Bringing financial judgment and technical craft to a question that matters: what can market history actually tell us?
Graham founded Chart Library to make empirical market research more accessible, more inspectable, and more useful. He leads the project's research direction and the development of the systems that turn its questions into working tools.
Graham's background in finance and institutional fixed income shapes the questions behind Chart Library: what is comparable, what evidence matters, and how much uncertainty remains after the analysis is done?
He holds a master's degree and a bachelor's degree in finance from Florida Atlantic University. With Chart Library, he brings that financial grounding into the design of empirical tests, data systems, and tools that readers can examine for themselves.
The project begins with a simple premise: access to useful research should not depend on an institutional subscription.
Graham's work spans the full path from a research question to a public result: defining comparable market states, developing historical retrieval and calibration systems, and making the evidence available through a website, API, and agent-accessible tools.
That combination matters. A method described on paper must still work on imperfect data. An attractive interface must still represent the evidence faithfully. A published number needs a source, a date, and a clear account of what it does—and does not—establish.
The result is Chart Library: an independent research project with a working market memory, a public study ledger, daily observations, and an inspectable coverage record.
The research is organized around specific questions and decision criteria. When a test does not support an idea, the negative finding remains useful: it narrows the next question and prevents an appealing explanation from becoming an unsupported claim.
Recent studies illustrate that distinction. Similar historical charts can help describe the range of outcomes without establishing which direction prices will move. Adding more context or more elaborate weighting does not necessarily improve the comparison.
This is an ongoing project, with methods that can be tested and limitations that can be challenged. The methodology describes the approach; the research ledger shows where it has—and has not—held up.
Tighter neighborhoods are genuinely more similar in what followed: the dispersion ratio falls monotonically as k shrinks, at every horizon. The tails narrow; the interquartile range barely moves. The state definition carries real information, but about how much, not which way.
One multiplier, ×1.48, fitted on the earlier half lifts held-out gap-day coverage from 0.70 to 0.83. The under-coverage lives in non-earnings gaps and in up-gaps; earnings gaps were already covered by the earnings conditioner, so the new one belongs on gap days that are not earnings sessions.
The event-banded set is wider, not narrower, and no better at equal coverage. The finding underneath mattered more: the served analog set for an event day is the state's twins, not the event's, and the served band under-covers event days at 5 days. That under-coverage is what study 103 repaired.