Handling losses and drawdowns is the psychological discipline of staying rule-governed while your account is below its prior equity high — accepting individual losing trades as the cost of doing business and tolerating clustered losing streaks without abandoning the system, oversizing to "make it back," or freezing into inaction. The core tension is that the human emotional response to losses is asymmetric and self-defeating: the pain of a loss is felt far more sharply than the pleasure of an equal gain, and that asymmetry pushes traders toward exactly the behaviors — revenge trading, deviating from rules, capitulating at the bottom — that turn a survivable drawdown into a fatal one. This node covers the behavioral side. The arithmetic of drawdown (the peak-to-trough math, the recovery asymmetry, Calmar/MAR/Ulcer) lives in the risk branch under Drawdown & Recovery (008-005-003) — don't duplicate it; reference it.

The behavioral mechanism

Two distinct concepts get conflated and should be kept separate. A loss is a single closed trade that went against you. A drawdown is the cumulative peak-to-trough decline of the equity curve — usually a string of losses (and sub-target wins). Both are statistically inevitable: even a strategy with a genuine edge produces losing streaks, because outcomes are a random draw from a distribution. For a system with a 50% win rate, a run of 5+ consecutive losers will appear regularly over a few hundred trades simply from chance — clustering is expected, not a signal the edge is broken.

The reason this is hard is loss aversion, formalized in Kahneman & Tversky's prospect theory (1979; the cumulative version, 1992). They estimated a loss-aversion coefficient (λ) of roughly 2.25 — losses loom about 2 to 2.5 times larger in felt utility than equivalent gains. Prospect theory also predicts that people become risk-seeking in the domain of losses: facing a sure loss, they reach for the gamble that might erase it. In trading this produces the well-documented failure cluster:

• The disposition effect — selling winners too early and riding losers too long. Odean's 1998 study of 10,000 brokerage accounts (1987–1993) found investors realized gains at a meaningfully higher rate than losses (reported proportions: PGR ≈ 0.148 vs PLR ≈ 0.098), holding losers in the hope they "come back."
• Revenge trading / "making it back" — after a loss, the question silently shifts from "Does this trade meet my criteria?" to "Will this trade recover my loss?" These are different questions and the second one has no edge.
• Drawdown-induced deviation — abandoning a working system mid-drawdown, or conversely oversizing to recover faster, which raises risk of ruin precisely when capital is thinnest.

See the sibling nodes Holding Losers Too Long (009-001-003) and Cutting Winners Early (009-001-002) for the disposition effect in detail.

How it's handled in practice

The recognized professional toolkit treats loss-handling as a systems problem, not a willpower problem — you engineer the discipline rather than summon it:

• Pre-commit the exit. A stop-loss or exit rule defined before entry removes the in-the-moment, loss-averse decision. The loss is "accepted" at entry, not negotiated at the trough.
• Fixed-fractional sizing (commonly cited 1–2% of equity at risk per trade) caps any single loss and, because risk scales with equity, automatically de-risks during a drawdown — the opposite of oversizing. See Position Sizing (008-001) and The 1–2% Risk-Per-Trade Rule.
• Think in R-multiples and expectancy, not dollars. Judging trades as +1R / −1R outcomes against a positive expectancy reframes a loss as one expected sample, not a personal failure. See R-Multiples & Expectancy (008-003). This is core to Van Tharp's framework: you trade your system's distribution, not the last trade.
• Drawdown circuit-breakers. Pre-set rules — e.g. a daily/weekly max-loss stop, or scaling size down after a defined account drawdown — that force a pause before the emotional response compounds. Used heavily by prop firms and disciplined discretionary traders.
• A trading journal / process scorecard. Grading whether each trade followed the rules separates process from outcome, so a rule-following loss is logged as a "good" trade. This is the central mechanism for not over-updating on a single bad result.
• Distinguish a drawdown from a broken edge. A drawdown within historically observed bounds (backtest/forward-test) is noise; one that exceeds the worst seen in testing is a signal to investigate regime change or model decay — a genuine reason to reduce or stop, distinct from emotional capitulation.

Standing & evidence

The behavioral phenomena here are among the better-supported findings in finance. The disposition effect has been replicated across many markets and asset classes (Odean 1998; Weber & Camerer 1998; later international studies). Prospect theory is one of the most-cited results in the social sciences and won Kahneman the 2002 Nobel.

Honesty requires two caveats. First, loss aversion as a universal, fixed-magnitude law is contested. Recent work (notably David Gal, and Yechiam & Hochman) argues some effects attributed to loss aversion are better explained by psychological inertia or increased attention to losses rather than a stable ~2× utility weighting, and that the effect can vanish for small stakes. The λ ≈ 2.25 figure is a useful average, not a constant. Second, the prescriptions for handling drawdowns (journaling, circuit-breakers, "trade the process") are professional best-practice and behaviorally well-motivated, but largely unquantified — there is no clean study proving a journal raises returns by X%. Treat them as risk-management hygiene with strong theoretical grounding, not as a measured edge.

Strengths & limitations

Disciplined loss-handling doesn't improve a strategy's raw edge — it protects your ability to keep applying the edge through the inevitable bad stretch, which is where most traders fail. The recovery asymmetry (a 50% drawdown needs a 100% gain back; see 008-005-003) means survival is the precondition for everything else, so the payoff is preventing the catastrophic tail, not boosting the average trade.

The single most common and most dangerous misuse is size escalation to recover faster — increasing risk after losses to "win it back." This couples the largest position sizes to the moment of lowest capital and worst psychological state, and is the classic route to ruin. The mirror-image failure is over-correction: cutting size to near-zero or stopping entirely after a normal losing streak, which guarantees you can't recover and often means quitting a sound system at the bottom. The skill is telling normal variance apart from a real regime break — which requires having measured your system's expected drawdown distribution in advance.

System relevance

Augustus is rule-driven, so it is structurally immune to revenge trading and disposition-effect emotion — but only if its inputs enforce the discipline. The operative cross-links are Drawdown & Recovery (008-005-003) for the math, Position Sizing / 1–2% rule (008-001) and R-Multiples & Expectancy (008-003) for the framing. Two caveats for consuming this node: (1) Augustus should size off the current account equity (auto-de-risking in a drawdown), never escalate after losses; and (2) the distinction between a normal drawdown (continue per system) and an out-of-sample drawdown (investigate / halt) should be tied to Cairn's measured track record of the live edge, not to any felt urgency to recover.

Sources

• Kahneman, D. & Tversky, A. (1979); Tversky & Kahneman (1992) — prospect theory, loss aversion λ ≈ 2.25 (via Wikipedia, Prospect theory / Loss aversion).
• Odean, T. (1998), "Are Investors Reluctant to Realize Their Losses?", Journal of Finance — disposition effect, PGR/PLR figures.
• Gal, D.; Yechiam & Hochman — contested status of loss aversion (summarized in Wikipedia, Loss aversion).
• Drawdown-recovery math: TradeZella, Investopedia, and sibling node 008-005-003.
• Van Tharp, Trade Your Way to Financial Freedom — R-multiples, expectancy, process-vs-outcome framing.
• Practitioner landscape: FXIFY, DayTradingToolkit, QuantStrategy.io (drawdown psychology guides — qualified as professional best-practice, not measured edge).