Psychology and how accounts die

OpeningA 92% win rate and a smaller account

A $100 stock with implied volatility of 30%. Every Monday you sell 10 puts with a 95 strike that expire in 7 days, collecting $0.20 on each, $203 a week.

The option’s Delta is −0.10, and backing it out from the option price, the probability of finishing in the money is 10.9%. The market’s quote is already telling you that, on average, about one week in nine will lose money.

For the first 11 weeks the stock never touches 95, every week’s options expire worthless, and you steadily collect $2,238. In week 12 the company releases bad news over the weekend, and the stock opens Monday at 88, down 12%. That week’s loss is (95 − 88 − 0.20) × 1,000 = $6,797.

A win rate of 11/12, about 92%, and yet the account is $4,559 smaller. Week 12 alone lost three times the profit of the first 11 weeks. The first 11 weeks are the bigger problem: each one reinforced the same belief, “this is easy money,” and before week 12 arrived you had no reason at all to doubt it.

This is what an options psychology problem typically looks like. The shape of the payoff distribution gives you the wrong feedback most of the time. This chapter skips general trading psychology and covers only the part specific to options: which biases options amplify, the paths along which they take an account to zero, and when you can see it coming.

15.1Which biases options amplify

Options shape payoffs into two extreme forms: for buyers, “frequent small losses, occasional big wins”; for sellers, “frequent small wins, occasional big losses.” Both shapes line up exactly with the tendencies people most easily get wrong.

At fair Black-Scholes prices, all four positions have an expectancy of about zero. Their shapes could hardly be more different: the 0DTE buyer loses everything 90% of the time and makes 8 times the cost on average when winning; the seller of the 7-day put keeps the full premium 90% of the time, and in the worst 1% loses 20 times the premium in one go.

Three amplified tendencies

• Overweighting small probabilities. People tend to feel “small chance, big payoff” events as weightier than their true probability. An option at $0.15 that might become $1.50 is too vivid a picture: a 10% chance feels like 30%. That makes buyers willing to overpay for OTM and 0DTE options.
• Drawing certainty from small samples. A seller with a 90% win rate often wins a dozen or more times in a row. People judge risk by recent experience, and after a dozen-plus trades without trouble they feel “nothing will go wrong.” But that 10% of losses rarely shows up in a sample of a dozen or so to begin with (Chapter 14: the probability of not seeing a 5% tail in 28 trades is 23.8%).
• Loss aversion and the disposition effect. People tend to take profits early and hang on to losers. Hanging on to a stock at worst means a slow decline; hanging on to an option also means facing expiration and Gamma: the longer a seller holds on, the bigger the Delta gets, and the longer a buyer holds on, the less time value is left.

Four amplifiers unique to options

15.2Buyer psychology: lottery tickets, 0DTE options and small losses that add up

The buyer’s problem isn’t any single loss. Each one is small, and the occasional win is highly visible, so a negative-expectancy habit can go on for a long time without being noticed.

The lottery mindset: looking at the payout, not the probability

Take an index ETF priced at $600 (prices are hypothetical). At 10 a.m. you buy a call that expires the same day, with a 603 strike (0.5% OTM), about 6 hours left and IV 16%. The mid is 0.146; filling at the ask costs half the spread, $0.01, more, for a cost of 0.156.

• Probability of finishing in the money: 11.8%.
• In the 10% or so of cases that win, the average gain is 7.8 times the cost and the median 6.0 times. That’s what the screenshots look like.
• Expectancy per trade: with perfectly fair pricing, about −7.2% of cost. If realized volatility that day is only 85% of implied, it drops to −49.4%.

Most of the −7.2% comes from that one cent: $0.01 is 6.8% of $0.146. The cheaper the option, the bigger the share that fixed cent takes. The same day’s 606 call (1% OTM) has a mid of just 0.008; adding a cent more than doubles the cost, and even with fair pricing the expectancy per trade is −57.6%. The cheapest lottery tickets have the worst odds.

Chasing 0DTEs: there’s a new draw every day

There are 0DTE options every day. Lose yesterday, and today there’s a fresh one to “win it back,” with no cooling-off period. In the example above, the chance of losing any one trade is about 88%, the chance of losing five days in a row is 0.882⁵ = 53%, and even ten in a row is 28.5%. Losing streaks are the norm, not bad luck.

What’s really dangerous is how you react after a loss. “I have to win it back today” makes each bet bigger and bigger, and that is how the first path to a dead account in Section 15.6 forms.

Small losses that add up: none of them hurts

At $300 a trade, no single loss sets off any alarm. But one trade every trading day is 20 a month, and even with perfectly fair pricing the expectancy is 20 × 300 × (−7.2%) = −$432; when realized volatility is only 85% of implied, it is −$2,964, half of a $6,000 monthly budget.

When buying OTM options makes sense

There’s nothing wrong in itself with buying cheap OTM options. In these three situations, there is a sound reason for it:

• You have reason to think realized volatility will exceed implied. For example, IV is low, there’s a clear catalyst, and the market hasn’t priced it in (the “low IV” row of the Chapter 14 matrix). Then the expectancy can be positive, provided you can say what the market has missed.
• It’s insurance, not a lottery ticket. An OTM put bought for a stock holding or a short-option position is there to cut off the tail, and its slightly negative expectancy is the insurance premium (Chapter 11). The test is whether it covers a specific risk you can’t afford to bear.
• It’s budgeted and defined in advance as a low-probability trade. Before entering, write down “this will most likely go to zero”; the money comes from the monthly lottery budget, not from your capital.

A 0DTE option that meets none of the three is, in essence, paid entertainment. That’s fine in itself, as long as you book it as entertainment, keep it out of your trading system’s results, and don’t use trading reasons to size up after losses.

15.3Seller psychology: the “easy money” illusion

The seller’s danger comes from feedback that is too good. A string of small wins convinces you the risk isn’t there, so you add when you should least add, and wait when you should most admit you’re wrong.

The “easy money” illusion

The chart below stretches the opening example to a year: every week you sell a 7-day put with a Delta of −0.15, priced at IV 30%, while the stock’s normal realized volatility is only 20% (the seller has an edge), with 8% annual drift. In week 47 the stock gaps down 18% in one move. The number of contracts is recalculated each week from a fixed “notional ÷ account equity” multiple, so it grows automatically as the account grows.

The most deceptive part of this curve is the first 46 weeks. Cut out any stretch of it and it looks like a low-volatility fund rising steadily. A selling strategy’s risk is concentrated in a very few days, and everyday volatility doesn’t reflect it at all. Measure a short-option position by its everyday drawdowns or volatility and you will systematically underestimate the risk.

Inflating size after a winning streak

A winning streak makes you feel “I’ve figured it out,” yet for sellers, streaks come almost built into the strategy’s structure: a strategy with a 90% win rate has a 35% chance of winning 10 in a row. Adding after a streak means pushing your size to its maximum just before the next loss arrives. Section 15.7 uses a simulation to show what this does to the spread of account outcomes.

Adding to a breached position to repair it

When a short leg is breached, a common impulse is to sell more at a lower strike, “using new premium to lower the cost basis.” An example: you sell 5 contracts of a 30-day 95 put at 1.33 each. On day 10 the stock falls to 91, IV rises to 38%, the 95 put rises to 5.53, and you are down $2,100 on paper. You then sell 5 contracts of a 20-day 90 put at 2.64, taking in $1,322. Net the paper loss against the new premium, and the books seem to show a loss of only $778.

On day 20 the stock is at 86, IV 45%. The original 5 contracts are down $3,928 and the 5 added ones another $1,185, for a total of −$5,112. Adding to repair didn’t reduce risk; it turned one loss into two, betting on the same direction with more exposure after that direction had already been proven wrong.

Skipping the stop to “wait for expiry”

The seller’s favorite line is “it’ll be fine if I just wait for expiry.” Look at what waiting buys you: a short 95 put, stock at 90, 10 days left, IV 40%, option price 5.63, of which 5.00 is intrinsic value and only 0.63 is time value. Waiting for expiry wins back at most that 0.63, and only if the stock doesn’t move at all; yet one standard deviation over 10 days is 90 × 0.40 × √(10/365) = ±5.96. You’re taking ±6 of risk for 0.63 of time value. The Chapter 14 simulation also shows that holding to expiry, compared with stopping out by rule, nearly doubles the average loss in the worst 5%.

15.4The Theta obsession

Theta is the compensation you receive for carrying Gamma risk, not interest. When realized volatility equals implied, it exactly pays for the losses you can expect to take.

Chapter 4 derived Θ ≈ −½σ²S²Γ. With Delta hedged, one day’s P&L on a short option is approximately:

Take a short 30-day ATM straddle (stock $100, IV 30%): |Θ| = 0.114 ($11.40 per straddle per day), Γ = 0.0925. The breakeven move is √(2 × 0.114 ÷ 0.0925) = 1.57, exactly one daily standard deviation, 100 × 0.30 ÷ √365 = 1.57.

If realized volatility is exactly 30%, the average daily Gamma loss is ½ × 0.0925 × 100² × 0.30² ÷ 365 = 0.114, matching Theta to the cent. Theta is just prepaying that expected loss. A seller’s real edge comes from realized volatility below implied: if realized is only 25%, the expected daily Gamma loss drops to 0.079, and only the remaining 0.035 is profit, about 30.6% of Theta. Of the $11.40 you see each day, 70% is prepayment for future losses.

What’s wrong with the “collecting rent” metaphor

Sell a 30-day 93.5 put (stock $100, IV 30%, Delta −0.19) and collect 0.963: 1.03% a month on the cash securing it, 12.4% a year in simple interest. It looks like steady rent. But a single gap to 80 loses 93.5 − 80 − 0.963 = 12.54 per share, 13 months of “rent.” Real rent never makes you pay back thirteen months of rent in one go; short puts do.

15.5Options culture on social media

The option returns you see online are a filtered sample: winners post, losers stay quiet. So the average you see is systematically higher than the true average.

The 2021 retail options boom

From the second half of 2020 into 2021, zero commissions and trading on phones drove a sharp rise in retail options volume, mostly purchases of short-dated OTM calls. In the January 2021 surge in GameStop and other “meme stocks,” calls bought by retail traders forced market makers to buy stock to hedge, and options amplified the move (Chapter 4, Case 3). Over the following years, index options gradually gained contracts expiring every trading day, and same-day expiries’ share of volume rose markedly.

The cultural legacy of that period has outlasted the prices themselves: options were spread as a tool for “turning a little money into a lot,” discussion focused on the return of single trades, and few people talked about win rate, expectancy or position size.

The “sell puts, collect rent” narrative

The counterpart to buyer culture is “sell puts, collect rent”: sell puts on stocks you’d be happy to own and treat the premium as rent; if you’re assigned, think of it as buying the stock at a discount, then sell covered calls to keep collecting rent (the wheel, Chapter 8). This narrative is half right: if you already planned to buy the stock at that price, selling a put does bring in more than a limit order would. The half it leaves out:

• By put–call parity, a cash-secured put and a covered call are the same position: you lose like a stockholder when the stock falls, and your upside is capped when it rises.
• The “rent” is compensation for gap risk; as calculated in the previous section, a single 20% drop eats 13 months of rent.
• “Happy to own the stock” is often something said after the fact. By the time you are actually assigned, the stock is often far below the strike, and by then you may no longer want to own it.

Survivorship bias in posted wins

Here is a simulation using the 0DTE option from Section 15.2: 1,000 people each make 20 trades in a month at $300 a trade (a $6,000 monthly budget). The results:

If only the top 5% post, the average you see is +$11,435, while the true average is −$517. Half of the people had at least one 10× trade within the month, so there’s never a shortage of screenshots. And the third who lost more than half their budget almost never show up in your timeline.

15.6Five paths to a dead account

Most dead options accounts died along one of five paths. What they share: each step looks reasonable on its own, and the early warning signs are obvious; it’s just that nobody was looking at the time.

Path 1: 0DTE options expiring worthless day after day

How it happens. It starts with one small 0DTE trade a day. The chance of losing any one trade is about 88%; the chance of 5 losing days in a row is 53%, and of 6 in a row 47%. After a run of losses comes “doubling up to win it back”: $200 → $400 → $800 → $1,600 → $3,200 → $6,400, a total of $12,600 put in over 6 days, 63% of a $20,000 account. Even raising each bet only 1.5×, 6 days take $4,156.

Path 2: naked selling meets a gap

How it happens. A $50,000 account sells 8 contracts of a 30-day 95 put (Delta −0.25) at 1.33 each. Notional exposure is $76,000, 1.5 times equity. Overnight the stock gaps to 80 and IV jumps to 70%; buying back at the open costs 16.57, a loss of $12,186, 24% of the account. With no stop and the stock at 75 at expiry, the loss is $14,933. A stop order is useless against a gap: it can only fill at the next price available.

Path 3: rolling bigger and bigger

How it happens. Every time the position is breached, you “roll out further to a lower strike for a net credit.” To get that net credit, each roll needs more contracts:

Every roll shows up on the broker’s screen as a “net credit,” yet the three together lost $6,178, the contract count went from 5 to 10, and notional exposure from $47,500 to $82,000. If the stock falls to 72 by expiry, the total loss is $9,178; even if it rebounds to 78, it’s still a $3,178 loss.

Path 4: the margin spiral

How it happens. A $50,000 account sells 12 contracts of a 30-day 90 put at 0.40 each, for notional exposure of $108,000. Margin below is estimated with the common Reg T formula for naked shorts (20% of the underlying’s value minus the out-of-the-money amount, with a floor of 10% of the strike, plus the option’s current price); many brokers set higher requirements of their own, so a second column uses 30%.

The stock fell 17% and equity fell 21.7%, but the margin requirement at 30% rose from $24,481 to $41,208, up 68%. The two lines move toward each other and cross on day 12: margin call. Being forced to buy back now means filling at prices set by 70% IV, about the worst level of the whole move. The essence of the spiral: just when you most need time, the margin requirement takes it away.

Path 5: losing both ways on event days

How it happens. The day before earnings, you buy 10 ATM straddles with 7 days to expiry (stock $100, IV 85%) for $9,383; the implied expected move is about ±9.4%. After earnings the stock rises 5% and IV falls back to 35%; the straddle is worth only 5.69, a loss of $3,689; a 5% drop loses $3,882 too. You lose whether you guessed the direction right or wrong, because the actual move was smaller than the move the market had already priced in.

The next quarter, having learned your “lesson,” you sell the same straddles instead and collect $9,383. This time the stock falls 14%, the straddle rises to 13.95, and you lose $4,567. Two events, one loss on each side, about $8,256 in total. The problem isn’t whether you chose to buy or to sell; it’s that neither time did you compare the expected move with historical moves. You just flipped based on the last result.

15.7Three tools: the impulse log, the warning-sign checklist and the streak review

Psychological problems are hard to solve with willpower. A more reliable approach is to turn them into a form you must fill in before placing an order, and questions you must answer every month.

1. The options impulse log

For any trade that isn’t in your plan and that you suddenly want to make, fill in this form first, then decide whether to do it. It doesn’t forbid impulses; its job is to keep impulses and plans in separate books. After a month, count: how many impulses, how many you acted on, and how the trades you made performed against your planned trades. Most people find that impulse trades have clearly worse expectancy and cluster right after a loss or a win.

[Options impulse log] Date ____  Time ____
Trade I want to make: underlying ____  structure ____  contracts ____  amount ____
What triggered me: sudden price jump / news / someone’s screenshot or trade call / just lost a trade / just won a trade / boredom / other ____
Is it on this week’s shortlist: yes / no
Can I write the three sentences: direction ____  volatility ____  time ____
Minutes since my last trade ____; last trade: win / loss ____
Max loss on this trade ____, ____% of this month’s risk budget
If it’s a 0DTE: lottery budget used this month ____ / ____
Look again after 15 minutes: still want it / no longer want it
Final: traded / skipped
Fill in after the close: result if traded ____; hypothetical result if skipped ____

When impulse trades lose money, by the four categories of Chapter 14 it’s almost always an “execution error” or a “risk management error,” rarely a “judgment error.” Track them separately from planned trades so impulse losses don’t contaminate your assessment of the strategy itself.

2. The early-warning checklist for the five paths

The early warning signs for each path in Section 15.6 are condensed into Checklist 15-3 at the end of this chapter. Go through it at the end of every month; if any item gets a check, go back to that path’s prevention rules.

3. The streak review

Right after a winning streak is the most dangerous time, because that’s when you most want to add. Set a trigger, such as 5 wins in a row for a seller, 3 for a buyer, or a new account high, and once it fires, do a review before trading again: are your positions bigger than last month? Were there trades in this stretch that “made money but were done wrong”? Rerun the stress test. Then take a day off.

The simulation below shows why. A seller-style strategy: 80% win rate, wins of 0.27R, losses of 1R, expectancy +0.016R per trade, a thin edge but a real one. Two sizing rules: a fixed 2% risk per trade; or scaling size up 1.5× after each consecutive win, to a maximum of 8×, and back to base size after any loss. 2,000 accounts each make 100 trades, using the same random numbers.

Why does this happen? At an 80% win rate, the chance of 6 wins in a row is 0.8⁶ = 26.2%, not rare at all. Scaling 1.5× per win, those 6 trades are sized 1, 1.5, 2.25, 3.38, 5.06 and 7.59×, making 20.78 × 0.27 = 5.61R in total; trade 7 is already at the 8× cap and loses 8R in one go. Everything the streak earned is given back, plus another 2.39R. The trade after a streak has exactly the same chance of winning as any other, yet it carries the biggest size.

The mean goes up because the few accounts that keep winning make enormous amounts; but you only live one path. For a single account, what matters is the median and the worst case, and both get worse.

CasesTwo ledgers

ToolsChecklist and self-test

Your checkmarks are saved only in your own browser.

Self-test: five questions