The practice
A principal-led practice for markets that are still being built.
Kinetic Alpha works on the structural questions that determine whether a new market functions: what the reference price should measure, how the contract should settle, how it should be collateralised, and who will actually use it once it lists.
Those questions get decided early, they get decided by a small number of people, and they determine whether a market works for a decade afterwards. The practice exists because a specific set of markets — compute, power, event contracts, perpetuals, tokenized collateral — are being built in public, quickly, and the structural questions they raise are not the ones the textbook answers.
The public research on this site is the working evidence. Every piece runs on primary data and ships with an interactive tool so the assumptions can be driven rather than taken on faith. The same capability is available on a commissioned basis, and the four engagement areas below each link to the published work behind them.
Principal
The practice is led by Daniel Kaufman. My background is in risk management, derivatives, and commodity-market infrastructure — clearing-house risk frameworks, energy and commodity derivatives pricing and decomposition, collateral modelling, and the margin-framework architecture that determines what capital efficiency looks like in practice.
That background shapes what the practice is good at. The energy decomposition library is the tool I spent years wishing existed. The perpetuals and event-contract work applies cleared-derivative framework thinking to underlyings that have never had it. Across every asset class it is the same three steps: read the rulebook, model the margin, understand the collateral. Most structural insight in modern derivatives sits in those three steps, and most of it never gets written down.
Practice areas
Four ways we engage — and the evidence behind each.
Index & benchmark design
New asset classes need reference prices before they can be hedged. We design and stress-test the methodology — construction, constituent selection, settlement windows, governance — and we read other people's methodologies for the assumptions they relocate rather than remove.
- ▸Methodology design and documentation: construction rules, constituent eligibility, settlement window, calculation agent responsibilities
- ▸Independent methodology review against IOSCO benchmark principles — design, data sufficiency, and disclosure
- ▸Denominator and normalization analysis: what an index's unit choice embeds, and the basis it leaves tradeable
- ▸Comparative index diligence across competing benchmarks in the same underlying
Contract & product design
A contract is the hour, the tail, the collateral, and the settlement rule — all at once. We design instruments end to end and pressure-test whether the thing a venue is about to list will actually be used by hedgers rather than only by speculators.
- ▸Full contract specification: underlying definition, settlement mechanics, funding or financing design, position and price limits
- ▸Perpetual design — funding-rate construction, cash-carry versus premium anchoring, and the no-arbitrage relationships that discipline a complex
- ▸Physical delivery and EFP mechanics: basis tables, delivery specification, convergence behaviour
- ▸Event and prediction contract architecture, including tiering, benchmark-versus-single-name liquidity, and scoring-rule sizing
- ▸Pre-listing gap analysis: what is missing from a filed complex, and what listing it would be worth
Risk, margin & collateral
How a derivative is margined determines who can hold it, at what size, and against what else. This is the practice's deepest bench — clearing-house risk frameworks, collateral mobility, and what happens to a margin model in the tail rather than the mean.
- ▸Initial and maintenance margin framework design, including portfolio and cross-commodity offsets
- ▸Clearing-house risk architecture: default waterfalls, liquidation horizons, concentration and wrong-way risk
- ▸Collateral eligibility, haircut construction, and mobility analysis — including tokenized and non-traditional collateral
- ▸Liquidation and cascade analysis: forced-flow estimation against observable depth, auto-deleveraging design
- ▸Scenario and stress design for non-storable, non-deliverable, or thinly transacted underlyings
Market structure research & diligence
Commissioned research on where a market is going and what that implies for a product decision — venue landscape, regulatory pathway, competitive positioning, and the empirical work underneath. Delivered as documents you can hand to a committee, and tools your team can drive.
- ▸Venue and competitive landscape analysis for a new or contested product category
- ▸Regulatory pathway assessment: listing routes, self-certification precedent, jurisdictional perimeter
- ▸Empirical market analysis on primary data — settlement records, order-book and index history, physical fundamentals
- ▸Build-out of interactive analytical tools for internal use, from the same engines behind the public dashboards
How we work
Four commitments that shape the output.
Primary data, not secondary summaries
Analysis runs on the source — ERCOT and PJM settlement records, exchange rulebooks, published measurement traces, CFTC and SEC filings — pulled and processed rather than cited from someone else's note. Where a figure is modelled, the model is stated and reproducible, and the script ships with the work.
Every conclusion ships with a tool
A document tells you what we concluded; an interactive artefact lets you test it. That is how a client interrogates a result instead of accepting it, and it is why the dashboards exist alongside the writing rather than as decoration.
Caveats stated before a critic finds them
Sample limits, borrowed distributions, and modelled versus observed figures are labelled in the work itself. When our own published numbers turn out to be wrong, we correct them in public and say what changed. Research that oversells its certainty is not usable for a product decision.
Compressed delivery timelines
Data ingestion, model implementation and the interactive tooling that ships with each piece are heavily automated. The cycle from question to working model runs in days rather than quarters, revisiting a finding when new information lands costs hours, and a principal-led practice can take on scope that would ordinarily require a team.
Why now
Three shifts that made this the work.
Perpetual futures came onshore. CFTC approvals, the Coinbase–Deribit pathway, spot-quoted futures and the litigation over whether funding-rate products are swaps — the regulatory building blocks for a new generation of US-listed structures, each re-opening questions about collateral, settlement and listing pathway.
Compute became a commodity in public. Five index administrators are now competing to own the benchmark on four different index technologies, with physical delivery, energy normalization and forward-curve construction all being decided at once. It is the first genuinely new major commodity in decades, it maps directly onto commodity frameworks, and it is now the practice’s deepest domain.
Tokenisation reached registered infrastructure. Programmable collateral inside a supervised perimeter changes the economics of secured financing, and the interesting questions there are haircuts, perfection and netting — which is to say, collateral questions.
Engagement shapes
How work typically gets scoped.
Diligence sprint
A focused two-to-four week question — is this contract designed correctly, what is the competitive landscape, does this methodology hold up. Delivered as a document plus the working model.
Design engagement
Full specification of an index, contract, or margin framework, from first principles through documentation a committee or regulator can read.
Ongoing advisory
Retained input on a product programme as it develops — design reviews, competitive monitoring, and analysis as filings and launches land.
The Lab
Applied AI beyond markets.
The same build approach behind the research here, applied to problems outside financial markets. These sit apart from the commercial practice — kept on the site because the method transfers.
DankeSuper
An AI-native data analysis project applying the same agent-driven build pattern behind Kinetic Alpha to problems outside financial markets.
Further applied-AI work
Data ecosystems, agent-built research workflows, and analysis pipelines in non-financial domains — surfacing here as they ship.
Open to partnerships, commissioned research, and consulting engagements.
If you are designing an index, listing a contract, building a margin framework, or trying to understand a market that does not have a settled structure yet — that is the work. Exchanges, venues, index providers, trading firms, and infrastructure builders all welcome.