A price-weighted technology index as a hedge and sector proxy

A reconstructible price-weighted basket

The 1982 Pacific Stock Exchange Technology Index was a price-weighted index: the cumulative value of one share of each of 100 listed and over-the-counter names. After adjustments for splits and stock dividends of at least 10% of issue value, holding an equal number of shares of each component could duplicate the basket.

Used that way, the series is an index-proxy, a reconstructible stand-in for a broader technology sector so a single position can be judged against diversified or regime-aware context.

As of 6 September 1995 the basket spanned 15 industries. Data communications was the largest industry weight at 14.4%, electronic equipment the smallest at 1.5%, and the average industry weight was 6.7%. Xerox was the largest single name at 2.99%.

Long-sample ties and yearly breakdowns

Correlation-analysis measures how closely relative returns or prices of two series move together, including how that relationship changes across calendar windows. Relative-return is annualized return above the contemporaneous average three-month Treasury bill yield, used to compare investments after a risk-free baseline.

Relative-return correlations of the technology index with the S&P 500 from March 1986 to January 1996 and with the NYSE Composite from September 1982 to January 1996 were 0.827 and 0.823. Annual correlations with the NYSE Composite ranged from 0.225 in 1995 to 0.941 in 1987.

From 1991 onward the technology index's average relative return was 19.50% per annum, compared with 7.70% for the S&P 500 and 7.62% for the NYSE Composite.

Price charts for 1995 to early 1996 show the three indices trending together over long horizons while the technology index can diverge over shorter windows, including a fourth-quarter 1995 trading range after an early-1995 advance.

How closely a small subset tracks the index

For January 1986 to January 1996, prices of the top 10 capitalized components correlated with the index at 0.991. The standard error was 5.36, or 5.7% of average index value, and the average weight was 1.97.

The top 20 correlated at 0.994. The standard error was 4.86, or 5.2% of average index value, and the average weight was 2.10. Fifteen of those 20 names were above 0.90, and Hewlett-Packard was the single highest at 0.965.

The same evaluation concluded that as few as 10 well-chosen stocks can correlate very highly with the index.

Sector hedge, residual drift, and leftover risk

The evaluation judged the technology index a closer hedge for technology-sector risk than the S&P 500 or the NYSE Composite. It also noted that prices tracked other technology indices only fairly and that relative returns tracked them poorly, with few observations on three of those series.

The evaluation also concluded that residual standard-error drift versus subindices is a candidate spread, and that long individual names judged cheap can be paired with a sector hedge. That pairing is a hedging-strategy: a rule-based use of an index or small stock subset to offset sector exposure while leaving room for stock-specific or residual drift.