Momentum-Index Constituents API for Systematic Research

Retrieve the maintained ten-stock momentum basket and its ranks for the actual rebalance date instead of reconstructing history with revised inputs.

A momentum-index constituents API should return the basket that was recorded at the rebalance, not a portfolio rebuilt later from today's universe or corrected data. Alphanume's Quant Galore Momentum Index dataset serves the maintained monthly ten-stock basket with the rebalance date, ticker, and within-basket momentum rank. Rank one is the strongest included name.

The Momentum Index guide describes the constituent workflow. The feed is useful for basket history, turnover, and post-rebalance research. Historical compounding is not an expected return, and the ten-name construction creates concentration and turnover risks that should remain visible.

Pull the recorded rebalance basket

Use date for a selected rebalance or date bounds for a sequence. An exact basket should contain ten distinct tickers with ranks one through ten. Preserve the served date as the observation key. Do not substitute an arbitrary month-end or first-day label.

The small schema is intentional. It answers what the basket held and how names ranked, not every intermediate feature used in selection. If the research question is about the maintained index, the recorded constituents are the primary input. If the question is about an independently replicated factor, document that it is a different portfolio.

Retrieve a date and check invariants

A single-date request is the cleanest way to reproduce one basket. For a historical sequence, request a range and group by date. The API accepts exact or logically valid date bounds. Save the raw response with the research run so a later strategy change does not alter the constituent history.

GET /v1/quant-galore-momentum-index
date=<selected rebalance date>

assert row_count == 10
assert unique(ticker).count == 10
assert sort(rank) == [1,2,3,4,5,6,7,8,9,10]

weights_equal[ticker] = 0.10
next_basket = query next recorded rebalance date
turnover = compare current ticker set with next ticker set

The dashboard documentation states that constituents update monthly at 4:05 PM America/New_York on rebalance. A simulation must trade no earlier than the chosen availability rule. If the research assumes same-close execution, that assumption needs market data and a realistic implementation model rather than being implied by the constituent record.

Measure turnover and concentration explicitly

Ten stocks can produce large issuer and sector concentration. Report equal-weight contribution, largest position impact, and any repeated sector exposure. Calculate additions, removals, and retained names at each rebalance. Even if nominal weights reset to equal, price movement between rebalances can create drift that depends on the chosen portfolio accounting.

• Keep constituent return attribution at the ticker level.
• Apply transaction costs to buys, sells, and weight resets.
• Handle delistings and corporate actions without dropping failed names.
• Compare recorded baskets only after their stated availability time.

Rank is relative within the selected basket. It is not a calibrated forecast of return magnitude. A rank-one name can lose money, and small differences in rank do not imply proportional differences in expected performance.

A multi-rebalance pull also needs basket-date validation. Group every row by served date, verify that each group has exactly ten names and complete ranks, then sort the dates rather than assuming a fixed number of calendar days between them. Store the next recorded rebalance date as the holding-period boundary. This protects the test from month-length and holiday assumptions and makes any unexpected constituent gap visible before return calculations.

Avoid revised-input reconstruction

The main failure mode is rebuilding an old basket from a current ticker universe, revised fundamentals, or a different price vendor, then assuming the output is the maintained index. That reconstruction can be a valid independent factor test, but it is not the historical constituent record. Label and compare the two series separately.

Another failure is reporting the basket's backtest without turnover, costs, or concentration. A ten-name monthly portfolio can be sensitive to a single constituent. Publish gross and cost-adjusted results, event-level contributions, missing-price treatment, and the exact trade-timing convention.

Benchmark choice should also be prospective. Compare the basket with a broad equity benchmark, a momentum benchmark, or both, but define them before reviewing the strongest period. If the maintained basket is used as a signal source rather than a portfolio, document the additional selection rule and distinguish its results from direct basket replication. Otherwise the API record and the tested strategy become impossible to separate.

Reproduce one basket transition

As a concrete next action, query one selected rebalance date and validate its ten unique tickers plus ranks one through ten. Save the exact request, raw response, returned date, row order, and validation output. Then attach point-in-time prices under a written post-publication rule without calculating a multi-rebalance turnover panel. That single-date contract audit keeps this API guide distinct from the historical constituent workflow.