In plain words — The stop does not state how much you want to lose. It marks the point below which the technical reason for the purchase no longer holds.
Where does the level come from?
In Chapter 4 Hurst buys after an upside break of a valid downtrend line, near a zone where several components are expected to reach lows. Chapter 5 returns to the channel of the shortest component visible on the daily chart and examines its lower boundary.
The reference is the lowest low reached before the VTL break. If price violates it after entry, Hurst reads the move as evidence that the expected cluster of lows did not complete or that the decomposition was wrong. That low becomes the initial cut-loss.
The historical procedure, step by step
First, the short component is identified and the two boundaries of its channel are sketched. The falling VTL leading into the low zone is then drawn. Entry may occur only after its upside break; at that moment, the preceding low is fixed as invalidation.
The level is not lowered to give the trade “more room.” If reached, the position is closed and the situation is reassessed. Any later entry needs a new signal. It does not follow from hoping the first scenario will become valid again.
In the Gruen case the book reports a purchase around 7½ and a cut-loss around 7⅛. Hurst estimates roughly a 5% price move and adds the commission costs of that period, reaching total risk near 9%. These are figures from one case and the 1961 cost structure, not a standard percentage to copy today.
Model level versus real order
“Stop” can mean two different things: the point at which the analysis is invalidated and the instruction sent to a broker. When triggered, an ordinary stop order generally becomes a market order. The stop price is therefore a trigger, not a guaranteed execution price.
During a downward gap or very fast market, the first available price may be substantially lower. A stop-limit order controls a price limit but may not execute at all. Brokers may also use different trades, quotations, or rules to determine activation. These operating features are not contained in the cyclic drawing and must be assessed separately.
How to make it testable today
A testable rule must state which series identifies the component, how the low is chosen, when the VTL is valid, and what data were available at that time. It must include gaps, slippage, commissions, unfilled signals, and every trade rather than only successful examples.
The logical cut-loss offers consistency: entry and invalidation come from the same hypothesis. Its limitation is that geometry may be ambiguous and monetary risk changes with the distance between entry and low. Position sizing therefore remains a separate decision.
Evidence boundary — Predefined invalidation disciplines a decision, but it neither caps the loss with certainty nor proves a cyclic edge. During gaps or poor liquidity, the execution price may be worse.
Sources
- J. M. Hurst, The Profit Magic of Stock Transaction Timing, Prentice-Hall, 1970, Ch. 5, pp. 86–88, Fig. V-1 — logical cut-loss construction and the Gruen case.
- U.S. Securities and Exchange Commission, Stop, Stop-Limit, and Trailing Stop Orders: Investor Bulletin, updated 2026 — triggers, price risk, and differences between stop and stop-limit orders.
Related entries
- Profit and exit management — complete framework
- Trailing and profit-taking — protection after the initial stop
- Valid trend line — entry and exit condition