Vanguard investment grade bond fund tracking error

Tracking error is one means of measuring active manager performance against an appropriate market index. Vanguard's investment grade bond funds are actively managed, defined asset class funds. The funds include three bond funds structured across three bond maturity levels: short, intermediate, and long term bonds. In addition to these three portfolios, Vanguard manages a high yield corporate bond fund and three corporate bond index ETFs. The corporate bond market is a very large market, with many illiquid securities. The following table shows the number of securities held by the funds and bond benchmark indices.

In May 2004, Vanguard instituted a change in the funds' fundamental investment policy, allowing the funds to invest in up to 20% of net assets in investment grade securities other than corporate bonds (this includes treasury and government agency debt issues). Prior to 2004, the funds invested in corporate bonds.

Tracking error
Table 2. provides long term average tracking errors for Vanguard's investment grade bond funds. Data extends from 1996 to present for the short, intermediate, and long term investment grade bond fund investor shares. Admiral share class data covers the period from 2002 (the first full year of operations of the share class) to the present. Detailed annual tracking error data, along with additional statistical measurement of fund returns, are provided in the footnote tables.

The relative performance of investor and admiral share classes can be best illustrated by viewing performance over a common holding period. The following table examines the average tracking error performance of each fund over the period beginning with the first full year of admiral share performance to the present. The lower expense admiral shares provide for lower tracking errors to benchmark performance.

Over the measurement period (prior to 2000 Vanguard benchmarked each fund to the Aggregate bond index), the short and intermediate term funds have produced an average tracking error measuring about 20 basis points higher than their expense ratios. Both funds have held portfolios that have lower average durations than the benchmark durations. This fact is reflected in the funds having betas lower than the benchmark. The long term fund has produced positive average tracking error over the measurement period. The fund's average duration has closely tracked the benchmark duration over the period (see the duration tab in the footnote tables for annual data). The funds also hold portfolios that do not precisely match the credit quality distribution of the benchmark index.