The Psychology of Investing

This is a university textbook on behavioral finance, not a trading psychology book, and you need to know that before you buy it. Nofsinger names each bias, then proves it with the academic study underneath and attaches a dollar cost. He tells you exactly what is wrong with you. He does not hand you a fix.

The one idea

The argument is that your investing errors are not random, they are predictable, and every one of them has a price tag. Traditional finance was built on two assumptions: people make rational decisions and people are unbiased about the future. Both are false in ways you can measure. So Nofsinger's move is to refuse to let any bias stay a feeling. Overconfidence is not a mood, it is a portfolio turning over 250 percent a year and netting 11.4 percent while the calmest investors net 18.5 percent from the same 18.7 percent gross return. Regret is not a vibe, it is selling winners that go on to beat the market by 2.35 percent while you keep losers that trail it by 1.06 percent.

Every chapter runs the identical loop, which is why the book reads fast. Name the bias. Show it in ordinary life, usually driving or gambling or buying a washing machine. Cite the study that found it in real brokerage accounts. Price the damage. The studies are the real content: 78,000 households at a discount broker, 10,000 trading accounts, 37,000 Swedish twins, 26 stock exchanges, 1,100 soccer matches. Not one person in this book has a name or tells a story. That is the point and it is also the trade.

Which is where it sits against the other psychology books here. Douglas, Tendler and Hougaard are first-person. They treat your mind as equipment you can retrain, and their evidence is the author and the author's clients. Nofsinger is the source material those books quote without footnotes. When Hougaard says normal thinking loses money, this is where the proof lives, in journals, with sample sizes. The cost is that a book built on population averages has almost nothing to say about you specifically at 10:15 on a Tuesday. There is no drill in here, no journal, no routine, no exercise. Its subject is investors and retirement savers, not traders.

What it actually teaches

There is no procedure in this book, so what follows is the diagnostic sequence it actually delivers: the biases stacked in the order Nofsinger stacks them, with the numbers he uses to prove each one.

1. START WITH THE VALUE FUNCTION

   Everything downstream runs on prospect theory. You do not judge money in absolute terms, you judge it against a reference point, and the purchase price is usually that point. Gains and losses then run through an S-shaped curve that is steeper on the loss side, so losing $500 hurts more than gaining $500 pleases. Chapter 1 makes you feel the miscalibration first with a ten-question quiz where you give a range you are 90 percent sure contains the answer. You should miss one. Most people miss five or more, and Nofsinger notes that most finance professors miss at least five too. Then the anchor demonstration: the Dow started 1929 at 300 and ended 2016 at 19,787, but with dividends reinvested it would have been 613,514, which is a 9.05 percent annual return. Almost nobody guesses inside their own 90 percent range, because they anchored on 19,787.

2. OVERCONFIDENCE, PRICED IN TURNOVER

   The chapter that does the most work. Barber and Odean sorted 78,000 household accounts from 1991 to 1996 into five turnover groups. All five earned the same 18.7 percent gross. The lowest turnover group (2.4 percent a year) netted 18.5 percent. The highest (over 250 percent a year) netted 11.4 percent. Seven points a year, paid in commissions and in worse stock picks: over the year after a swap, the stock sold beat the stock bought by 5.8 percent. Turnover also splits by who you are. Across 38,000 households from 1991 to 1997, single men turned over 85 percent a year, married men 73, married women 53, single women 51. And going online made it worse: 1,607 investors who switched from phone to internet trading went from 70 percent turnover to 120 percent, and from earning 18 percent a year (2.35 above the market) to 12 percent (3.5 below it). The mechanism is the illusion of knowledge. In a football prediction experiment, giving people five blocks of information instead of one moved accuracy from 64 percent to 66 percent and moved confidence from 69 percent to 79 percent.

3. THE DISPOSITION EFFECT

   Shefrin and Statman's name for selling winners too early and riding losers too long, driven by seeking pride and dodging regret. Odean measured it across 10,000 accounts from 1987 to 1993: when investors sold a winner, that sale represented 23 percent of the paper gains available to them, but a loser sale represented only 15.5 percent of the paper losses available. Net effect, investors are 50 percent more likely to sell a winner than a loser. Holding periods give it away too, in an older sample of 75,000 round trips: positions closed inside a month returned 45 percent annualized, one to six months 7.8 percent, six to twelve 5.1 percent, over a year 4.5 percent. In Israel, individual investors held winners 20 days and losers 43. It is not a retail-only problem: professional futures traders, corporate option holders and 30 percent of mutual funds show it, and those funds underperform the rest by 4 to 6 percent a year. One oddity worth knowing: the disposition effect disappears in stocks that have recently split, because the split scrambles the reference point.

4. YOUR RISK APPETITE MOVES WITH YOUR LAST RESULT

   Thaler and Johnson ran 95 students through two-step gambles with real money. After a $15 windfall, 77 percent took a coin-flip bet they would otherwise refuse. Without the windfall, only 41 percent took it. That is the house money effect. After losing $7.50, 60 percent declined the next bet, which is the snakebite. But the break-even drive beats both: after a loss, a majority accepted a double-or-nothing flip even when told the coin was not fair. Coval and Shumway found the professional version in 426 Chicago Board of Trade bond futures traders in 1998, who raised afternoon risk after losing in the morning, and lost on those trades on average. Then the part that should bother you most: memory does not preserve any of this. Architects surveyed about their own fund returns overstated them by 6.22 percent and overstated their margin over the market by 4.62 percent. Educated AAII members overstated by 3.40 and 5.11. German online investors overstated their annual returns by 11.6 percent, and experience barely helped (13.2 percent for beginners, 10.3 for the experienced).

5. THE SHORTCUTS: STEREOTYPE, FAMILIARITY, ONE OVER N

   Representativeness is judging by resemblance, which is why people confuse a good company with a good stock. Lakonishok, Shleifer and Vishny sorted US stocks by five-year sales growth from 1963 to 1990: the glamour decile returned 11.4 percent the following year against 18.7 percent for value, and 81.8 percent against 143.4 percent over five years. Familiarity is worse because it feels like prudence. The US is 43 percent of world market cap and US investors hold 87 percent domestic; Japan 91 percent, the UK 72 percent. The average US household holds 30 percent of its portfolio in companies headquartered within 250 miles. Georgians own 16 percent of Coca-Cola. Professional managers buy firms headquartered on average 100 miles closer to their own office than the typical US company. And the diversification people think they have is often just an even split across whatever is on the menu. TWA pilots, offered five stock funds and one bond fund, ran 75 percent equities against a 57 percent national average. University of California employees, offered one stock fund and four bond funds, ran 34 percent. In one 7,000-account 401(k) study, 47 percent held no equities at all and 22 percent held nothing but, so 69 percent were completely undiversified.

6. MOOD, CROWD AND BIOLOGY

   The back half widens out and gets less reliable, so read it knowing that. Across 26 exchanges, sunny days beat the most miserable weather days by 24.6 percent annualized (15 percent on the NYSE, 22.1 in London, 4.1 in Copenhagen). Across 48 markets, returns ran 3 to 5 percent a year lower in the seven days around a full moon. A national soccer loss knocks 0.21 percent off that country's market the next day, 0.38 after an elimination match, 0.49 in the World Cup, with no matching bump after wins. On the crowd side, the useful number is investment clubs: the press claimed 60 percent of clubs beat the market, but a study of 166 clubs over five years found gross returns of 17 percent against the S&P's 18 percent, and 14.1 percent net, with roughly 60 percent underperforming. The survey the press was quoting had a 5 to 10 percent response rate. Chapter 12 is new to this edition and is the most interesting: across 37,000 Swedish twins, genetics explain about 29 percent of how much of your portfolio sits in stocks (44.5 percent for those under 30) and roughly a fifth to a third of the biases themselves. Investment skill drops sharply after age 70, costing an estimated 3 percent a year in risk-adjusted return, over 5 percent for older investors with large portfolios.

7. THE PRESCRIPTION, SUCH AS IT IS

   Chapter 11 is the only place the book tries to fix anything, and it is thin. Six strategies: understand the biases, know why you are investing (with a written dollar goal, not 'I want to travel'), have quantitative criteria you check before every purchase, diversify (15 stocks across industries and sizes, own very little of your employer, hold some bonds), control your environment, and set reminders. The environment rules are the concrete part: check your stocks once a month rather than once an hour, place all buys and sells on one fixed day of the month, review the whole portfolio once a year against your written goal. Then four rules of thumb: avoid stocks under $5, treat chat rooms and message boards as entertainment only, do not trade outside your criteria unless you genuinely believe you know more than the market, and strive to earn the market return rather than beat it. The chapter's most persuasive evidence is not about you at all, it is about defaults. Save More Tomorrow took 162 employees from a 3.5 percent savings rate to 11.6 percent over three pay raises, while non-joiners went from 5.3 to 7.5.

What it looks like Monday morning

You measure yourself instead of trusting your memory, because the strongest finding in the book is that your recollection of your own returns is inflated by 3 to 12 percent. Pull your last twelve months of statements. Write down your actual return before you look it up, then look it up, and keep the gap. Compute your turnover. Then run Odean's split by hand: for every position you closed, was it up or down, and what else was sitting in the account unrealized at that moment? If you are consistently closing gains while the losses stay open, you have the disposition effect on paper rather than as a suspicion, and you now have a number to move.

The behavioral rules are harder to lift straight across, because they were written for someone with a 401(k) and not for someone with a chart open. Nofsinger's remedy for overtrading is to trade once a month on a fixed date, which is not advice an active trader can take. What does transfer is the environment principle underneath it: decide in advance, in writing, while calm, what qualifies as a trade, and check the position less often than you want to. Also stop treating a company you admire as a stock you should own, because the glamour decile lost to value by seven points a year in the study he leans on.

Reproduced as printed in Table 3.1 of the 6th edition, including its arithmetic slip: a $833 basis makes the gain $167 and the tax $25.05, not the $177 and $26.55 shown. The conclusion holds either way, and the point is that the disposition effect makes people sell the position that costs them money to sell.

By learning about your psychological biases, you can overcome them and increase your wealth.John R. Nofsinger, The Psychology of Investing

Who it is for