Skip to content
Canli Capital

Canli Capital

Arhan Canli

I build and publish a systematic trading engine. Everything on this site that carries my name as author is my own work, and all of it is written so that somebody who does not know me can check it without asking me anything.

What the work is

ALPHAC is a cross-asset book of paper-traded sleeves — crypto funding carry, US equity cross-sectional momentum, multi-asset time-series momentum, and a point-in-time consumer-price-surprise size spread — combined at equal core weights with a separately disclosed directional overlay. The engine behind it handles the whole path: point-in-time data, factor construction, covariance-based sizing, a volatility target and drawdown ladder, execution against a live order book, and publication into a signed record. That path is walked stage by stage on the systems page, with the measurement that proves each stage linked from it.

The record it produces is deliberately larger than the part that worked. 82 research documents are published, organised into 13 subject hubs, alongside 28 measurements each rendered with its own claim boundary. 46 candidates have been killed and published in full against 3 that survived. 162 distinct hypothesis identities have been consumed against a declared budget of 320, because a deflated result is only honest if the number of attempts behind it is counted in public.

The live paper record began 2026-08-07 and has accrued 15 days. That is short, it is stated as short everywhere it appears, and no conclusion is drawn from it that a record of that length cannot support.

What is deliberately not on this page

There are no degrees here, no former employers, no awards and no assets under management. That is not modesty and it is not an oversight — it is the same rule the rest of the site follows. A credential is a claim you would have to take on trust, and the entire argument of this record is that you should not have to take anything on trust. If a line of biography would change how you read the numbers, it is doing work the numbers should be doing themselves.

So the only personal claim made on this page is one you can check with a command, and it is in the next section.

The one claim here you can verify

I have committed USD 1,000 of my own capital at first live (real-money) deployment of this book, into the same book, at the same time, on the same terms as any external capital. That commitment is Ed25519-signed, so it cannot be quietly edited, softened or backdated after the fact.

It is a small number and saying so is the point: it is what I can actually commit, stated exactly, rather than a figure chosen to impress. No real capital is deployed today. This is a forward commitment that activates at first live deployment, and until then it is a promise with a signature on it and nothing more.

curl -sO https://canlicapital.com/glassbox/founder_commitment.json
curl -sO https://canlicapital.com/glassbox/reproduce.py
pip install cryptography
python3 reproduce.py --dir .

The same key signs the 406-entry append-only chain behind the track record, which has been running since 2026-06-27. The full verification instructions cover both, including what each check cannot prove.

How the work is done, and where it has been wrong

The rules I hold myself to are written down and enforced in code rather than kept as intentions: a hypothesis is registered before its data is opened, a threshold is fixed before it is measured, a gate is mutation-tested by breaking the thing it guards, and a result that is chosen after seeing which population clears it is called selection and published as such.

The most useful thing I can offer a sceptical reader is not the wins. It is that the corrections are published in the same place and the same format as everything else — a withdrawn figure, a defect found in our own gate, an earlier answer superseded by a better measurement of the same thing. The measurements include several that exist only because something we had already published turned out to be wrong.

Elsewhere

Code: https://github.com/arhancanli. The research library is at /research, the published measurements at /measurements, and the instructions for checking any of it at /verify.

Nothing on this site is investment advice, an offer, or a solicitation. The book trades on paper and no real capital is deployed.