P34 Technical Report · June 2026

Computable
Markets

Learning When Not to Trade: A Generalized Profit-As-Regression Decision Model for Biased, Partially Observed Business Markets

P34 is a decision model, not a price forecaster. It learns profitable portfolio selection from biased historical business data where only prior accepted trades have reliable outcomes.

Read the Report Get in Touch

P34 · Difficult market

+$14.8k

Realized profit under declining margin and hidden bias

Baseline · Same market

−$219.8k

Tuned GBDT-R predicted +$126.6k but lost it all

P34 · Stationary market

+$228.6k

Matches optimistic baseline in favorable conditions

Calibration

71.7%

Prediction calibration in the difficult benchmark

01 / The model class

Profit-As-Regression

PARML optimizes against final economic outcomes — not intermediate forecasts. The target is realized portfolio profit, not predicted price. The model internalizes execution risk, adverse selection, inventory loss, and margin compression by learning from what actually happened.

02 / The problem

Biased Observation

Businesses never observe the complete market. They only see outcomes of trades they accepted. Historical data is shaped by a prior policy the model cannot see. Standard regressors trained on these labels over-enter trades when conditions shift.

03 / The architecture

Universe Selection

P34 constructs multiple plausible interpretations of the market, filters dominated universes, corrects for false-positive risk, and selects portfolios using meta-level telemetry. No-trade is always an available action.

Where P34
Applies

01 Wholesale and retail trading
02 SME and micro-lending
03 Freight and trucking brokerage
04 Crypto and prediction markets
05 GPU capacity allocation
06 Construction materials hedging
07 Electricity and energy markets
08 Liquidation and resale markets

The
Architecture

A business provides data tables, action schemas, and constraints. P34 returns an executable portfolio: selected trades and volumes.

Decision pipeline

  • 01Menu grounding and safe label expansion
  • 02Market and business regime detection
  • 03Base profit prediction with false-positive correction
  • 04Universe construction and Pareto filtering
  • 05Drift extrapolation under regime hypotheses
  • 06Universe selection via meta-telemetry
  • 07Portfolio execution with boundary-aware sizing

The interface

The business menu: a structured set of candidate actions a business can actually take.

Request Technical Report

Why
PARML

A business whose decisions are encoded in spreadsheets, manual judgment, and brittle rules is difficult to scale, audit, and transfer. A business whose decision process is represented by a retrainable profit-directed model becomes more stable, more observable, and more adaptive.

P34 is not a forecaster attached to a rule engine. It is an operational asset that converts business complexity into repeatable decision quality.

What current systems cannot do