ETF NAV and Tracking Error — How Accurately Does It Follow the Index?
An ETF that "tracks the S&P 500" — does it really move exactly with the index? In fact it diverges bit by bit, and there is a separate metric that measures how much.
Tracking Difference vs. Tracking Error
There are two metrics for measuring how well an ETF follows its underlying index.
Tracking difference is "the ETF's return minus the index's return" over a certain period. It is usually negative, because costs make it lag the index slightly. It shows "on average, how much it lagged."
Tracking error is how much that difference fluctuated from day to day, that is, the volatility (standard deviation) of the difference, annualized. It shows the "consistency of the tracking."
Tracking difference looks at direction (how much it lagged), while tracking error looks at stability (does it follow consistently). The smaller both are, the better the ETF.
Why It Diverges from the Index
There are several reasons an ETF cannot perfectly match the index.
First, the management fee. The cost taken bit by bit each day is the biggest reason it consistently lags the index, and it mainly affects the tracking difference (direction).
Second, sampling. When only some, rather than all, of the stocks in the index are chosen and held, it moves slightly differently from the index, creating tracking error.
Third, cash drag, rebalancing costs, and the time lag in reinvesting dividends also create divergence.
Implications When Choosing an ETF
Even ETFs tracking the same index have different tracking differences and tracking errors. An ETF with low costs and precise management follows the index more closely.
For long-term investors, the tracking difference is especially important. The margin by which it lags the index slightly each year, compounded, creates a performance gap decades later. So when choosing an ETF, it is good to look not only at the total expense ratio but also at "how well it has actually followed the index" (its past tracking difference and tracking error).
That said, there is no guarantee that past tracking performance will hold the same in the future. It is safer to use it only as a reference metric and not as a basis for a definitive prediction.
よくある質問
Q. If the tracking error is near zero, is it unconditionally a good ETF?
In terms of consistency, yes. But you should look at it together with the tracking difference (on average, how much it lagged). An ETF with both values small is better than one that lags consistently (small tracking error) but by a large amount each year (large tracking difference).
Q. How do I keep from confusing the premium/discount with tracking error?
The premium/discount is "how far the market price has diverged from NAV at this very moment" (a trading-moment issue), while tracking error is "how much management performance has diverged from the index" (a long-term management issue). It is easy to remember that one is the price when you buy, and the other is the accuracy of management.
関連ページ
📋 結果は過去のデータに基づくものです。過去のリターンは将来のリターンを保証しません。
📋 本サービスは投資アドバイスではなく、投資を理解するための教育目的で提供されています。