Insider’s Note · 4 May 2026
What Evrydiki actually does
The most direct description we can write of what we do, what we don’t, and why it matters. Liquid futures, a portfolio of systematic models, fractional-Kelly sizing under fat tails, and a strict line between methodology (open) and signals (private).
Evrydiki Research · 7 min read · Introduction · Methodology · Boutique
Most boutique systematic firms either over-promise ("uncorrelated alpha, every quarter") or under-explain ("proprietary, can't say more"). We try not to do either. This page is the most direct description of Evrydiki we can write — what we do, what we don't, and why we think it matters.
What we are
Evrydiki is a Cyprus-based boutique systematic trading firm.
- Boutique — a small number of qualified investors at any given time. Capacity is finite by design.
- Systematic — every trade entered, sized, and exited under a written rule. No discretionary overrides during market hours.
- Trading firm — not an advisory practice, not an academic research outfit. The work product is real positions in real markets, with real P&L.
What we trade
Liquid futures. Specifically: equity index, fixed income, FX, and selected commodity contracts where execution costs are low enough that systematic strategies remain viable after slippage and financing.
We don't trade equities, options, or crypto in the live portfolio — not because they're bad markets, but because our edges are sized to futures liquidity and the operational story stays simpler.
How we trade — the multi-model picture
We run a portfolio of models, not a single strategy. Each model is a self-contained system with its own:
- Signal logic
- Position sizing
- Drawdown budget
- Exit conditions
Models cover different time horizons (intraday to weekly), different stylistic premia (momentum, mean reversion, term structure), and different markets. The portfolio's resilience comes from diversification across models, not from any single signal being exceptional.
If a model fails its own internal stop conditions, it's retired and documented publicly. We do not let dead strategies haunt the portfolio.
How we size — fractional Kelly under fat tails
This is where most boutique systematic firms quietly get it wrong.
The textbook Kelly criterion assumes returns are normally distributed. They aren't — Mandelbrot showed this in the 1960s and the empirical record has only gotten clearer since. Under fat-tailed returns (L-stable distributions), full-Kelly sizing is fragile. Even half-Kelly understates drawdown when tail clusters arrive.
We size positions under a heavy-tailed correction to the Kelly formula, calibrated against a range of empirically plausible tail indices for each market we trade. The discipline isn't to outguess the next move — it's to size such that no plausible tail breaks the firm.
The full derivation is in our working paper: Mandelbrot–Kelly: from theory to position sizing.
Two intellectual influences
Our work is shaped by two traditions, neither of which is original to us. We just take both seriously at the same time.
- Benoît Mandelbrot — markets are fractal, prone to extreme moves, and resistant to averaging. The right response is sizing, not prediction.
- Ray Dalio — every recurring decision should be encoded as an explicit principle, examined when it fails, and revised in the open. Discretion is reserved for the few decisions that cannot yet be written down.
We hold both. The first protects us from over-confidence in any single model. The second protects us from cumulative drift in our own decision making.
What we publish
The thinking — methodology notes, working papers on sizing and risk construction, regime commentary, and post-mortems on retired models.
You can read most of this work without ever speaking to us. That's on purpose. If our methodology cannot survive open scrutiny, our returns won't either.
What we keep private
The signals — live model parameters, current positions, entry and exit triggers, intra-day allocation. Anything that would erode the edge if disseminated.
The boundary is not arbitrary. The framework is examinable; the trades are not.
How we engage
By selection. By appointment. By process.
- By selection — we engage a small number of qualified investors at any given time. Capacity is finite.
- By appointment — conversations begin only after a fit on principles, ticket size, and jurisdiction. There are no cold calls and no public sales pitches.
- By process — every step, from research to onboarding to monthly reporting, runs on a written process. If we cannot describe a step, we don't take it.
If you're a qualified investor and the work above resonates, the Request Access form is the door. We respond within two business days, either way.
What this is not
A few things Evrydiki is deliberately not:
- Not an "uncorrelated alpha guaranteed" pitch. Drawdowns happen; we budget for them ex-ante.
- Not a black-box hedge fund. The methodology is publishable precisely because the value isn't in obscurity.
- Not an aggressive growth story. We are intentionally small and intend to stay so. The strategy capacity is limited; we'd rather serve a few investors well than many investors badly.
If you read this and the differences from typical systematic marketing read as boring, you're our audience.
This is an Insider's Note. For the full archive of methodology notes and working papers, see research.