn't happen":

| Asset Class | Typical Drawdown (90th percentile) | Max Historical Drawdown | Recovery Time (from max) |
|------------|-----------------------------------|------------------------|------------------------|
| US Large Cap Equities | -15% to -25% | -56.8% (2007-2009) | ~5.5 years |
| US Small Cap Equities | -20% to -30% | -54% (2007-2009) | ~4 years |
| International Equities | -20% to -35% | -57% (2007-2009) | ~5 years |
| Emerging Market Equities | -25% to -40% | -65% (2007-2009) | ~4 years |
| US Aggregate Bonds | -3% to -5% | -18% (2020-2022) | Still recovering |
| High-Yield Bonds | -10% to -20% | -33% (2007-2009) | ~2 years |
| Commodities | -20% to -40% | -69% (2008-2020 peak to trough, index) | Varies widely |
| REITs | -20% to -35% | -68% (2007-2009) | ~5 years |

**The lesson**: Every major asset class has experienced drawdowns of 30-70%. If you hold an asset class, you WILL experience a drawdown of at least 20-30% at some point. The question is whether you've planned for it.

## Recovery Math: The Asymmetry of Losses

The relationship between loss and required recovery gain is non-linear:

| Loss | Gain Needed to Recover | Key Insight |
|------|----------------------|-------------|
| -10% | +11.1% | Manageable; normal market fluctuation |
| -20% | +25.0% | Starting to be painful; takes ~1 year in average markets |
| -30% | +42.9% | Difficult; takes ~2 years |
| -40% | +66.7% | Very difficult; takes ~3 years |
| -50% | +100.0% | You need to DOUBLE your money just to get back to even |
| -60% | +150.0% | You need 2.5x your remaining capital |
| -75% | +300.0% | You need 4x your remaining capital |
| -90% | +900.0% | Effectively permanent impairment |

**The critical insight**: The recovery math is why defense is more important than offense. Avoiding a -50% drawdown is worth more than capturing a +100% gain — because after a -50% loss, you need that +100% gain just to get back to zero.

This asymmetry means:
- Capital preservation should be the first priority, return generation the second
- Stop-losses should be tight enough to prevent losses from crossing into "difficult to recover" territory (roughly -20% at the portfolio level)
- Concentration should be limited so no single position can inflict unrecoverable damage

## Tudor Jones's Stop-Loss Discipline

Paul Tudor Jones's approach to stop-losses is absolute:

1. **Define the loss before entry**: Before taking any position, determine the exact price level where the thesis is invalidated
2. **Honor it without exception**: When the stop is hit, exit. No rethinking, no "let me give it a bit more room," no averaging down
3. **The stop is the cost of being wrong**: Every trade has a cost of failure. The stop defines that cost. Accept it as a business expense.

### Why Stops Work

- **They cap the downside**: A stop at -10% means you can never lose more than 10% on that position (barring gaps)
- **They remove emotions**: The decision to sell is made when you're rational (at entry), not when you're panicking (in a drawdown)
- **They preserve capital for the next opportunity**: Capital lost to a failed trade that should have been stopped out is capital that can't be deployed to the next winner
- **They enforce anti-martingale discipline**: You can't "average losers" if you've already exited

### Common Objections to Stops (and Why They're Wrong)

| Objection | Reality |
|-----------|---------|
| "I got stopped out and then it reversed" | This will happen. It's the cost of insurance. The times the stop saves you from -50% far outweigh the times you get whipsawed. |
| "Stops don't work in gaps" | True — gaps can blow through stops. This is why position sizing matters: size the position so even a gap-through-stop doesn't cause unrecoverable damage. |
| "My time horizon is long enough to wait it out" | Some drawdowns don't recover (see: Japanese equities 1989-present, value stocks 2010-2020). Infinite time horizon is a fantasy. |
| "Buffett doesn't use stops" | Buffett has an insurance float that provides permanent capital with no redemption pressure. You don't. |

## Buffett's Alternative: Hold Through If the Thesis Is Intact

Warren Buffett's approach is the philosophical opposite of Tudor Jones's:

- **Don't sell because the price dropped**: Price is what you pay; value is what you get. If the value hasn't changed, the lower price is an opportunity, not a reason to sell.
- **Hold through drawdowns IF**: You understand the business deeply, the competitive moat is intact, management is trustworthy, and the long-term thesis is unchanged.
- **Sell when the thesis breaks**: Not when the price drops — when the fundamental reason you bought no longer holds.

### When Buffett's Approach Works

| Condition | Why It Works |
|-----------|-------------|
| You have permanent capital (no redemptions, no leverage) | You can literally wait forever; price eventually converges to value |
| You understand the business at the level of an owner | You can distinguish between price noise and thesis impairment |
| The company has a durable competitive advantage | Temporary setbacks don't destroy long-term value |
| You have the psychological constitution to hold through -50% | Most people say they do, but very few actually can |

### When Buffett's Approach Fails

| Condition | Why It Fails |
|-----------|-------------|
| You have leveraged capital or redemption risk | You may be forced to sell at the bottom regardless of thesis |
| You don't deeply understand the business | You can't distinguish thesis impairment from noise |
| The competitive advantage is eroding | Holding through is just slow bleeding |
| You're wrong about the thesis but can't admit it | "The thesis is intact" becomes a rationalization for holding a loser |

## The Cut-or-Add Decision Framework

When a position is losing money, you face the most consequential decision in investing: do you cut the loss or add to the position? This framework provides a systematic approach:

### Question 1: Has the Thesis Changed?

| Answer | Action | Rationale |
|--------|--------|-----------|
| Yes — fundamentals have deteriorated | Cut the position | The reason you bought no longer holds. Price is confirming, not contradicting, reality. |
| No — fundamentals are unchanged | Proceed to Question 2 | The loss may be noise, not signal. |
| Uncertain — you're not sure | Cut to half position | When in doubt, reduce. You can always re-enter if clarity improves. |

### Question 2: Is This a Temporary Dislocation or Permanent Impairment?

| Answer | Action | Rationale |
|--------|--------|-----------|
| Temporary: market-wide sell-off, sentiment shift, technical selling | Consider adding | The loss is driven by factors unrelated to your thesis. |
| Permanent: competitive position eroded, regulation changed, technology disrupted | Cut the position | No amount of waiting will restore value that's been structurally destroyed. |
| Unclear | Hold at current size, set a deadline for clarity | Give yourself a bounded window to determine which it is. |

### Question 3: Do You Have the Capital and Time to Wait?

| Answer | Action | Rationale |
|--------|--------|-----------|
| Yes: no leverage, no near-term liabilities, long time horizon | Adding is viable | You can afford to be patient. |
| No: leveraged, facing redemptions, or need liquidity soon | Cut regardless of thesis | You don't have the luxury of being right eventually. |
| Partially: some constraints but some flexibility | Reduce to a size you can hold without stress | Right-size the position for your actual constraints. |

### The Integration

```
Position is losing money
  |
  v
Has the thesis changed?
  |
  Yes --> CUT
  |
  No / Uncertain
  |
  v
Temporary dislocation or permanent impairment?
  |
  Permanent --> CUT
  |
  Temporary / Unclear
  |
  v
Do you have the capital and time to wait?
  |
  No --> CUT (or reduce to manageable size)
  |
  Yes
  |
  v
Consider adding — but only if:
  - Position size after adding still respects risk budget
  - Portfolio heat is not overheated
  - You'd be willing to buy this at today's price if you had no position
  - You're not adding out of ego ("I must be right")
```

**The ego test**: Would you buy this stock today at this price if you had no existing position? If the answer is yes, adding may be rational. If the answer is "no, but I already own it so I should hold" — that's the sunk cost fallacy, and you should cut.

## Drawdown Budgets

A drawdown budget allocates the maximum acceptable drawdown across strategies and positions. It answers: "How much am I willing to lose from each source before I stop out?"

### Setting Drawdown Budgets

| Level | Budget | Trigger Action |
|-------|--------|---------------|
| Individual position | -5% to -15% of position value (varies by strategy) | Hit stop-loss, exit position |
| Strategy sleeve | -10% to -20% of sleeve allocation | Reduce all positions in the strategy by 50% |
| Total portfolio | -15% to -25% of portfolio value | Move to maximum defensiveness: reduce to 50% invested, 50% cash |
| "Red line" | -30% of portfolio value | Full risk-off: move to 80%+ cash, T-bills only; reassess entire approach |

### Allocating Drawdown Budget Across Strategies

Total portfolio drawdown budget should be allocated such that no single strategy can consume more than its share:

**Example**: Total portfolio drawdown budget = -20%

| Strategy | Allocation | Drawdown Budget | Max Contribution to Portfolio Loss |
|----------|-----------|----------------|-----------------------------------|
| Core equity | 60% | -25% of sleeve | -15% of portfolio |
| Tactical trades | 20% | -15% of sleeve | -3% of portfolio |
| Speculative | 10% | -50% of sleeve | -5% of portfolio |
| Hedges | 10% | N/A (cost, not loss) | Premium drag only |

Note that the sum of potential contributions (-23%) slightly exceeds the budget (-20%). This is intentional — it assumes not all strategies will hit maximum drawdown simultaneously. If they do (a crisis), the portfolio-level stop at -20% triggers first.

## The Uncle Point

Everyone has an "uncle point" — the level of pain beyond which they cannot function rationally. Knowing your uncle point before entering a position is essential because:

1. **It defines your true risk tolerance**: Not what you say in a questionnaire, but the actual level of loss that causes you to deviate from your plan
2. **It sets the position sizing ceiling**: If your uncle point is -15% of portfolio, you cannot take positions that could collectively lose more than 15%
3. **It prevents catastrophic behavioral errors**: If you know you'll panic at -30%, you can structure the portfolio so it never reaches -30%

### Finding Your Uncle Point

Honest self-assessment questions:

| Question | What It Reveals |
|----------|----------------|
| What's the largest loss I've ever experienced? How did I behave? | Your revealed preference under stress |
| At what drawdown level would I start checking my portfolio multiple times per day? | Where anxiety begins to impair judgment |
| At what drawdown level would I lose sleep? | Where the loss is affecting your life, not just your portfolio |
| At what drawdown level would I feel compelled to sell everything? | Your true capitulation point |
| What drawdown would make me question my entire investment approach? | Where the loss becomes an identity crisis |

**The honest answer to the last question is your uncle point.** Size and structure the portfolio so it never reaches this level. If your uncle point is -25%, the maximum portfolio drawdown budget should be -20% (with 5% margin of safety).

### Adjusting the Portfolio to Your Uncle Point

| Uncle Point | Portfolio Implication |
|------------|---------------------|
| -10% | Very conservative: low equity allocation, heavy hedging, mostly cash/bonds |
| -15% to -20% | Moderate: balanced allocation, systematic hedging, defined stops |
| -25% to -30% | Aggressive: higher equity, concentration possible, but with strict stops |
| -40% or beyond | Very aggressive: can tolerate full market drawdowns; few investors are genuinely here |

**Warning**: Most investors overestimate their uncle point. They say -30% when they actually capitulate at -15%. The 2020 and 2022 drawdowns provided real-world calibration for many investors. Use actual experience, not hypothetical tolerance.

## Practical Framework: Pre-Trade Drawdown Checklist

Before entering ANY position, complete this checklist:

### 1. Define the Stop-Loss

- **Where is the stop?** Specific price level where the thesis is invalidated
- **Why that level?** Based on technical support, fundamental floor, or maximum acceptable loss — not arbitrary
- **Is the stop realistic?** Can it be hit by normal volatility without invalidating the thesis? If so, widen it and reduce position size accordingly

### 2. Calculate Maximum Position Loss

- **Position size x distance to stop = maximum dollar loss**
- **Maximum dollar loss / portfolio value = portfolio impact**
- Is the portfolio impact within the drawdown budget for this position type?

### 3. Assess Recovery Implications

- If the stop is hit, what percentage of portfolio is lost?
- What recovery gain is needed? (Use the asymmetry table above)
- Is the recovery achievable in a reasonable timeframe?

### 4. Plan the Emotional Response

- At what drawdown level will I feel uncomfortable? (Be honest)
- What will I do when I feel that discomfort? (Plan: check thesis, not price)
- Who will hold me accountable to the plan? (Trading partner, journal, rules)

### 5. Pre-Commit to the Decision Framework

Write down and sign:
- "If the thesis changes, I will cut the position regardless of loss size"
- "If the stop is hit, I will exit. No exceptions, no averaging down"
- "If I'm unsure, I will reduce to half position and reassess"
- "I will not make any decision in the first hour after a major move"

### 6. Check Portfolio Context

- What is the current portfolio heat?
- What is the current total drawdown from peak?
- Am I close to any budget thresholds that would be triggered by this position losing?
- Do I have other positions correlated with this one? What's the combined worst-case?

### 7. Document Everything

Record in the trade journal before entry:
- Thesis (why you're buying)
- Entry price
- Stop-loss level and rationale
- Target price and rationale
- Position size and rationale
- Maximum acceptable loss in dollars and portfolio percentage
- What would make you add to the position
- What would make you cut before the stop
- Review date (when you'll reassess regardless of price action)

### Decision Summary

```
Before any trade:
  |
  v
Define stop-loss level (where thesis is invalidated)
  |
  v
Calculate max loss (size x distance to stop)
  |
  v
Is max loss within drawdown budget? --No--> Reduce size until it is
  |
  Yes
  v
Plan emotional response at each drawdown level
  |
  v
Pre-commit to cut/add decision framework IN WRITING
  |
  v
Check portfolio heat and correlation with existing positions
  |
  v
Document everything: thesis, entry, stop, target, size, review date
  |
  v
Enter position
  |
  v
During the trade: check thesis weekly, not price hourly
  |
  v
Stop hit? --> Exit. No debate.
Target hit? --> Take profits per plan.
Thesis changed? --> Cut. Even if it hurts.
Nothing changed? --> Hold. Don't touch it.
```

## Related Skills

- **`position-sizing`** (Risk Architecture) — consult when setting pre-trade position sizes that respect drawdown budgets and uncle points before entering any position
- **`second-level-thinking`** (Value & Quality) — consult when building contrarian conviction during drawdowns; second-level analysis determines whether a drawdown is noise or signal
- **`rebalancing-logic`** (Adaptive Monitoring) — consult when deciding whether to act on drawdowns through systematic rebalancing rather than emotional ad-hoc decisions