Skip to main content
Track Index proxy comparison
15 / 19
Library

2014issue C0359-61

Headline index levels mix a changing basket with a changing divisor

A price-weighted 30-name index cannot be treated as the same object across decades because both the member set and the divisor change. Editorial reading: winner-in, loser-out reconstitution can make a later high look like a healthier regime when the series is only a changing index-proxy.

  • A price-weighted 30-name index is not a like-for-like object across decades because both the member set and the divisor change.
  • Scaling can dominate the headline number, as when a formula change lifted the published level overnight or a later divisor turned the same dollar move into more points.
  • Membership changes systematically swap names in stabilization or degeneration for names in takeoff or acceleration.
  • A later high does not by itself prove that the same economic or social regime is healthier than an earlier low.
Entries in this reading2 entries

An index-proxy is not a fixed object

A price-weighted 30-name index cannot be treated as the same object across decades because both the member set and the scaling divisor change, so later point totals are not a like-for-like comparison with earlier totals. An index-proxy is a headline basket used as if it were a continuous market thermometer, even though its members and scaling change over time.

The divisor changes what one point means

The divisor is the scaling factor that keeps an index continuous after splits and substitutions, and that therefore changes how many points a one-dollar move in the basket produces. On 1 October 1928 the basket was enlarged to 30 names and the divisor was set at 16.67 instead of 30, so a share-price sum of 3984 produced 239 rather than 132.8, an overnight 80 percent lift in the published level from the formula change alone.

By 2014 the divisor had fallen to about 0.1557 from a value above 1 in 1985, so a one-dollar move in the 2014 basket added roughly 6.4 times as many index points as the same dollar move under the 1985 formula. A published level near 16437 under the 2014 divisor would have been about 2559 if the 1985 divisor of 1 had still been used, showing how scaling, not only prices, can dominate the headline number.

Membership follows transition-phase, not a frozen roster

Membership changes systematically swap names in stabilization or degeneration for names in takeoff or acceleration, which raises the chance the published series trends up relative to a frozen roster of the same original firms. Transition-phase is a life-cycle stage of a constituent or industry, used here to describe why weak names leave a headline index and strong names enter it.

From 1980 onward the 30-name set absorbed seven information, communications, and technology firms plus five financial firms, concentrating the proxy in the industries then treated as the engines of that transition. Documented 2004-2013 substitutions removed names after large multi-year declines or after they were judged no longer representative, and added names after large multi-year gains or after industry-mix arguments, which is winner-in, loser-out maintenance rather than a fixed portfolio.

A later high is not a regime proof

Market-regime-classification sorts conditions into phases such as takeoff, acceleration, stabilization, or degeneration so a single reading is judged against the membership and formula then in force. Because an index point is computed differently at different dates, a later high does not by itself prove that the same economic or social regime is healthier than an earlier low. The series is a changing proxy, not a constant regime thermometer.

Educational research material, not investment advice. Historical source context does not establish present-day performance.
15 of 19 in the Index proxy comparison track
201744-45 pp.Next on Index proxy comparisonScreening ETFs by liquidity, index fit, and rankActive traders used average-volume lists as a liquidity filter, hunting for volume and narrow bid-ask spreads before choosing an executable ETF.
All readings on this track · 19 readings
  1. 1992Country regime inside global allocation and index proxies
  2. 1992Intermarket confirmation for long-duration bond-fund timing
  3. 1993Paired bond and currency proxies with weekly crossover confirmation
  4. 1995Walk-forward evaluation of a municipal futures timed fund switch
  5. 1999Regime-gated allocation with bounded index leverage
  6. 1999Testing trend following with cash-price controls
  7. 2002A capital-preservation case for index-proxy allocation
  8. 2003A shared weekly-average grid for four Asian index proxies
  9. 2005Index-etf-core weights, a growth-index-clock, and an implementation-cost-ledger
  10. 2005European index proxies as one weekly-regime panel
  11. 2006Index-fund proxies as intermarket regime instruments
  12. 2006Constructing metal option exposure with mining proxies and implied volatility
  13. 2010Matched straddles on levered versus unlevered index proxies
  14. 2013Inheritance as an index-proxy and allocation case
  15. 2014Headline index levels mix a changing basket with a changing divisor
  16. 2017Screening ETFs by liquidity, index fit, and rank
  17. 2019Leveraged commodity proxies fail the futures test
  18. 2020Constructing pre-listing paths for new fund sleeves
  19. 2020A sleeve after cost-drag, judged by an index-proxy, sized in a stock-bond mix
All 28 readings tagged Index proxy comparison
Also on Index proxy comparison5 readings