Tracking Error, Information Ratio & Active Share

Definition

Tracking error measures how far a fund's returns stray from its benchmark's returns, period by period. Take the difference between the fund's return and the benchmark's return in each period — the active return — and compute the standard deviation of those differences:

Active return (each period)  a  = fund return − benchmark return

Tracking error               TE = standard deviation of the active returns

  a  = how far the fund landed from its benchmark in one period
  TE = how widely those gaps scatter, annualized like any other
       standard deviation (monthly figure × √12)

A tracking error of 0% would mean the fund matched its benchmark exactly in every period. A tracking error of 5% means the fund's yearly result typically lands within a few percentage points of the benchmark — sometimes ahead, sometimes behind. Like any standard deviation, it measures spread, not direction: a fund that beats its index erratically and one that trails it erratically can post the same tracking error.

One sibling term is worth separating immediately. Tracking difference is the *average* gap between fund and benchmark — the mean, not the scatter. For an index fund it is usually a small, steady negative number, and its biggest ingredient is simply the fund's fee: a fund charging 0.10% per year will, all else equal, trail its index by about 0.10% per year. That is the expense ratio showing up in the return stream. Tracking error tells you how *noisy* the gap is; tracking difference tells you how *big* it is on average.

Two companion metrics complete the picture. The information ratio (covered below) divides average active return by tracking error to ask whether the deviation paid off. Active share approaches the same question from the holdings side: what percentage of the portfolio differs from the index at all.

Why It Matters

Tracking error means opposite things depending on what kind of fund you own — and reading it with the wrong lens is one of the most common analytical mistakes income investors make.

For an index fund, small tracking error is a quality grade. An S&P 500 tracker like VOO has exactly one job: deliver the index's return, minus as little as possible. A tracking error of a few hundredths of a percent says the fund is executing well; a large one signals sloppy replication, illiquid holdings, or operational drag. Here, less is simply better, and the companion number to check is tracking difference, which should sit close to the fund's expense ratio and no worse.

For an active fund, tracking error is the size of the bet — not a flaw. An actively managed dividend strategy *must* deviate from the index to have any chance of beating it; zero tracking error would mean zero active management. The question is never "why isn't this zero?" but "am I being paid for this?" — which is what the information ratio measures.

For covered-call and option-income funds, large tracking error is structural. A fund like SPYI holds S&P 500 stocks but sells call options on top, deliberately trading away upside for income. Against the raw S&P 500, its return stream *cannot* track closely — it lags in rallies and cushions in selloffs by design. Measuring such a fund's tracking error against the plain index tells you the overlay is working, not that the fund is broken. The fair yardstick is a buy-write index, which bakes the same call-selling mechanics into the benchmark itself.

The practical payoff: tracking error sorts the fund universe into honest categories. Near zero means true indexing; moderate with a full fee attached may mean closet indexing; large means a genuine active bet or a structural strategy — and the follow-up question becomes whether the results justify it.

Information Ratio

The information ratio (IR) connects the two halves of the story: it divides the fund's average active return by its tracking error.

Information ratio  IR = active return ÷ tracking error

  active return  = fund return − benchmark return, averaged per year
  tracking error = standard deviation of those same differences
  IR             = benchmark-beating return earned per unit of deviation risk

Think of it as "alpha per unit of activeness." Two funds can each beat their index by 1% a year, but if one did it with a 2% tracking error and the other needed 8%, the first delivered its edge four times more efficiently — a distinction the raw outperformance number hides completely.

Worked example (illustrative). Suppose an active dividend fund beats its benchmark by an average of 1.2% per year, with a tracking error of 4%:

IR = 1.2% ÷ 4% = 0.30

An IR of 0.30 is respectable for a real-world active fund: year-to-year results scatter within a few points of the benchmark, but the average lands 1.2% ahead — a modest, persistent edge relative to the size of the bets. A negative IR means the deviation is *costing* you: the fund is different from the index and worse.

The IR is a cousin of the Sharpe ratio: Sharpe measures return per unit of *total* volatility against a risk-free rate, while IR measures return per unit of *benchmark-relative* volatility against an index.

Active Share

Tracking error and IR are returns-based: they look at results after the fact. Active share is the holdings-based cousin. It measures the percentage of a portfolio that differs from its benchmark index — 0% means the fund holds exactly the index, 100% means not a single overlapping position at the same weight.

Active share is the classic detector for closet indexing: charging active management fees on a portfolio that is mostly just the index. A fund with 20% active share and a 0.85% expense ratio is, in effect, charging a wildly inflated fee on the 80% that is plain index exposure you could own for a few basis points. Low active share plus a high expense ratio is paying active prices for index exposure — the one combination with no redeeming story.

One nuance matters enormously for income investors: active share only sees the stocks. A covered-call fund that holds the full S&P 500 but writes calls on top can show a *low* active share while posting a *huge* tracking error, because the options overlay reshapes the return stream without changing the stock list. When the holdings-based and returns-based measures disagree like this, the strategy is living in the overlay, not the stock picks.

Example

Here is how the three metrics line up across four fund profiles an income investor might compare. All numbers are illustrative, measured against the plain S&P 500:

Reading each row:

• The index fund's 0.05% tracking error is a quality certificate. The IR is not meaningful — the fund is not trying to beat anything. If its average lag roughly equals its expense ratio (say, −0.03% for a 0.03% fee), it is delivering exactly what it promises.
• The closet indexer deviates a little (2%) and loses by doing so: with an average active return of −0.8% per year, its IR is −0.8% ÷ 2% = −0.40. The 25% active share confirms most of the portfolio is just the index, and the fee explains most of the shortfall.
• The genuine active fund — think of a rules-based dividend strategy in the spirit of SCHD, whose concentrated dividend portfolio differs sharply from the broad market — runs a real 5% tracking error and an 85% active share. Averaging +1.5% of active return per year gives it an IR of 1.5% ÷ 5% = +0.30: the deviation is being paid for. Its loose fit to the index is exactly what a lower R-squared would show.
• The covered-call fund looks disastrous against the raw index in a strong bull year: if the index gains 25% and the fund gains 15%, the active return is −10%, and −10% ÷ 8% gives an IR of −1.25. But that is a benchmark error, not a fund failure. A covered-call ETF sells its upside for income on purpose; scored against a buy-write index that does the same, its tracking error shrinks dramatically and its IR becomes a fair test of execution.

Common Mistakes

• Judging every fund by the index-fund lens. Small tracking error is a virtue only for funds that promise to track. Demanding it from an active or option-income fund is asking the fund to stop doing the thing you bought it for.
• Benchmarking a covered-call fund against the raw index. Its tracking error will be structurally huge and its IR will look terrible in every rally. Use a buy-write index as the yardstick, or you will "discover" the same misleading shortfall every year.
• Confusing tracking error with tracking difference. One is the scatter of the gaps, the other is their average. An index fund can have a tiny tracking error and still lag its index every single year by its fee — that steady lag lives in tracking difference.
• Celebrating outperformance without checking the IR. Beating the index by 2% with a 12% tracking error (IR ≈ 0.17) is a coin-flip-grade result; beating it by 1.2% with a 4% tracking error (IR = 0.30) is a far stronger signal of skill.
• Missing closet indexers. Moderate tracking error, low active share, and a full active fee is the pattern to hunt for — never pay twenty times the index-fund rate for what is mostly index exposure.
• Trusting active share alone for overlay strategies. Options-based funds can match the index's holdings and still behave nothing like it. Read the holdings-based and returns-based measures together, and treat any disagreement as information.

FAQ

What is a good tracking error?

It depends entirely on the fund's job. For an index fund, lower is better — well-run large-cap trackers commonly show a few hundredths to a few tenths of a percent, and anything much beyond that deserves an explanation. For actively managed funds, tracking error is a design choice, not a defect: benchmark-aware strategies often run 2–6%, and concentrated ones higher still. For covered-call funds measured against the raw index, large tracking error is structural and says nothing about quality. There is no universal "good" number — only a number that does or does not match what the fund claims to be.

What is a good information ratio?

For a real-world active fund, an IR around 0.3 sustained over a full market cycle is solid, roughly 0.5 is very good, and long-run figures near 1.0 are rare. Beware of high IRs computed over short windows — one lucky year can produce a spectacular number that means little. Also confirm the benchmark is fair: an option-income fund scored against a buy-write index can post a sensible IR while looking hopeless against the raw index. A persistently negative IR means the fund's deviations are systematically costing you money relative to just holding the benchmark.

Why does my covered-call ETF trail the index?

Because it is supposed to — in up markets. A covered-call fund sells call options on its holdings, collecting premium income today in exchange for giving up gains above the strike price. In a strong rally the index runs ahead while the fund's upside is capped, so a large gap versus the raw index is the strategy working as designed. The honest comparisons are against a buy-write benchmark that sells calls the same way, and on after-distribution total return rather than price alone. If the fund trails a *buy-write* index persistently, that is a genuine execution problem worth investigating.

What is the difference between tracking error and tracking difference?

Tracking difference is the average gap between a fund's return and its benchmark's — for index funds, usually a small negative number dominated by the expense ratio. Tracking error is the standard deviation of the period-by-period gaps — how much they bounce around, regardless of direction. A fund can have near-zero tracking error while reliably lagging by its fee every year, or a small average lag that arrives erratically. Index-fund due diligence should check both: tracking difference for cost, tracking error for execution quality.

What active share counts as closet indexing?

A common rule of thumb from the academic research that coined the term: active share below roughly 60% suggests a fund is substantially hugging its benchmark, while genuinely active stock-pickers typically run 80% and above. The number matters most alongside the fee — a 30% active share fund charging index-fund prices is merely conservative, but the same portfolio charging 0.8–1.0% is charging active prices for mostly passive exposure. And remember the overlay caveat: options-based funds can have low active share while being highly active in their return behavior, so active share alone cannot classify them.

How do tracking error and R-squared relate?

They describe the same relationship from opposite ends. R-squared measures how much of a fund's movement is *explained* by its benchmark — an index fund sits near 100. Tracking error measures the size of the movement the benchmark does *not* explain, so the two move inversely: near-zero tracking error implies an R-squared near

1. Neither says whether the deviation helped; for that you need alpha

for the size of the edge and the information ratio for its efficiency.