Based on John R. Nofsinger, Investment Psychology (投资心理学) (5th Edition)
Nofsinger's central argument is that traditional finance theory — built on the assumption of rational, utility-maximizing agents — systematically fails to explain how real people actually make investment decisions. The Efficient Market Hypothesis (EMH) assumes that prices reflect all available information because rational investors will arbitrage away any mispricings. Behavioral finance challenges this by documenting that investors are predictably irrational in ways that create persistent, exploitable patterns and, more importantly, that destroy individual portfolio returns.
| Pillar | Description |
|---|---|
| Limits to Arbitrage | Even when mispricings exist, structural barriers (cost, |
| risk, short-sale constraints) prevent rational traders from | |
| fully correcting them | |
| Psychology of Decision-Making | Cognitive biases, heuristics, and emotional states cause |
| systematic deviations from rational behavior |
The book's practical premise is that you cannot fix what you cannot see. Most investors believe they are rational while simultaneously falling prey to dozens of well-documented biases. The cost is enormous: Barber and Odean's foundational research shows that the most active individual traders underperform passive benchmarks by 6-7% annually — not from bad stock-picking, but from behavioral errors layered on top of transaction costs.
Nofsinger organizes behavioral finance into a coherent decision-making model:
The goal is not to eliminate biases (impossible) but to build systematic defenses that reduce their impact on portfolio outcomes.
Overconfidence is the single most damaging bias in investing. Nofsinger identifies it as the "mother of all biases" because it amplifies every other psychological error. If you believe your information is better, your analysis sharper, and your timing superior, you will trade more, diversify less, and take risks you do not understand.
People believe that more information leads to better decisions. In reality, beyond a certain threshold, additional information increases confidence far more than it increases accuracy.
Key research finding: When analysts receive more data points about a company, their confidence in their forecast rises dramatically, but their forecast accuracy barely improves and sometimes declines. Information overload leads to pattern-finding in noise.
| Information Level | Confidence | Accuracy | Net Effect |
|---|---|---|---|
| Minimal (5 items) | Low | Moderate | Appropriately cautious |
| Moderate (15 items) | High | Moderate | Overconfident — dangerous zone |
| High (30+ items) | Very High | Moderate | Illusion of expertise |
Implementation rule: More research does not always mean better decisions. Define in advance what information is decision-relevant, collect only that, and decide. Set a research time limit per position.
Investors who actively choose their own stocks (vs. being randomly assigned) believe their portfolios will outperform — even when the selection mechanism is demonstrably random. Control over the process creates a false sense of control over the outcome.
Manifestations in investing:
Implementation rule: Ask yourself: "Does this action actually change the probability of a good outcome, or does it merely feel like it does?" If the latter, stop.
When an investment succeeds, overconfident investors attribute it to their skill. When it fails, they attribute it to bad luck, market manipulation, or unforeseen events. This asymmetric attribution prevents learning from mistakes and inflates confidence over time.
The feedback loop:
Win → "I'm skilled" → Increase position size / trade more
Lose → "Bad luck" → No adjustment to strategy
Net result → Escalating risk without escalating skill
Implementation rule: Keep a decision journal. For every trade, record the thesis, expected outcome, and actual outcome. Review quarterly. If your wins and losses are both "because of me," something is wrong. If your wins are skill and losses are luck, you are almost certainly self-attributing.
Daniel Kahneman and Amos Tversky's Prospect Theory is the theoretical backbone of behavioral finance. Nofsinger devotes substantial attention to it because it explains more investor behavior than any other single framework.
The pain of losing $1,000 is psychologically approximately 2 to 2.5 times as intense as the pleasure of gaining $1,000. This asymmetry is hardwired — it likely served survival purposes in ancestral environments where a loss (of food, shelter, safety) could be fatal, while an equivalent gain was merely helpful.
Prospect Theory replaces the smooth, concave utility function of classical economics with an S-shaped value function:
Value
^
| /---- (gains: concave, diminishing sensitivity)
| /
| /
|------/-----------> Outcome
| /
| /
| / (losses: convex, steeper than gains)
| /
Key properties:
| Behavior | Prospect Theory Explanation |
|---|---|
| Holding losers too long | Selling realizes the loss, making it "real" and painful |
| Selling winners too early | Taking gains feels good; risk of losing gains feels bad |
| Risk-seeking with losses | In loss domain, investors gamble to break even |
| Risk-averse with gains | In gain domain, investors lock in profits prematurely |
| Reference point fixation | Purchase price becomes anchor regardless of fundamentals |
People do not process probabilities linearly. They overweight small probabilities and underweight large probabilities:
Implementation rule: When evaluating any position, force yourself to write down the probability of each outcome. Then ask: "Am I overweighting the exciting but unlikely scenario? Am I underweighting the boring but probable one?"
Mental accounting is the tendency to treat money differently depending on where it came from, where it is kept, or what it is earmarked for — even though money is perfectly fungible.
Investors mentally segregate their portfolio into separate accounts:
Total Wealth (should be treated as one portfolio)
├── "Safe" bucket — bonds, cash, CDs (for security)
├── "Income" bucket — dividend stocks (for living expenses)
├── "Growth" bucket — stocks (for wealth building)
└── "Speculation" bucket — options, crypto (for excitement)
The problem: These buckets are managed independently, ignoring correlations. An investor might hold bonds "for safety" while simultaneously taking leveraged bets in the speculation bucket, resulting in a total portfolio risk profile they would never consciously choose.
Evaluating each investment in isolation rather than as part of a total portfolio. A stock that looks terrible alone might be an excellent diversifier. A "safe" bond might be redundant if you already hold 60% fixed income.
Implementation rule: Review your total portfolio as a single entity at least quarterly. Calculate aggregate statistics — total equity exposure, sector concentration, geographic allocation, correlation matrix — rather than evaluating each position in its own mental bucket.
The disposition effect is one of the most robust findings in behavioral finance: investors sell winners too early and hold losers too long. This is a direct consequence of prospect theory's value function operating around the reference point of purchase price.
The mathematics of destruction:
| Scenario | Action | Consequence |
|---|---|---|
| Stock up 20% | Sell to lock profit | Miss potential further gains; realize taxable |
| gain; pay capital gains tax immediately | ||
| Stock down 20% | Hold hoping to | Tie up capital in underperformer; miss |
| break even | opportunity cost of redeployment; potential | |
| further losses |
Tax consequence: The disposition effect is particularly devastating after taxes. Selling winners generates immediate tax liability. Holding losers delays the tax benefit of loss harvesting. The investor does the exact opposite of what tax optimization demands.
After a loss, investors become risk-seeking — they want to gamble to get back to their reference point. This leads to:
Money already spent should not influence future decisions — yet it does. Investors continue holding a bad position because they "already have so much invested." The sunk cost is gone regardless of the future decision. The only relevant question is: "Given where this stock is now, would I buy it today?"
Implementation rule: For every position currently in your portfolio, ask: "If I held cash instead, would I buy this stock today at this price with this information?" If the answer is no, sell. Your purchase price is irrelevant to the stock's future prospects.
Heuristics are mental shortcuts that work well in everyday life but create systematic errors in complex domains like investing. Nofsinger catalogs the most destructive ones.
The first piece of information received about a value disproportionately influences subsequent estimates. In investing:
Implementation rule: When analyzing a stock, deliberately avoid looking at the current price first. Estimate intrinsic value from fundamentals alone, then compare to market price. If you cannot do this, you are anchoring.
Judging the probability of an event by how much it resembles a stereotype or pattern, rather than by base rates and statistical logic.
Common manifestations:
Events that are easy to recall (vivid, recent, emotional) are judged as more likely.
Implementation rule: When assessing risk, use historical base rates, not recent memory. Build a probability reference sheet: "What is the actual frequency of a >20% drawdown? A >40% drawdown? How long do recoveries take historically?"
Investors overwhelmingly prefer what they know:
The preference for the current state of affairs. Changing a portfolio requires effort and triggers loss aversion (what if the change is wrong?). The result: investors keep default allocations, stay in underperforming funds, and fail to rebalance — not from analysis but from inertia.
Implementation rule: Schedule a quarterly portfolio review with a specific checklist. The default should be "review and potentially change" rather than "keep unless something dramatic happens."
Humans are social animals. When others are buying, we feel safety in joining them; when others are selling, we feel danger in standing alone. Herding produces bubbles and crashes — the crowd amplifies directional moves far beyond fundamental value.
The herding cycle:
Stage 1: Smart money identifies opportunity → modest price increase
Stage 2: Early followers join → price increase accelerates
Stage 3: Media coverage → mainstream investors pile in → euphoria
Stage 4: Late-comers buy at peak → "this time is different"
Stage 5: Smart money exits → initial decline
Stage 6: Panic selling → crash → media coverage → more selling
Stage 7: Capitulation → bottom → cycle restarts
An information cascade occurs when individuals, observing the actions of those before them, rationally decide to follow the crowd regardless of their own private information.
Example: You research a stock and believe it is overvalued. But you see that five respected analysts have buy ratings, three friends own it, and the stock keeps going up. You abandon your private signal and buy. If enough people do this, the crowd's "wisdom" is actually just the first few people's opinion amplified.
When cascades break: Cascades are fragile. A single credible contrarian signal — a short-seller report, an earnings miss, a fraud revelation — can reverse the cascade instantly, producing the violent reversals we see at market tops.
Financial media amplifies biases rather than correcting them:
| Media Behavior | Bias Amplified |
|---|---|
| Sensational headlines | Availability bias, fear, recency |
| Featuring star fund managers | Survivorship bias, representativeness |
| Reporting daily market moves | Overtrading, narrow framing, loss aversion |
| "Expert" predictions | Anchoring, authority bias, illusion of knowledge |
| Success stories | Self-attribution, overconfidence |
Implementation rule: Limit financial media consumption. News is noise on investment time horizons. Set specific, scheduled times to check portfolio performance (monthly, not daily). Unsubscribe from real-time alerts.
Research demonstrates robust links between mood and financial decisions:
These findings are disturbing precisely because they have nothing to do with fundamentals.
People judge risks and benefits by how they feel about something, not by objective analysis. If an investment makes you feel good (exciting story, admired brand, social cachet), you perceive its risks as lower and its returns as higher. If an investment makes you feel bad (boring company, stigmatized industry, complex structure), you perceive risks as higher and returns as lower.
The inversion of reality: Often, the investments that feel best are the most dangerous (late-stage bubbles feel euphoric) and the investments that feel worst are the most rewarding (buying during a panic feels terrible but historically produces the best returns).
Nofsinger maps the emotional cycle of investing:
Optimism → Excitement → Thrill → Euphoria (maximum financial risk)
↓
Anxiety → Denial → Fear → Desperation → Panic
↓
Capitulation → Despondency → Depression (maximum financial opportunity)
↓
Hope → Relief → Optimism (cycle restarts)
Critical insight: Emotional extremes correspond inversely to financial opportunity. When you feel best about the market, forward returns are lowest. When you feel worst, forward returns are highest. Your feelings are a contrarian indicator.
Implementation rule: Develop a personal "emotion thermometer." Before making any significant portfolio decision, rate your current emotional state on a scale from -5 (panic) to +5 (euphoria). Decisions made at extremes (below -3 or above +3) should trigger an automatic 48-hour cooling-off period.
Cognitive dissonance is the psychological discomfort experienced when holding two contradictory beliefs. In investing, it typically arises when new information contradicts an existing position.
Example: You own Stock X. A credible analyst publishes a detailed bearish report. This creates dissonance: "I am a smart investor" conflicts with "I own a stock that a smart analyst says is overvalued." Resolution strategies:
All four strategies preserve the ego but destroy the portfolio.
The tendency to seek, interpret, and remember information that confirms pre-existing beliefs. Confirmation bias is the operational arm of cognitive dissonance — it is how dissonance avoidance plays out in practice.
In investment research:
The asymmetric information diet:
After buying Stock X:
Read: bullish articles, bull case forums, management interviews
Ignore: bearish reports, short-seller analyses, competitor threats
Result: ever-increasing conviction disconnected from reality
Implementation rule: For every investment thesis, actively seek the strongest counter-argument. Assign a "devil's advocate" role — or become your own by writing the bear case before you buy. If you cannot articulate three credible reasons your thesis might be wrong, you do not understand the investment well enough to own it.
People systematically overvalue immediate rewards relative to future ones — and the discount rate is not constant. The difference between "now" and "one year from now" feels much larger than the difference between "ten years from now" and "eleven years from now."
Consequences for investing:
Even investors who intellectually understand the right strategy often lack the self-control to execute it. Nofsinger frames this as a battle between the "planner self" (rational, long-term) and the "doer self" (emotional, present-focused).
| Planner Self Says | Doer Self Says |
|---|---|
| Stay the course during downturns | "This time is different — sell now!" |
| Rebalance systematically | "Winners are winning, why would I sell them?" |
| Save 20% of income | "I deserve this purchase now" |
| Ignore daily market moves | "Let me just check the portfolio..." |
| Diversify globally | "I know US tech, I'll stick with that" |
The most effective strategy is removing the decision from the moment of temptation:
Primary biases: Overconfidence, illusion of control, self-attribution
Overconfident investors trade 45-75% more than average. Each trade incurs costs (commissions, bid-ask spread, market impact, taxes). The most active quintile of traders underperform the least active quintile by approximately 6-7% annually.
Rule: Impose a mandatory trading budget. Limit yourself to N trades per quarter. Each trade requires a written thesis filed before execution.
Primary biases: Familiarity, home bias, illusion of knowledge, representativeness
Investors hold a median of 3-4 stocks in individual accounts when theory suggests a minimum of 20-30 for basic diversification. They concentrate in familiar sectors, domestic markets, and employer stock.
Rule: Define minimum diversification criteria before building a portfolio. No single position >5% of total portfolio. No single sector >25%. Domestic allocation should not exceed global market-cap weighting by more than 20 percentage points.
Primary biases: Availability, affect, overconfidence, recency
After a crash, investors overestimate future risk (availability of recent pain). During a bull market, investors underestimate risk (recency of gains, positive affect). Objective risk measures (volatility, drawdown frequency, correlation) do not change nearly as much as perceived risk.
Rule: Use quantitative risk measures, not feelings. Define risk tolerance in advance using drawdown scenarios: "I can tolerate a X% portfolio decline over Y months without changing strategy." Write it down, sign it, revisit only annually.
Primary biases: Disposition effect, sunk cost, anchoring, loss aversion
Selling is the hardest investment decision because it triggers multiple overlapping biases. Nofsinger identifies selling as the decision most corrupted by psychology.
Rule: Define sell criteria at the time of purchase — before any bias can form:
Sell Rules (defined at purchase):
1. Stop-loss: Sell if position declines X% from purchase price
2. Thesis violation: Sell if the original investment thesis is broken
3. Valuation target: Sell if price reaches Y (pre-calculated fair value)
4. Time limit: Re-evaluate if thesis has not played out within Z months
5. Opportunity cost: Sell if a clearly superior opportunity emerges
Nofsinger draws important distinctions between how biases manifest differently in professional and individual contexts.
| Bias | Typical Manifestation |
|---|---|
| Overconfidence | Excessive trading, concentrated portfolios |
| Disposition effect | Holding losers, selling winners |
| Home bias | 80%+ domestic allocation |
| Familiarity | Employer stock, local companies, own industry |
| Herding | Buying tops, selling bottoms, chasing performance |
| Status quo | Keeping default allocations, failure to rebalance |
| Framing | Evaluating positions individually, not as portfolio |
Institutions are not immune — they are simply biased differently:
| Bias | Typical Manifestation |
|---|---|
| Career risk | Herding to consensus ("no one got fired for buying IBM") |
| Short-termism | Managing to quarterly benchmarks, ignoring long-term |
| Anchoring to benchmarks | Tracking error aversion overrides return maximization |
| Groupthink | Investment committees reinforce consensus, suppress |
| dissent | |
| Window dressing | Buying winners and selling losers before quarter-end |
| to make holdings look prescient | |
| Overconfidence in models | Mistaking model precision for accuracy |
| Herding across firms | Everyone owns the same "must-own" stocks |
Individual investors have the structural advantage of patience — no quarterly reporting, no career risk, no benchmark tracking — but they squander it through behavioral errors. Institutional investors have the informational and analytical advantage but are constrained by career incentives that produce their own behavioral distortions.
Implementation insight: The individual investor's greatest edge is the ability to do nothing, hold for decades, ignore benchmarks, and act contrarily. These are precisely the actions that behavioral biases prevent.
Nofsinger identifies recurring patterns that cost investors the most money:
| # | Mistake | Root Bias(es) |
|---|---|---|
| 1 | Trading too frequently | Overconfidence, illusion of control |
| 2 | Holding losers too long | Loss aversion, sunk cost, anchoring |
| 3 | Selling winners too early | Disposition effect, risk aversion in |
| gains domain | ||
| 4 | Under-diversification | Familiarity, overconfidence, home bias |
| 5 | Chasing past performance | Representativeness, recency, herding |
| 6 | Buying high, selling low | Herding, fear and greed cycle |
| 7 | Ignoring fees and taxes | Mental accounting, narrow framing |
| 8 | Overweighting employer stock | Familiarity, loyalty, illusion of |
| knowledge | ||
| 9 | Checking portfolio too often | Myopic loss aversion, illusion of |
| control | ||
| 10 | Reacting to financial media | Availability, authority bias, recency |
| 11 | Anchoring to purchase price | Anchoring, reference dependence |
| 12 | Confirmation bias in research | Cognitive dissonance, confirmation bias |
| 13 | Panic selling during downturns | Loss aversion, availability, fear |
| 14 | Failing to rebalance | Status quo, disposition effect |
| 15 | Overcomplicating the portfolio | Illusion of control, overconfidence |
Nofsinger cites research estimating the aggregate cost of behavioral errors:
Knowing about biases does not cure them — awareness alone is necessary but insufficient. Nofsinger emphasizes that structural solutions (changing the environment) beat willpower solutions (trying harder to be rational).
Replace discretionary decisions with rules wherever possible:
Discretionary: "I'll sell when I feel the stock has peaked"
Systematic: "I'll sell 25% of the position when it reaches 2x my purchase price,
25% at 3x, and trailing-stop the remainder at 20% below peak"
Discretionary: "I'll buy more if it drops to a good price"
Systematic: "I'll add to the position only if (a) the thesis is intact, (b) the
price drops >20% from my entry, and (c) the total position remains
under 5% of portfolio"
Adapt the aviation/surgery checklist revolution to investing. Before every significant portfolio decision, require completion of a standardized checklist:
Pre-Trade Checklist:
Bind your future self to rational behavior before the emotional moment arrives:
The simplest and most effective debiasing tool. Most impulsive trades that feel urgent will feel much less compelling after 48 hours. The market will still be there.
Protocol:
External accountability dramatically reduces behavioral errors:
Run this assessment quarterly to identify which biases are currently affecting your portfolio:
Portfolio Bias Audit:
OVERCONFIDENCE INDICATORS
[ ] Have I traded more than [budget] times this quarter?
[ ] Do I hold fewer than 15 positions?
[ ] Is any single position >7% of my portfolio?
[ ] Do I believe I can consistently beat the market?
[ ] Have I recently told someone about a successful trade?
Score: ___/5 (higher = more overconfident)
LOSS AVERSION / DISPOSITION INDICATORS
[ ] Am I holding any position mainly because "it'll come back"?
[ ] Have I sold any winner in the last quarter for non-fundamental reasons?
[ ] Is my average holding period for losers longer than for winners?
[ ] Do I feel physically uncomfortable thinking about selling a loser?
[ ] Have I checked my losers' purchase prices in the last week?
Score: ___/5 (higher = more loss-averse/disposition-prone)
HERDING / SOCIAL INDICATORS
[ ] Did I buy anything this quarter because someone recommended it?
[ ] Did I sell anything because of a media story?
[ ] Do I check financial news more than once per day?
[ ] Am I in any investment chat groups that influence my decisions?
[ ] Have I recently changed strategy based on market consensus?
Score: ___/5 (higher = more socially influenced)
ANCHORING / FAMILIARITY INDICATORS
[ ] Is more than 70% of my portfolio in domestic equities?
[ ] Do I own stock in my employer or industry?
[ ] Am I holding any position because of its past price, not future value?
[ ] Do I evaluate positions by what I paid rather than what they're worth?
[ ] Am I avoiding unfamiliar asset classes or geographies?
Score: ___/5 (higher = more anchored/familiarity-biased)
EMOTIONAL INDICATORS
[ ] Have I made any trade while feeling strong emotions (fear, excitement)?
[ ] Did I check my portfolio more than twice this week?
[ ] Do I feel anxious when I can't check my portfolio?
[ ] Have I lost sleep over an investment decision this quarter?
[ ] Do I feel personally attacked when someone criticizes my holdings?
Score: ___/5 (higher = more emotionally driven)
TOTAL SCORE: ___/25
0-5: Low bias risk — maintain current protocols
6-12: Moderate — review and tighten systematic rules
13-18: High — activate all debiasing protocols; consider advisor consultation
19-25: Critical — stop discretionary trading; move to fully systematic approach
Maintain a structured decision journal for every trade:
Entry Template:
Date: _______________
Asset: _______________
Action: Buy / Sell / Hold (reaffirm)
Position Size: _______% of portfolio
THESIS (one paragraph max):
_________________________________________________________________
DISCONFIRMING EVIDENCE (minimum 3 points):
1. _______________________________________________________________
2. _______________________________________________________________
3. _______________________________________________________________
SELL CRITERIA (defined now, before ownership bias forms):
Stop-loss level: $_____ (___% below entry)
Target price: $_____ (___% above entry)
Thesis review date: _______________
Maximum hold period: _______________
EMOTIONAL STATE: ___/10 (1=panicked, 5=neutral, 10=euphoric)
If below 3 or above 7: DELAY 48 HOURS
CHECKLIST COMPLETED: [ ] Yes [ ] No
If No: DO NOT EXECUTE
POST-TRADE REVIEW (fill in 3 months later):
Outcome: _______________
Was thesis correct? _______________
What did I learn? _______________
What bias, if any, influenced this decision in retrospect? _______________
A quarterly scoring system to detect portfolio-level bias drift:
METRIC TARGET ACTUAL STATUS
─────────────────────────────────────────────────────────────────────
Domestic allocation ≤70% ____% [ ]
Largest single position ≤5% ____% [ ]
Top 5 positions concentration ≤25% ____% [ ]
Number of holdings ≥15 ____ [ ]
Sector concentration (top sector) ≤30% ____% [ ]
Trades this quarter ≤[budget] ____ [ ]
Avg holding period (winners) ___months ____ [ ]
Avg holding period (losers) ___months ____ [ ]
Winner/Loser hold ratio ≥1.0 ____ [ ]
(If <1.0, disposition effect is active)
Portfolio checked per week ≤2 ____ [ ]
Unrealized losses held >12mo ≤2 positions ____ [ ]
(If >2, loss aversion is active)
Trades driven by media/social 0 ____ [ ]
Trades with completed checklist 100% ____% [ ]
Once per year, conduct a comprehensive behavioral audit:
Nofsinger's work distills into principles that every investor should internalize:
Knowing about biases does not prevent them. You need structural defenses — rules, checklists, automation, accountability — that operate when your rationality fails.
When investing feels most exciting, risk is highest. When it feels most terrifying, opportunity is greatest. Never trust emotional comfort as a signal of investment quality.
For most investors, doing less — trading less, checking less, researching less, reacting less — would improve returns more than any strategy change. The default should be to do nothing; action should require justification.
Your purchase price, the 52-week high, and the "round number" target have no causal relationship to what a stock will do next. Only future cash flows and the price you pay for them matter.
Under-diversification is not a strategy — it is a bias (familiarity, overconfidence, illusion of knowledge) masquerading as conviction. Genuine conviction with proper position sizing is wise; concentration from behavioral blindness is reckless.
A good process will sometimes produce bad outcomes (randomness exists). A bad process will sometimes produce good outcomes (luck exists). Over time, only process matters. Judge yourself by the quality of your decisions, not the results of individual trades.
The market does not know or care what you paid. Your break-even point, your cost basis, and your profit target exist only in your mind. The stock will go where fundamentals and sentiment take it, indifferent to your psychology.
The crowd is loudest and most unanimous at tops and bottoms — precisely when it is most wrong. When "everyone knows" something about the market, that information is already priced in and the contrarian position is likely more profitable.
If you cannot stop yourself from panic-selling, automate your contributions and rebalancing. If you cannot stop checking prices, delete the app and check monthly via statements. Change the environment rather than fighting your nature.
In a market where informational edges are competed away in milliseconds, the last remaining edge for individual investors is behavioral. The investor who can remain rational when others are emotional, patient when others are impulsive, and disciplined when others are panicking will outperform — not because they are smarter, but because they are more self-aware.
The ultimate lesson of Investment Psychology is humility. The market is a mirror that reflects every cognitive bias, emotional weakness, and self-deception we carry. The investors who succeed are not those who believe they have conquered their biases, but those who build systems acknowledging they never will.