# GO / NO-GO decision

GO / NO-GO decision is a pre-trade verdict indicating whether a proposed stock trade passes an analysis check, fails it, or needs changes under a separate MODIFY verdict.

## What it means

A pre-trade check can reduce a complicated assessment to three labels: GO, NO-GO or MODIFY. GO means the proposed trade or basket passes that check as submitted. NO-GO means the check identifies a reason not to proceed as submitted. MODIFY means changes to the proposal should be assessed before a new decision is made.

These labels describe the outcome of a particular analysis, not what a stock is worth or where its price will go. A GO is not an instruction to place an order; a NO-GO is not a prediction that the stock will fall. The proposed quantity and execution conditions matter because changing them changes the proposal being assessed.

## How it is calculated

There is no universal formula for a GO / NO-GO decision. A pre-trade system applies its own inputs, rules and model assessments. For a simple, **hypothetical illustration only**, a rule could be written as `verdict = GO if P >= 50% and R = 0; MODIFY if P >= 50% and R = 1; otherwise NO-GO`, where:

- `P` is a hypothetical model-estimated success probability, measured as a percentage.
- `R` is a hypothetical, unitless change flag: `0` means no change was flagged, and `1` means a change was flagged.
- `50%` is an illustrative cutoff, not a general market standard or PreTrAIde’s full decision rule.

A real verdict can involve more than one threshold. The label alone does not explain which inputs drove it, so the accompanying risk assessment and reasoning are important.

## Worked example

Suppose a private investor constructs a hypothetical basket of two US stocks. Line A is a proposed purchase of 100 shares at US$20 per share, or `100 × US$20 = US$2,000`. Line B is a proposed purchase of 50 shares at US$40, or `50 × US$40 = US$2,000`. The basket’s stated share value is US$4,000, excluding trading costs.

1. In the illustrative rule above, assume the basket receives `P = 60%` and `R = 1` because its proposed execution conditions need review. The result is MODIFY, not GO.
2. Suppose the investor changes those conditions and runs a fresh hypothetical check. If the new inputs produce `P = 60%` and `R = 0`, that rule returns GO. The changed proposal, not the original one, passed.
3. If the check instead produces `P = 45%`, the illustrative rule returns NO-GO regardless of `R`.

These invented inputs demonstrate how labels work; they are not actual PreTrAIde outputs or estimates of investment returns.

## Why it matters before you trade

A verdict puts a decision point between planning and order entry. It can prompt a review of position size, liquidity, volatility, estimated execution costs and the conditions under which an order might fill. For a basket, it also helps distinguish a concern affecting one line from a concern affecting the basket as a whole.

Read the explanation rather than treating the label as a command. A model assessment may depend on data quality and assumptions, while actual fills and subsequent prices remain uncertain. The investor remains responsible for deciding whether to act and for placing any order through their broker.

## How PreTrAIde uses it

PreTrAIde performs **pre-trade analysis** on a user-built stock basket before any order is placed. Each stock has a buy or sell side, quantity and execution conditions. The analysis returns GO, NO-GO or MODIFY for the basket and each line, alongside a success probability, risk assessment, suggested execution strategy and plain-language reasoning.

Its success probability is a model estimate between 5% and 95%, not a guarantee or a return forecast. Quantitative scores, market data, news and analyst ratings inform the analysis. Guardrails can move a verdict only in a more conservative direction: a suspended stock is always NO-GO, and a GO with a success probability below 50% is downgraded.

For a MODIFY verdict, PreTrAIde offers “Let AI Modify” to rework the basket within defined limits. PreTrAIde is analysis software, not a broker or investment adviser; users place any trades themselves.

## Common mistakes

- **Reading GO as a recommendation or guaranteed fill.** It is a decision-support verdict, not an order or a promise.
- **Reading NO-GO as a price forecast.** It describes the analyzed proposal, not the stock’s future return.
- **Treating MODIFY as GO.** A proposed change should be evaluated as a changed proposal.
- **Ignoring the level of the verdict.** A line-level result and a basket-level result answer different questions.
- **Confusing success probability with expected return.** A percentage estimate is not a projected gain in US$ or percent.

---
Source: https://pretraide.com/glossary/go-no-go-decision