Home/Methodology

Methodology

Validation comes before presentation.

The most common weakness of the consulting industry is the unvalidated claim. Tradist Capital brings academic peer-review practice to commercial delivery: every output must be traceable to its source and independently reproducible.

Five principles.

These are not marketing copy; they are working standards included in our contract annexes.

We do not predict; we estimate. The difference between the two is measurable uncertainty. Our Working Principle

Research first, implementation second

No model is coded before the literature review and a comparative analysis of existing approaches are complete.

Reproducibility

Every analysis is documented so that a third party can reproduce it — including data sources, parameter sets and random seeds.

An estimate is not a claim of certainty

All outputs are reported with confidence intervals and error metrics. A presentation that hides uncertainty is treated as a methodological defect.

Scope discipline

No component outside the contracted scope is added to a delivery without written agreement. In engineering, surprise is not a virtue.

Regulatory awareness

Capital-markets regulation, data-protection law and related frameworks are considered at the start of the design, not at the end.

Delivery Standards

What every delivery contains.

  • Assumption log — every assumption the model rests on, with its rationale, in writing
  • Data lineage — which data, from which source, through which transformations
  • Uncertainty report — not point estimates; confidence intervals and error metrics
  • Limitations section — what the model cannot do, and under which conditions it breaks
  • Reproduction protocol — a step-by-step protocol for a third party to reach the same results

Let's start with your institution's question.

The first meeting is for jointly clarifying which of our practice lines your need falls into. It creates no obligation.

Request a Meeting Send an Email