08 · Price Alerts and Off-Screen Discipline
Price alerts are the cheapest — and most underrated — discipline tool in any charting platform: you set a price condition, and the system calls you back when it triggers. Their purpose is not "catching the move" but outsourcing the "condition → action" trigger to the system, redeeming your attention from the screen. Watching charts costs more than time — it erodes the very judgment you need at the decisive moment.
💡 One-Sentence Summary
One-sentence summary: Alerts replace watching, not deciding. An alert only brings you back to the screen; the grounds for action always come from the trading plan.
1. The Cost of Watching Charts
"Watching the screen means being responsible" is a beginner's instinct, yet what actually happens is usually the opposite:
| Cost | Mechanism |
|---|---|
| Attention drain | Constant scrolling makes the brain process meaningless fluctuations, dulling it for the signals that truly matter |
| Emotional triggers | Every refresh is a temptation: even without reaching a condition level, you itch to "do something" |
| Forced noise intake | Short-timeframe watching shows mostly random movement; constant exposure to high noise inevitably breeds an urge to act |
| De facto loss of position control | Staring at unrealized P&L makes emotion ride the numbers, tempting early profit-taking or panic stops that break the plan |
The habit of watching 1m / 5m charts all day is especially harmful: you are not "tracking the market" — you are force-fed the highest proportion of noise. The finer you look, the easier it is to read random movement as signal and act outside the plan. This resonates with human weakness: the loss aversion and recency effects covered in the behavioral finance chapter are amplified severalfold in the watching scenario.
💡 Attention Is Part of Position Sizing
Decision quality depends on attention quality. Spending limited attention on the key moments of plan execution — rather than on every minute's fluctuation — is itself a form of risk control.
2. The Right Role for Alerts: Outsourcing "Condition→Action," Not the Decision
A trading plan (see the trading plan chapter) decomposes a trade into judgeable conditions: "when price reaches X, do Y." What an alert does is refreshingly simple: the system watches the condition for you and calls you back when it triggers.
The crucial boundary:
- Alerts replace watching: you need not poll the market with your eyes; the system neither misses nor tires nor slips.
- Alerts do not replace decisions: a triggered alert ≠ automatic execution. "Price broke the alert level" means one condition is met; whether to enter, how much, and where the stop goes — all of that still comes from the plan.
One sentence draws the line: "An alert is an alarm clock, not an auto-trader." The alarm means "come back and handle this," not "buy now."
3. How to Set Alerts That Actually Work
3.1 Bind Alerts to Plan Key Levels, Not Round Numbers
The most common mistake is "hanging alerts on round numbers by feel": 30000, 50000, and other psychological levels. Such levels are neither your entry condition nor your stop basis; when one triggers you have no playbook — you have wired noise straight into your notification bar.
The right approach: every alert maps to a key level in the trading plan —
| Alert type | Which plan item it binds to | What to do when it fires |
|---|---|---|
| Breakout confirmation level | The entry rule's "break X with a confirmed close" | Recheck the plan: are all confirmation conditions met? Enter only if yes |
| Stop warning level | The buffer one step before the stop rule (e.g., warn at −0.7% for a −1% stop) | Rehearse the exit: check whether the plan calls for reducing or exiting — do not improvise a new stop |
| Add-position trigger | The position-management rule "add near pullback to Y" | Verify add-on preconditions (trend intact, total exposure within limits) before acting |
| Structure-level approach alert | 0.5%~1% before a major higher-timeframe support/resistance | Enter observation mode and wait for the confirmation signal |
Set this way, every triggered alert has a playbook — the direct benefit of sourcing alerts from the plan.
3.2 Prefer Conditional Alerts Over Single Prices
Most platforms support conditional alerts ("price crosses above," "drops more than N%," "recovers above the average price"). When a condition can be written, write it — an alert phrased like the plan itself reduces the "translation cost" at trigger time.
4. Alert Overload: Too Many Alerts Equal No Alerts
The fatal flaw of the alert mechanism is overload: too-frequent triggers → the brain files them as background noise → the one alert that matters gets missed. This is the same disease as the "confirmation bias" in multi-timeframe analysis: the more information, the easier to hear only what you want to hear.
Suggested caps:
- ≤ 5 active alerts per instrument; ≤ 10 across the whole portfolio.
- Every alert must answer two questions: "Which item of the plan does it map to?" and "What will I do when it fires?" If either answer is missing, delete it.
- Clean up regularly (e.g., during the weekly review — see the trading journal and checklists): alerts unrelated to the current plan, pure round-number alerts, and alerts that fired last week without any action.
🔄 Keep Only Alerts Sourced from the Plan
The sole test of whether an alert deserves to exist: "Does it come from the current trading plan?" Alerts hung outside the plan — however convenient — open a back door for impulsive trading.
5. After the Alert Fires: Triggered ≠ Enter
The standard post-trigger routine is a fixed three steps:
1. Stop: do not order immediately. The alert only says "the price condition
is met," not "the trade is on."
2. Check: return to the trading plan and verify every premise attached to
this trigger — direction allowed? Volume confirmed? Total exposure within
limits? Is today even a planned execution day?
3. Act: all conditions met → execute per the plan (including stop and size);
any condition missing → record "triggered but skipped" with the reason,
and close the case.The "triggered but skipped" record matters enormously — it is as valuable as a winning trade: it is your evidence for distinguishing "the alert system works" from "the condition design works." If reviews show a flood of skipped triggers, your alert conditions are looser than your plan and need tightening.
The reverse trap deserves equal warning: treating alerts as signals and entering on every trigger. That compresses the plan's entry conditions into a single price condition, skipping every confirmation step in the instant you press the button — the alert becomes an accelerator for impulsive trades instead of a discipline tool.
6. Using Desktop Notifications and Sound as Tools
The delivery channel itself affects discipline:
- Tiered delivery: stop warnings — which demand immediate action — get sound and strong notifications; structure-approach alerts — which merely invite observation — get silent push. Strong notifications everywhere recreates overload.
- Kill market-feed push: charting apps' default "big move" and "trade tape" notifications are unrelated to your plan — turn them all off. They are the front door for noise.
- A post-trigger ritual: when a strong alert fires, the first thing you open at the screen is the trading plan page — not the chart page. That fixed order is itself discipline training.
💡 The Tool-Chain Loop
Alerts (standing watch) + the trading plan (decision basis) + the trading journal (behavior record) form a loop: the alert brings you back, the plan tells you what to do, the journal records what you actually did.
⚠️ Risk Warning
Price alerts can be delayed or missed, and in extreme moves price can blow through both the alert level and the stop level in an instant; a triggered alert is not a trade signal, and executing without checking the plan invites losses. No alert mechanism can replace standing stop-loss orders and position management.