Research
What we look for, and why it might exist.
Every family has a stated reason to exist, a counterparty who pays for it, and a condition under which it is retired.
Information domains
Each domain is an independent view of the market. All of them clear the same measurement before they are allowed to matter, and slow-moving context tilts conviction without firing anything.
Price and volume
What traded, where, and how much.
Measured
Microstructure proxies
Who is trading, inferred from the shape of bars.
Measured
Cross-asset and macro
The state of related markets, rates and volatility.
Measured
Calendar and time structure
Cycles, windows and the clock.
Measured
Institutional flow and positioning
Where large holders are leaning.
Planned
Derivatives and options
What the options market is pricing in.
Planned
News and narrative
What is being said, and how fast it spreads.
Planned
Fundamentals
Earnings, releases and revisions.
Planned
Temporal and seasonal
Patterns tied to the clock and the calendar.
Same time of dayParticipants with daily schedules, such as funds rebalancing and orders sliced across the day, act at similar times and leave a repeating footprint.CounterpartyTraders whose timing is set by a schedule, not by price.Retired ifThe pattern fails on later data, or cost per session consumes it.Overnight versus intradayDifferent participants are active in each window, and information arrives at different times.CounterpartyHolders who cannot trade overnight, and those who must trade the open.Retired ifThe split carries no information once the market is removed.Calendar and period-end positionRebalancing, payroll flows and reporting dates concentrate flows into predictable windows.CounterpartyInvestors whose flows are fixed by the calendar.Retired ifIt is among the most published of effects, so it is retired if it does not survive being counted among every other look.Cross-sectional
One name measured against its peers.
Short-term reversalDemand for immediate liquidity pushes price past fair value, and it recovers as liquidity returns.CounterpartyThose who need immediate execution and pay for it.Retired ifReversal is explained by a tilt in the market, or disappears after cost.Residual and industry momentumInformation about a single firm spreads slowly once the common moves are stripped out.CounterpartySlow-reacting holders of the firm.Retired ifThere is no persistence on later data.Lead and lag
Information that reaches one market before another.
Fund to constituentA fund and its holdings are priced in different places by different participants, and the faster one leads.CounterpartyThose pricing the slower side from stale information.Retired ifThe delay shrinks below what cost allows.Cross-asset and cross-sector transferAttention and capital are limited, so related markets do not reprice at the same moment.CounterpartyParticipants who watch only one market.Retired ifThe lead disappears when measured at prices that could be traded.Flows and events
Trading that happens for reasons other than price.
Period-end markingA reporting date creates an incentive to show certain holdings or levels.CounterpartyManagers acting for a reporting date, not for price.Retired ifThe effect does not survive the count of looks.Tax-loss sellingTax rules create a deadline that forces sales.CounterpartyTaxable holders working to a deadline.Retired ifThe pressure cannot be told apart from ordinary reversal.Options expiryDealers hedging large positions buy and sell in ways that dampen price, then free it.CounterpartyDealers hedging a book, not expressing a view.Retired ifHeavily optioned names and a control group behave alike.Index reconstitutionIndex funds trade on a known date, in known size.CounterpartyFunds that must buy or sell on a schedule.Retired ifA clean point-in-time record of the changes cannot be had, or the effect is gone.Price structure and anchoring
Where price has been, and why it matters again.
Price anchoringParticipants anchor on reference prices and under-react when price approaches them.CounterpartyAnchored participants.Retired ifThe sign or size does not hold on later data.Volume profile and structurePrices accepted by many participants become reference points for later decisions.CounterpartyTraders who ignore where the crowd has been.Retired ifThe levels add nothing once simple price measures are accounted for.Volume and order flow
Who is trading, and how urgently.
Participation and order flowInformed or urgent trading leaves traces in volume and in the shape of bars.CounterpartyPassive providers of liquidity.Retired ifThe proxy taken from bars is too noisy to survive cost.Microstructure proxiesSpread and impact show up in the high, low and close of a bar.CounterpartyThose who trade thin names without accounting for what the trade costs.Retired ifThe estimates do not match the cost actually paid.Discovery
Letting a governed search propose what to measure.
Generated familiesDiscovery without preconceptions, where every expansion of the grammar is counted as a look.CounterpartyNone is assumed in advance. Anything found must name one before it earns weight.Retired ifNothing found survives the counted selection.