Reconstitution Algorithm
Install and import#
npm install fintech-algorithmsimport { calculate } from "fintech-algorithms/index-and-benchmark-engineering/governance-and-maintenance/reconstitution-algorithm";Signature#
calculate(data)Chooses the membership at a scheduled review, using an incumbent buffer so a name hovering at the boundary is not swapped in and out every period. The buffer is what separates an index from a churn machine.
Parameters#
| Name | Type | Notes |
|---|---|---|
data | { candidates: Candidate[]; targetCount: number; incumbentBuffer: number } | incumbentBuffer lets an existing member stay while it ranks within the buffer beyond the cutoff — asymmetric on purpose, because turnover costs the tracking funds real money. |
Returns#
{ selectedIds, additions, deletions, targetCount }
The new membership with the explicit adds and drops it implies.
Errors#
- When targetCount exceeds the candidate count — throws
Complexity: time O(n log n),
space O(n).
Worked example#
verified This is the worked example published in the article, replayed by the test suite on every run. The output cannot drift.
Input#
{
"candidates": [
{
"id": "A",
"score": 98,
"incumbent": true
},
{
"id": "B",
"score": 94,
"incumbent": true
},
{
"id": "C",
"score": 90,
"incumbent": false
}
],
"targetCount": 5,
"incumbentBuffer": 1
}Call#
calculate(data)Returns#
object with 4 fields: selectedIds, additions, deletions, targetCount
{
"selectedIds": ["A", "B", "C", "D", "F"],
"additions": ["C"],
"deletions": [],
"targetCount": 5
}Diagrams#
Calculation flow#
Reconstitution Algorithm calculation flow
flowchart LR
A["Point-in-time inputs"] --> B["Validate units and timing"]
B --> C{"Contract feasible?"}
C -->|No| D["Reject with reason"]
C -->|Yes| E["Calculate Reconstitution Algorithm"]
E --> F["Recompute invariants"]
F --> G{"Checks pass?"}
G -->|No| D
G -->|Yes| H["Publish audited output"]
Reconstitution Algorithm methodology state
stateDiagram-v2
[*] --> FrozenInputs
FrozenInputs --> Validated: contract passes
FrozenInputs --> Rejected: missing or infeasible
Validated --> Calculated: apply named rule
Calculated --> Audited: invariants pass
Calculated --> Rejected: invariant fails
Audited --> Published: version and timestamp recorded
Published --> Revised: approved correction
Revised --> FrozenInputs: rebuild from retained source state
How it works#
This page states the contract — how to call it correctly. The article explains the concept: why it works, and where it breaks.
References#
- S&P Dow Jones Indices Index Mathematics Methodology — S&P Dow Jones Indices
- S&P DJI Equity Indices Policies & Practices — S&P Dow Jones Indices
- FTSE Russell Capping Methodology — FTSE Russell, LSEG
- FTSE Russell Index Policy and Methodology Library — FTSE Russell, LSEG
- MSCI Global Investable Market Indexes Methodology Library — MSCI
- MSCI Minimum Volatility Indexes Methodology — MSCI
- S&P Risk Control 2.0 Indices Methodology — S&P Dow Jones Indices
- Principles for Financial Benchmarks — International Organization of Securities Commissions
- Regulation (EU) 2016/1011 — European Union
- Portfolio Selection — Harry Markowitz
- On the Properties of Equally-Weighted Risk Contributions Portfolios — Sébastien Maillard, Thierry Roncalli, and Jérôme Teïletche
- Fundamental Indexation — Robert Arnott, Jason Hsu, and Philip Moore
- FTSE Currency Hedging Methodology Overview — FTSE Russell, LSEG