What it means
In PreTrAIde, MODIFY flags a proposed stock basket for changes before any order is placed. It is one of three possible verdicts—GO, NO-GO and MODIFY—returned for both the basket and each stock line. A line has a side (buy or sell), a quantity and execution conditions.
The verdict is decision support, not an instruction or investment recommendation. It does not mean an order has been changed, submitted or executed. A basket-level MODIFY can call attention to the proposed basket as a whole; the line-level verdicts let the user inspect individual entries.
How it is calculated
MODIFY is a categorical verdict, not a number with a standalone formula. PreTrAIde does not publish a fixed threshold or equation that converts its inputs into MODIFY. The following expression describes the output, not a calculation rule:
Basket verdict ∈ {GO, NO-GO, MODIFY}
Here, basket verdict means the outcome for the basket as a whole; GO, NO-GO and MODIFY are the three possible labels. Each line also receives a verdict. PreTrAIde supplies a separate success probability between 5% and 95%, a risk assessment, a suggested execution strategy and plain-language reasoning. The probability is a model estimate, not a forecast of returns or a guarantee.
Its pre-trade analysis combines deterministic quantitative estimates with AI reasoning anchored on quant scores, news and analyst ratings. Quantitative inputs include estimated spread, participation in daily volume, market impact, daily volatility, 1-day 95% value at risk, liquidity and concentration. Guardrails can correct an AI output that contradicts the data and move a verdict only in a more conservative direction; for example, a suspended stock is always NO-GO.
Worked example
Suppose a hypothetical basket contains four buy lines of 25 shares each. For illustration, assume each stock is priced at US$100, making each line US$2,500 and the basket US$10,000. Each line therefore represents 25% of the basket by value.
- The user enters the four lines and their execution conditions, then runs one pre-trade analysis.
- Suppose the returned basket verdict is MODIFY. The user reviews the basket- and line-level verdicts, risk assessments, success probabilities, suggested execution strategies and reasoning. These hypothetical inputs do not imply that four equal-weight lines automatically receive MODIFY.
- If the user chooses Let AI Modify, an illustrative revised basket might retain three original lines at US$2,500 each and replace the fourth with one new US$2,500 line. Three of four original lines remain; each position is 25%, below the feature’s 40% maximum. This is an example of a possible rework, not a promised result.
The revised basket is a proposal. The example assumes unchanged prices solely to keep the arithmetic simple; prices and quantities can affect actual weights.
Why it matters before you trade
A MODIFY verdict distinguishes a proposed trade that may merit adjustment from a simple GO or NO-GO label. It gives the user a point to examine line-level risks and execution conditions before placing orders with a broker. For example, a large proposed quantity relative to daily volume can matter for estimated market impact, while a heavily weighted line can affect concentration.
Neither the verdict nor an altered basket removes market risk or establishes that a later trade will succeed. The user remains responsible for deciding what, if anything, to execute.
How PreTrAIde uses it
PreTrAIde runs its PreTrade analysis before an order is placed. Each analysis uses one analysis credit. When the verdict is MODIFY, Let AI Modify can rework the basket automatically once per analysis without another credit. It can make up to five AI passes to trim, exit or replace lines, add at most two new lines, retain at least half of the user’s original lines and keep every single position at or below 40% of the basket.
PreTrAIde is analysis software, not a broker, investment adviser or portfolio manager. It never receives, transmits or executes an order; users place any trades themselves through their own broker.
Common mistakes
- Treating MODIFY as an automatic trade. Auto-modify reworks a proposed basket; it does not place an order.
- Assuming MODIFY means every line must change. Basket and line verdicts are reported separately, and auto-modify keeps at least half the original lines.
- Reading the success probability as an expected return. It is a bounded model estimate, not a return forecast.
- Assuming the example is a scoring rule. Equal weights and round prices make the arithmetic visible; they do not determine the verdict.