Home › Learn › Lesson 48

The emotional cycle — fear, hope, doubt, greed, disappointment, excitement

By TradingPrimer · Published 11 Sep 2026

The usual treatment of this subject hands you a list of feelings and asks you to be stronger. That has never worked for anyone, and there is a better reason than weakness: a feeling cannot reach your account directly. It has to act through something you do with your hands, and there are only two such things once a position is open. This lesson follows those two channels, prices them in R and in dollars, and ends with the two ratios that tell you which one is leaking — neither of which requires you to be honest about how you felt.

A quarter taken off the winning side, a quarter added to the losing side, and the scale sits dead level — which is what an expectancy of zero looks like. The illustration is a metaphor; every number in this lesson is in the charts below.

What is the emotional cycle, and why is it one thing and not six?

It is the sequence of feelings that arrives, in a fairly predictable order, as one position moves against you, moves for you, and finally closes — and then spills into the position after it. The course this site is built on lists twelve emotional states a trader passes through, from anger and fear through to pride, shame and envy, each one paired with the behaviour it produces. Six of them attach specifically to a trade you are holding or have just closed, and those six are the subject of this lesson: fear, hope, doubt, greed, disappointment and excitement.

Most writing on this topic treats them as six separate faults needing six separate cures. That framing is where the whole subject goes wrong, because it is not what the evidence looks like when you go and measure it. Here is the claim this lesson is built on, and it is testable: a feeling can only reach your account through one of two doors. Either it moves an exit on the trade you are already in, or it changes which trade you take next. There is no third door. Position size would be a third door, but size is set before entry and belongs to Lesson 42 and Lesson 47; once the order is filled, the only levers still in your hand are where you get out and what you do afterwards.

So the six stations collapse into two mechanisms. Four of them — fear, hope, doubt, greed — move exits. Two of them — disappointment, excitement — move selection. Which means you do not need six fixes. You need two guards, and a way of telling which guard is failing.

Read the diagram once more and notice the ordering. Doubt appears twice, at station 1 and station 3, and it is genuinely the same feeling both times: a loss of confidence in the level you chose. What differs is which direction price happened to move first, and therefore which exit the doubt goes and touches. That is the clearest evidence that these are not separate diseases. The same internal state produces an opposite-looking action depending on whether the position is red or green.

Which two numbers does the cycle actually move?

The average win and the average loss. Nothing else. Every one of the four in-trade stations ends in one of two physical acts: closing a position earlier than the plan said, or later than the plan said.

It helps to see how tiny the acts are. Suppose you trade a method that wins 40% of the time with a planned stop at 1R and a planned target at 2.5R — a reasonable, unglamorous swing system. Its expectancy is 0.40 × 2.5 − 0.60 × 1 = +0.40R per trade. Now let the cycle do its work, mildly. On the trades that go green you close at 1.875R instead of 2.5R, because at station 3 you would rather bank something than watch it come back. On the trades that go red you get out at 1.25R instead of 1.00R, because at stations 1 and 2 the stop looked a little tight and you gave price some room. Neither of those would strike anyone as emotional trading. Both are a 25% drift.

The expectancy is zero. Not reduced — gone, to three decimal places, on a system that still wins 40% of the time and still has a genuinely positive edge written into its rules.

The dollar version decomposes exactly, and the decomposition is worth carrying around. Over 100 trades at a flat $100 risked per trade, the planned edge is 40R, or $4,000. The forty winners lose 0.625R each to early exits: 40 × $62.50 = $2,500. The sixty losers each cost an extra 0.25R: 60 × $25 = $1,500. Add them: $2,500 + $1,500 = $4,000, which is the entire edge. Two habits that each feel like nothing at the moment of acting split the whole year's profit between them.

One detail in that second figure deserves its own note, because it is the reason this failure survives so long in so many accounts: widening a stop is not only a chart decision, it is a money decision. Your position was sized so that 1R equals the percentage of the account you agreed to risk. Let the exit slide to 1.25R and a 2% risk quietly becomes 2.5% — above the ceiling set in Lesson 40, without a single conscious decision to raise it. On a $10,000 account planning to lose $100 a trade, you are now losing $125, sixty times over the next hundred trades.

How much drift can a system survive before the edge is gone?

There is an exact answer, it takes one line of algebra, and it is worth knowing because it turns a vague worry into a threshold you can check against your own statement. Write expectancy as E = pW − (1 − p), where p is the win rate and W the planned reward in R against a 1R loss. Let the cycle shave the winners by a fraction x and stretch the losers by the same fraction x. Setting the result to zero and solving gives:

x = E ÷ (pW + 1 − p)

The denominator is just the gross size of an average trade, win or lose. So the rule reads: your tolerance for sloppiness is your edge divided by the size of a typical trade. For the 40% / 1:2.5 system that is 0.40 ÷ 1.6 = 25.0%, exactly the drift used above.

What makes the number alarming is how it compares with each habit taken alone.

Either habit on its own is survivable for a long time. A trader who only ever cuts winners early can be 40% sloppy about it and still break even. A trader who only ever gives losers room can be two-thirds sloppy. But the cycle does not hand out one habit per person — it hands both to everyone, because fear on a red trade and doubt on a green one are the same reluctance viewed from two sides. Run both and the budget collapses to a quarter, which is the amount of drift nobody notices.

Why does the win rate go up while the account goes down?

Because the win rate measures which exit you reached, and you just moved the exits closer. This is the part of the lesson that catches careful people, and it deserves a measurement rather than an assertion.

Here is the test. Simulate 200,000 trades as price paths wandering between two barriers, with the drift calibrated so the planned version — stop at −1.00R, target at +2.50R — wins 40% of the time. Then replay the identical paths with the cycle's exits in place and count again.

The win rate climbs 11.3 points, from 39.7% to 51.0%. A trader reviewing his month would see a system that has crossed from losing more often than not to winning more often than not, and would reasonably conclude that whatever he changed was an improvement. It was not. The R per trade fell.

This is the same structural trap that Lesson 46 found in position sizing, arriving by a completely different route. There, the damage hid because changing your size does not change how often you are right. Here, the damage hides because changing your exits changes how often you are right in the flattering direction. Both leaks are invisible to the one statistic almost everybody tracks, which is a good argument for not tracking it on its own.

That table has one row in it that contradicts the lesson, and the honest move is to point at it rather than hope nobody looks: pushing the stop out to −1.25R raised the simulated average to +0.471R. That is a real feature of the model and a real limitation of it, and the section near the end explains exactly why — and what it means for your own stop.

Why do some traders survive the cycle and others do not?

Partly temperament, but much less than people assume. The bigger factor sits in the identity above: because tolerance equals edge divided by average trade size, two systems with the same expectancy can have wildly different room for error, and the trader never chose that room — the system shape did.

Look at the top and bottom rows. A system winning 30% of the time at a planned 1:3 tolerates 12.5% drift. A system winning 60% at 1:1.5 tolerates 38.5% — about three times as much. Take the comparison to two systems earning the same +0.50R and the gap is still there: 30% at 1:4 gives 26.3%, while 60% at 1:1.5 gives 38.5%.

Now put that next to what those two systems feel like to trade. The 30% system loses seven trades out of ten. It produces long strings of losses, and every one of them is another invitation to stations 1 and 2. The 60% system is right most days and rarely tests anyone's patience for long. So the shape that generates the most emotional pressure is also the shape that withstands the least of it. That double bind is, as far as we can tell, the real reason so many people abandon trend-following systems as “not working” while running an identical edge in a higher-win-rate form without difficulty.

The practical consequence is not “avoid low-win-rate systems”. It is that if you run one, your exit discipline has to be roughly three times tighter than a colleague's for the same result, and that requirement is arithmetic rather than moral. It also argues for keeping your first system somewhere in the middle of that table while you are still learning what your own numbers are.

What do disappointment and excitement cost on their own?

They never touch an open trade. They decide which trade exists at all — and because of that, they leave a completely different fingerprint: your exits look fine, your R-multiples look fine, and the account still underperforms the system.

Disappointment after a loss skips the next valid setup. Excitement after a win takes one that does not qualify. Price them both. Start with 100 setups that meet every one of your criteria, each worth +0.40R. A B-grade trade — one that fails a criterion — is generously assumed to be a coin flip before costs, so fees and slippage make it −0.05R. Let k be the share of occasions on which the feeling actually moves your hand.

At the bottom row you placed 80 orders instead of 100 and kept roughly a third of the money. More screen time, more commission, more decisions, less profit. And notice which half does the damage: the twenty B-grade trades cost 2R between them, while the sixty skipped A-grade setups cost 24R. The expensive emotion is disappointment, not excitement — the trades you did not take, which appear nowhere in any journal, cost twelve times more than the bad ones you did take.

That is worth pausing on, because it inverts the usual advice. “Stop overtrading” is the warning everyone hears. On these numbers, “stop under-trading after losses” is the larger problem, and it is almost never named because a skipped trade generates no entry, no fee and no regret at the time.

What actually stops it, if willpower does not?

Two guards, matched to the two mechanisms — and both of them work by moving the decision earlier in time, not by making you stronger at the moment of pressure. The teaching notes behind this course are blunt about why: by the time your hand is on the order button, the feeling has already made the decision, and suppressing it is too late. The only thing that reliably works is removing the window in which it gets to act.

Guard one, for the exits. Both exits are written on the ticket before the entry exists, and each one carries a reason rather than a number: “stop below the swing low at 41,800”, not “stop at 41,800”. Written that way, moving the stop requires you to claim the swing low moved, which is a claim about the chart that you can check and be wrong about — instead of a private feeling that price needs room. Same on the other side: “target at the prior high”, so closing at 1.875R has to be justified as a structure that is not there any more.

Guard two, for the selection. A fixed number of criteria, counted before entry, with a hard rule attached: below the count, no trade, regardless of how attractive it looks; at or above the count, the trade is taken, regardless of what the last one did. The count is the part that closes both doors at once. Disappointment cannot skip a qualifying setup because the count already qualified it, and excitement cannot add one because the count already rejected it.

The course adds a third layer for when the cycle has already got away from you, and it is worth having written down before you need it, because nobody invents good rules mid-drawdown. Cut the position size right down, so that the emotional weight of each decision drops. Set a maximum loss for one trading day, and when you hit it, stop for the day. Stop trading for a few days after a string of losses, to review them rather than to recover them. None of these three is about feeling better. They all work by shrinking how much a bad decision is allowed to cost while you are most likely to make one.

Then measure, because that is what turns this from advice into a process. Take your last twenty closed trades and compute exactly two ratios: realised average win ÷ planned average win, and realised average loss ÷ planned average loss. The first tells you whether stations 3 and 4 are active; the second, stations 1 and 2. Both are built from a number you wrote before the trade and a number the exchange recorded after it, so neither can be tidied up in hindsight. Compare their combined size against your own x from the identity above. If they sum past it, the edge is gone, and no amount of new setups will bring it back until the exits hold.

Practice corner

Pull your last twenty closed trades, and for each one write down four numbers you already have: the stop you planned, the target you planned, the price you actually left at, and whether it was a win. Two columns, twenty rows, ten minutes. Then compute the two ratios — realised average win over planned average win, realised average loss over planned average loss — and compute your own tolerance, x = expectancy ÷ (pW + 1 − p), from the same twenty trades.

Then do the part most people skip, because it is the only part that identifies the station rather than the symptom. Sort the twenty rows by what the previous trade did. If the loss ratio is worse after losing trades, you are at stations 1 and 2. If the win ratio is worse after winning trades, you are at stations 3 and 4 protecting a run. And count the setups you logged but did not take: if that count is concentrated after losses, the expensive station is 5, and everything above says it is costing you twelve times what your bad entries are.

Make the exits impossible to move quietly. The two ratios only mean something if the planned stop and target existed as real orders before the trade did. On the exchanges below you can attach a stop and a take-profit to the entry as one bracket, so the plan is resting on the book instead of in your head — and any later change shows up as an order you had to cancel, which is exactly the audit trail this exercise needs. Links below are affiliate links; the exchange pays us if you sign up, at no cost to you.

Binance→ OKX→ Bybit→

What people get wrong about the emotional cycle

• “It is six different problems.” It is two: exits and selection. Four stations move an exit, two move the next trade. Six cures is why the subject feels unfixable — you are treating symptoms that share a plumbing system.
• “Excitement and greed are the dangerous ones.” On the numbers here the expensive station is disappointment: sixty skipped A-grade setups cost 24R against 2R for the twenty bad entries that excitement added. The costly emotion is the quiet one that produces no orders.
• “Small deviations are fine.” A quarter off the winners and a quarter onto the losers is the whole edge, exactly. The deviations that kill accounts are specifically the ones too small to feel like deviations.
• “My win rate would tell me.” It moved from 39.7% to 51.0% while the system got worse. Win rate is a statement about where you put your exits, not about how good you are.
• “It is a discipline problem, so it is about me.” Partly. But your tolerance for drift is set by your system's shape — 12.5% for one, 38.5% for another at similar expectancy. Two people with identical self-control get different results from the same slip.
• “Journaling how I felt will fix it.” The journal is written by the same mind that made the decision. The two ratios are written by the exchange.

When this lesson is wrong

If you have no edge, none of this applies and tightening your exits will not save you. Every number here starts from a system with genuinely positive expectancy; the lesson is about protecting one, not manufacturing one. Executed perfectly, a losing method simply loses more reliably. If your twenty-trade sample says expectancy is negative, the identity returns a negative x, which is the arithmetic's way of telling you there is nothing here to protect. Go back to Lesson 39 and the setup itself.

The simulation genuinely disagrees with the arithmetic on one point, and we are not going to hide it. In that barrier model, pushing the stop from −1.00R to −1.25R improved the average from +0.392R to +0.471R. The reason is a property of the model: in a random walk with drift, expectancy is the drift multiplied by time spent in the market, so a wider stop keeps you in longer and therefore earns more. Real setups do not behave like that. Their edge lives in a window shortly after entry, and the target was chosen at a structural level rather than at an arbitrary distance. So the simulation is used here for one claim only — what happens to the win rate — and the expectancy claim comes from the arithmetic, which needs no model at all because it is applied directly to numbers you have already realised.

That said, the model is pointing at something true, and it has a real consequence for your stop. If your stops are genuinely too tight for the instrument's noise, widening them is correct. Crypto is volatile enough for this to be a common, honest mistake. The condition is that you widen the stop and cut the position size to match, so that 1R remains the same share of the account. Widening the stop while leaving the size alone is the failure this lesson is about; widening both together is Lesson 43 doing its job.

Not every early exit is doubt. Structure changes. A level that justified a 2.5R target can stop existing halfway there, and closing then is the plan working, not the cycle. The test is whether the closing rule existed before the trade. If it did, it is a rule; if you invented it while watching, it is a station. This is also why the two ratios are meaningless for anyone whose targets were never set by a rule — fix the target-setting first, then measure.

One limit on the figures themselves. The 40% / 1:2.5 system, the 0.40R A-grade trade and the −0.05R B-grade trade are stated assumptions, not measurements of any market. What is exact is the identity, which applies to whatever numbers your own account produces. The simulation used 200,000 paths at a fixed seed with no fees; adding costs pushes every conclusion here further in the same direction, never back.

Frequently asked questions

Is this the same thing as the market's Fear and Greed Index?

No, and it is worth keeping the two apart because they are measured on different objects. The Fear and Greed Index is a reading of the crowd — volatility, momentum, dominance and survey data blended into one number about the market. The cycle in this lesson happens inside one person, on one position, and it can run in the opposite direction to the crowd. You can be at the greed station on a trade that is working beautifully while the index says the market is fearful. Nothing in this lesson can be read off an index, because the inputs are your own planned exit and your own realised exit.

My win rate went up after I started taking profits earlier. Doesn't that mean it worked?

It means the thing you measured moved, which is not the same as the thing you care about moving. Shaving the target makes the target easier to reach, so more trades finish as wins almost by definition. In the simulation here, moving both exits pushed the win rate from 39.7% to 51.0% — an 11.3 point improvement on paper — while the money went the other way. If you want to know whether earlier exits helped, the number to compare is average R per trade before and after, not the percentage of trades that ended green.

What if my target really was too far away and cutting early was the right call?

Then it was a plan change, not an emotion, and the way to tell them apart is when the decision was made rather than how it felt. A plan change has a rule behind it that existed before the trade: close at the prior swing high, close if the daily candle closes back inside the range, close before a scheduled event. An emotional exit has a reason invented while the position is open. Write the closing rule on the ticket at entry and the ambiguity disappears — you will be able to read off, months later, which exits were rules and which were inventions.

Do I need to journal how I felt on each trade to use this?

No, and a feelings journal is usually the weaker tool, because the entry is written by the same mind that made the decision. Two ratios do the work instead: realised average win divided by planned average win, and realised average loss divided by planned average loss. Both come from numbers you wrote down before the trade and numbers the exchange recorded after it, so neither can be flattered. The feelings are worth noting once the ratios tell you which station to look at.

Sources and assumptions. The twelve emotional states and the behaviour each one produces, the list of emotion-driven habits (taking profit early with no signal, moving the stop out to give price room, entering with no confirmation, letting a winner turn into a loser), and the three recovery rules (cut the size right down, set a maximum loss for one trading day, stop trading for a few days) are from Part 5 of the free slide course. The point that emotion must be blocked before the hand reaches the button rather than suppressed at the button, and that a fixed written procedure is the only reliable block, is from the teaching notes behind that course. The break-even drift identity x = E ÷ (pW + 1 − p) is derived in the lesson and exact. The win-rate figures come from 200,000 simulated barrier paths per row, seed 20260911, on a drift-calibrated random walk with no fees; they are reported for the win-rate claim only, for the reason given in the section above. The selection-arm table assumes a +0.40R A-grade trade and a −0.05R B-grade trade. Full working: de-cuong-bai/do-so-bai-48.py. Published 11 Sep 2026.

Next: The trading journal — the six columns worth keeping, and why the two ratios in this lesson are the only ones that cannot be rewritten by the person filling them in.