Notes from the Research Desk
Essays on adaptive markets, model risk and the discipline of evidence.
Written for researchers, students and investment professionals who want to understand not only what a model says, but when it should be trusted.
These notes are written for readers who work with models rather than merely consume their outputs. Each one begins from a practical problem encountered in research or portfolio management, states the assumptions that make the problem interesting, and ends with the conditions under which the proposed answer should be reconsidered.
The series follows three recurring themes. The first is environment dependence: the recognition that correlations, liquidity and model effectiveness are not constants, and that a strategy validated in one market regime may behave differently in another. The second is validation discipline: the habit of treating a backtest as a hypothesis and asking what evidence would falsify it. The third is education: the translation of institutional research practice into forms that students and independent analysts can use.
None of the notes offers a forecast or a recommendation. Their purpose is methodological. A reader should finish each one with a clearer sense of what a model depends on, and a shorter list of reasons to trust it without question.
Reading guide
For practitioners: start with the model-risk note, then the allocation note. Together they describe the validation loop that ORION is designed to support.
For students: begin with the education note, which explains why the other two are written the way they are.
For researchers: all three notes link back to the ORION architecture and the VEA methodology.
Adaptive Allocation When Correlations Stop Behaving
Diversification is a claim about co-movement. When liquidity contracts, that claim can fail precisely when it is needed most. A framework for allocating across market environments rather than assuming one.
Backtest, Then Ask What Could Go Wrong
A backtest is a hypothesis, not a credential. Six questions that separate genuine strategy evidence from the assumptions embedded in historical data.
Teaching Finance as a Discipline of Doubt
The most useful thing a finance course can teach is not a formula but a habit: state the conditions under which your conclusion would be wrong.
Research context
The framework behind these notes
Each note draws on the ORION research environment and the VEA methodology. For the technical architecture, see the four-layer ORION framework; for the institutional context, see Verdora Excellence Alliance.