Home › Explore Investing Strategies › Technical analysis
Plan before entry: reason, exit condition, entry price and stop
Four lines to write before buying, sizing from the stop. One breakout on QQQ and SPY since 2000 with only the stop changed, and how often entries filled.
In one line
Decide the reason to sell and the stop distance before you buy, and let the stop distance set the size. On QQQ a 10% trailing stop cut the max drawdown from −83.0% to −46.3% but lowered the annual return from 8.8% to 7.2%, and after 45 of its 49 stop-outs the close came back above the sale price within 20 trading days.
Article
A plan made after buying is already late. Once the price starts moving, judgment gets tied to the purchase price and the stop drifts lower bit by bit. The most practical way to reduce the loss aversion and disposition effect from the psychology article is to move the decisions to before the price moves. Writing those decisions down in four lines is the plan before entry.
Write four lines before you press buy: why you buy (the reason), what breaks the case (the exit condition), what price you buy at (the entry price), and where you get out (the stop). This article covers what the four lines mean, the research behind them and position sizing, then calculates what stops and entry prices did on QQQ and SPY. Figures are as of September 2026 and were calculated by the operator with this site's backtest engine.
What the four lines mean
- Reason: write a fact you can check. Not "it looks like it will go up" but "it cleared the prior high on volume while above the 200-day line" or "quarterly earnings grew three quarters in a row". Only a fact can be checked later.
- Exit condition: the reason turned upside down. If the reason was a breakout, it is a close back below the breakout price; if the reason was earnings growth, it is an announced earnings decline. It is not a price level but the condition under which you admit the reason was wrong.
- Entry price: not the current price but "I buy at this price", such as the first pullback after a breakout or a spot near support.
- Stop: the price at which the position goes regardless of the reason. It limits the loss before the exit condition is confirmed, and its distance from the entry price sets the position size.
Why writing it in advance helps
In a 1999 American Psychologist paper the psychologist Peter Gollwitzer called plans set in advance in the form "if X happens, I will do Y" implementation intentions. In a 2006 meta-analysis of 94 independent tests (over 8,000 participants), Gollwitzer and Paschal Sheeran found an effect size of d = 0.65, medium to large, including on disengaging from failing courses of action. In investing, the exit condition and the stop are the "if X happens".
Without them, the yardstick drifts to the purchase price. In Terrance Odean's 1998 study of 10,000 US brokerage accounts, investors sold winners about 1.5 times as readily as losers, and the winners they sold beat the losers they kept by 3.4 percentage points relative to the market over the next year (Investor psychology).
Size the position from the stop distance
The gap between the entry price and the stop is the risk per share. In the approach popularized by Van K. Tharp's Trade Your Way to Financial Freedom (first edition 1998), that amount is called 1R and trade results are counted in multiples of R. Once you decide how much one stop-out may cost, the size follows by dividing the allowed loss by the risk per share. With an allowed loss of 1% of the account and an 8% stop distance, the position is 12.5% of the account. A wider stop means a smaller position, a tighter one a larger position.
Turtle trading put volatility into this calculation. A unit was sized so that a move of N, the 20-day average range, moved the account by 1%, and the stop sat 2N from the entry price (Turtle trading). Applied to QQQ, N had a median of 1.56% of the price since 2000, and the position that holds a 2N stop-out to 1% of the account was 32.1%. It shrank sharply on volatile days: 6.0% on January 5, 2001, when N was widest (8.40%), and 7.2% on November 20, 2008 (6.95%), while in 2017, when N was small (median 0.84%), it was 59.5%.
What stops actually did (as of September 2026)
We fixed one entry rule and changed only the stop. When the close is above the prior 20-day high, the whole account buys at the next trading day's open. The only exit is the stop, filled at the stop (or at the open if the price opens below it), after which the rule waits for the next breakout. A fixed stop is measured from the purchase price, a trailing stop from the highest price since the purchase. $10,000 went in once in January 2000, with dividend-reinvested prices, a 0.15% cost per trade and no interest on cash. The same rule can be tested in the backtest tool with the custom price-breakout rule (20-day entry), the strategy exit signal turned off, and a stop or trailing stop set.
| Stop | QQQ annual return | QQQ max drawdown | QQQ stop-outs (back within 20 days) | SPY annual return | SPY max drawdown | SPY stop-outs (back within 20 days) |
|---|---|---|---|---|---|---|
| None | 8.8% | −83.0% | 0 | 8.2% | −55.2% | 0 |
| Fixed 8% | 11.0% | −61.6% | 10 (8) | 7.2% | −54.4% | 8 (7) |
| Fixed 15% | 10.5% | −71.0% | 6 (6) | 6.6% | −60.4% | 5 (5) |
| Trailing 10% | 7.2% | −46.3% | 49 (45) | 7.3% | −43.2% | 28 (26) |
| Trailing 20% | 9.2% | −66.0% | 16 (16) | 5.9% | −56.7% | 11 (10) |
The number in parentheses counts stop-outs after which the close came back above the sale price within 20 trading days. With no stop, the rule buys the first breakout and holds to the end, so it is almost the same as buy and hold (buy and hold returned 8.6% a year on QQQ and 8.2% on SPY).
A fixed stop measures only from the purchase price. All 10 of QQQ's 8% stop-outs came in 2000 to 2002. During the 83% decline the rule cut each loss at about 8%, and after buying again on October 16, 2002, the price never fell below 92% of that purchase price, so the position was held to the end. The 11.0% a year came from that one stretch. SPY's 8% stop stayed tied to a March 2003 purchase price, so it took almost all of the 2008 decline, sold on March 2, 2009, a week before the March 9 bottom, and bought back on March 24 at a price 14.3% higher. SPY's 15% stop added the 2008 decline to the losses from stop-outs and re-entries in 2000 to 2002 and fell to −60.4%, deeper than buy and hold.
A trailing stop keeps working after the price rises. A 10% trailing stop cut the max drawdown from −83.0% to −46.3% on QQQ and from −55.2% to −43.2% on SPY, and in exchange lowered the annual return from 8.8% to 7.2% on QQQ and from 8.2% to 7.3% on SPY. After a stop-out, the close came back above the sale price within 20 trading days in 45 of 49 cases on QQQ and 26 of 28 on SPY. On QQQ the next breakout bought back at a median 5.1% above the sale price, and above it in 33 of 49 cases. On the other side, the close fell more than 10% below the sale price within 60 trading days 17 times on QQQ and 9 times on SPY. Avoiding those few large declines cost dozens of buybacks, and the QQQ 10% trailing rule paid $2,934 in costs.
A stop does not guarantee the maximum loss. In 7 of the 49 QQQ trailing-stop exits, the price opened below the stop and the sale filled at the open. The largest loss among QQQ trades with the fixed 8% stop was −8.7%, 1.08 times 1R.
In a 2014 Journal of Financial Markets paper, Kathryn Kaminski and Andrew Lo showed that when prices follow a random walk, stop-loss rules lower expected returns, and that they can add returns only when trends persist (momentum). With monthly US stock data from 1950 to 2004, rules that switched into long-term government bonds after a stop added 0.5 to 1.0 percentage points a month during stopped-out periods. That matches the result above, where the fixed stop helped only in a long decline like 2000 to 2002.
Entry price: when the planned price never comes
Since 2000, QQQ closed above its prior 20-day high for the first time in a run on 459 days (through May 2026, so that 60 trading days can be seen afterward). A limit order 2% below that close, left open for 10 trading days, would have filled 46.2% of the time. When it filled, the return from the fill to 60 trading days after the signal averaged +2.6%; when it did not, buying at the next open would have averaged +3.5%. The plan to buy 2% cheaper saved money every time it filled, but missed the moves that went straight up without a pullback. Per signal, the limit plan was worth +1.2% and buying at the next open +2.2%. On SPY the fill rate was 32.9%, with +2.5% when filled and +2.7% when missed, a small gap, but per signal the limit plan was worth +0.8% and buying at the open +2.0%.
A missed opportunity does not show up as a loss in the account, but it has a cost. When you set an entry price, put into the plan that you do not buy if the price never comes, and decide in advance whether you accept that cost.
When the template holds and when it does not
- It holds when the reason is a checkable fact: a reason you cannot check, such as "a good company", gives you no exit condition.
- It fails when the stop is tighter than normal swings: QQQ's 2N had a median of 3.1% since 2000, and a tighter stop gets hit by everyday swings. Set the stop wider than the swings and cut the position to match.
- Indexes and individual stocks differ: the results above are the record of index ETFs that did recover. William O'Neil's 7 to 8% stop was meant for growth stocks (O'Neil's cup-with-handle breakout), and on a stock that never recovers, the loss a stop prevents can be far larger.
- Without a re-entry rule, a stop is half a plan: the calculation above bought back at the next breakout. A stop with no condition for buying back means sitting in cash through whatever rise follows.
Related concepts and search terms
- R-multiple, expectancy: counting each trade's result in multiples of the initial risk (1R), and computing a rule's expectancy from its win rate and average R.
- Fixed fractional position sizing: putting only a set share of the account (for example 1%) at risk on each trade and sizing from it, as this article does.
- ATR stop, chandelier exit: setting the stop distance as a multiple of the average true range (ATR). Measured down from the highest price, it becomes a trailing stop.
- Implementation intention, pre-mortem: a plan of the form "if this happens, I do that", and a method that assumes failure first and writes down the reasons.
Search terms: trading plan, stop loss backtest, trailing stop, position sizing 1% rule, R-multiple, Van Tharp, ATR stop, Kelly criterion, implementation intentions, trading journal.
Drawbacks and risks
- Stops do not hold through gaps. If the price opens below the stop, the loss is larger than 1R.
- Stops and re-entries add trades. The QQQ 10% trailing stop filled 100 times since 2000, and for Korean residents overseas stock gains above ₩2.5 million a year are taxed at 22% (including local income tax).
- Sizing from the stop distance makes the position larger as the stop gets tighter. At 1% risk, a 1% stop means the whole account, so a separate cap on position size is needed.
Operator's assessment
A stop protects against losses and costs money at the same time. With the same 20-day breakout entry on QQQ and SPY and only the stop changed, a 10% trailing stop cut QQQ's max drawdown from −83.0% to −46.3% but lowered its annual return from 8.8% to 7.2%, and after 45 of its 49 stop-outs the close came back above the sale price within 20 trading days. A fixed stop measures only from the purchase price, so it did nothing once the price had risen, and its excess return on QQQ came from one stretch, 2000 to 2002. So we first set the stop distance wider than normal swings, then cut the position to match. At 1% risk that is 12.5% for an 8% stop and 32.1% for QQQ's usual 2N (3.1%). The entry price follows the same logic: a plan to buy 2% cheaper went unfilled more than half the time on QQQ, and the misses were the bigger gainers. Confluence Zone's presets and alerts let the tool check two of the four lines, the reason and the exit condition; the entry price, stop and size have to be set for your own account.
What this article does not cover
This article does not give entry or stop prices for any stock. It covers only the four-line template and the calculations; you set the values from your own reasoning. The backtests apply one entry rule to two index ETFs, QQQ and SPY (dividends reinvested, 0.15% cost per trade, no interest on cash, taxes and currency left out). Stops can work out differently on individual stocks that never recover. Everything here is reference information based on past, public data and is not investment advice.
Articles and questions on the same topic
Drawdowns and investor psychology
- Strategy articles: The math of loss and recovery: a halving needs a doubling, Averaging down and pyramiding: which way to add
- Backtest questions: What would investing in QQQ (Nasdaq-100) every month have returned?, Doubling monthly contributions when the market is 20% below its high: did it beat the same amount every month?, Stopping or continuing monthly investing during a crash: how different were the results?
- Investor profiles: Contrarian Bargain-Buying Strategy (John Templeton), Contrarian Low-P/E Strategy (David Dreman)
Asset allocation and rebalancing
- Strategy articles: Asset allocation: cash 30, stocks 70, bonds and rebalancing, All Weather and the Permanent Portfolio: four-season allocations tested with ETFs
- Backtest questions: How much less does a 70% stock, 30% cash mix fall than 100% stocks?, HFEA (55% UPRO, 45% TMF, rebalanced quarterly): did it still beat SPY after 2022?, TQQQ, QLD and QQQ cross-rebalancing: how does it compare with simply holding?
Short-term rule trading (grid, volatility breakout)
- Strategy articles: Grid trading: buy at each line down, sell one line up, Volatility breakout: buy at open + k × yesterday's range
- Backtest questions: Grid trading on QQQ and TQQQ: did it beat simply holding?, Grid trading with lines 3%, 5% or 10% apart: which spacing did best?, Larry Williams volatility breakout on QQQ: did it beat simply holding?
Related
Turtle trading · O'Neil's 7 to 8% stop · Investor psychology · Turtle breakout backtest · Support, resistance and breakout · Presets and alerts in the analyzer
Basics and trading strategies
← Dual momentum (GEM): switching between stocks and bonds on 12-month returns · Contents · Reading indicators: moving averages, MACD, RSI and volume →
See these metrics on a real stock →
About & methodology · Explore Investing Strategies · Glossary · Questions and answers · Rule status · Sample analysis · Pricing · Terms · Investment disclosure · Privacy policy · Open the analyzer · 한국어 · 日本語 · 繁體中文 · 简体中文