Projecting readiness…
Portfolio implementation & control
Mandate reconciles holdings, models, tax lots, restrictions, and approvals into one explainable readiness state — then orders the work required to clear what blocks the next rebalance.
Account
Alder Family Tax-Managed Equity SMA
SMA-2048 · US Large-Cap Direct Index, tax-managed
Mandate status
BlockedPrimary blocker
Client restriction compliance
Proposed buy of CDRP (Cedar Point Energy Corp), 3,464 shares / $220,830, conflicts with RSTR-001: "Exclude Cedar Point Energy Corp from all sleeves of this account." (effective 2025-11-01, source Client IPS v4 (synthetic), section 6.2, exclusion schedule).
MANDATE-RESTRICT-004 · hard stop
Control status
Next action · Data Steward
Request and ingest a fresh snapshot for every stale or missing material source.
Unblocks 4 downstream controls
The question this product answers
Can this customized account safely enter its next rebalance? If not, why not, who owns the resolution, what must happen first, and which client or trading outcome is at risk?
Teams already have portfolio systems, custodians, trading tools, warehouses, spreadsheets, tickets, and policy documents. The failure happens between them. The model says what the account should own. The custodian says what it does own. Tax lots decide what can be sold and at what consequence. Restrictions decide what must not be bought. Approvals decide who may authorize an exception. Every team can see a fragment, and no one can reconstruct the complete reason an account was considered ready.
Mandate is an overlay, not a replacement system. It reconciles the minimum evidence a rebalance needs, evaluates effective-dated controls, produces one readiness projection, orders the resolution work by dependency, and preserves a trace from source data to decision to human approval.
Operating principles
One portfolio state from fragmented sources
Holdings, lots, models, mandates, and approvals each carry a source, a version, a freshness age, and a verification state. A fact with no provenance is not a fact.
Effective-dated constraints determine readiness
Deterministic controls produce explicit outcomes and expose their inputs, calculation, and dependencies. Unknown stays unknown; a contradiction stays a contradiction.
Named humans release material actions
Actions are owned and ordered by what they unblock. Language assistance is bounded. A named human approves the exact projection that enters the handoff.
The synthetic rebalance
Every step appends events to the canonical state, the same projection function re-evaluates the case, and readiness changes because the facts changed.
The account exists and the rebalance window is known, but nothing has been evaluated. Reference data is loaded; the position, tax-lot, and transaction sources for this cycle have not arrived.
Readiness after
Not evaluatedSource snapshots arrive and the case opens. Every control evaluates against recorded facts. Material blockers appear, a wash-sale match is routed to a named reviewer, and the controls that could not evaluate each name the dependency they are waiting on.
Readiness after
BlockedA refreshed custodian file arrives and the holdings difference is resolved by naming an authoritative record with a reason. Position-dependent controls now evaluate and the work queue reorders itself.
Readiness after
BlockedThe restricted buy is replaced by an in-group substitute. The impact on model deviation, cash, concentration, and projected tax is shown before a portfolio manager approves it with a reason.
Readiness after
BlockedThe completed tax-lot ledger arrives, the lot selection is confirmed, and the projected realized result becomes computable. The two wash-sale matches are reviewed and documented rather than cleared.
Readiness after
Needs authorized reviewA portfolio manager reviews the projection version, control outcomes, trades, and remaining warnings, then authorizes the release. A checksummed synthetic handoff package is generated. No order is created or routed.
Readiness after
Handoff approvedHow the decision works
Models propose trades. Controls evaluate readiness. Authorized people release the handoff.
Custodian, internal accounting, model, security master, mandate, tax lots, transactions.
Recorded facts and an append-only event log. Nothing derived is stored here.
Effective-dated controls and deterministic calculations, evaluated against a fixed clock.
Decides readiness
One projected state, with every control decision and its evidence attached.
Exceptions ordered by what they unblock, each with an owner and an SLA.
A named human releases the exact projection. The package is checksummed, never an order.
The assistant reads the projection. It cannot set a control outcome, change canonical state, approve an exception, or release a handoff — and a test asserts the projection is identical regardless of what it returns.
Implemented versus simulated
Being precise about this is part of the product, not a footnote to it.
Start at the baseline, ingest the sources, clear a stale feed, resolve a mandate conflict with an approved substitution, restore tax-lot integrity, and authorize the release. Reset returns to the identical baseline.