Proposal: benchmark-relative Information Ratio for fixed-income fund analytics

[thayroncarlessi]

Motivation

Fixed-income fund performance should be evaluated relative to an appropriate benchmark rather than through absolute return alone.

I propose adding a benchmark-relative Information Ratio metric, initially supporting CDI for Brazilian cash-like and low-duration fixed-income funds while keeping the benchmark interface extensible to alternatives such as IRF-M, IMA-B, and credit indices.

This would also create a foundation for fixed-income analytics without assuming that CDI is appropriate for every mandate.

Definition

For periodic portfolio and benchmark returns:

• Active return: a_t = R_p,t - R_b,t
• Tracking error: annualized standard deviation of a_t
• Information Ratio: annualized mean active return divided by annualized tracking error

The implementation should align portfolio and benchmark observations to the same dates and periodicity before calculating active returns.

Optional interpretation bands

These should be configurable and documented as an operational classification rather than a universal market standard:

• Weak: < 0.30
• Acceptable: 0.30–0.50
• Good: 0.50–1.00
• Exceptional: >= 1.00

Suggested initial scope

• configurable benchmark return series;
• CDI as the first Brazilian benchmark;
• frequency and business-calendar alignment;
• consistent annualization;
• explicit handling of zero tracking error, insufficient history, missing observations, and negative Information Ratios;
• output containing active return, tracking error, Information Ratio, benchmark identifier, observation count, frequency, and annualization factor;
• deterministic unit tests;
• documentation explaining when CDI is and is not an appropriate benchmark.

A possible result contract could be:

{
  "portfolio_return_annualized": 0.128,
  "benchmark_return_annualized": 0.114,
  "active_return_annualized": 0.014,
  "tracking_error_annualized": 0.021,
  "information_ratio": 0.667,
  "classification": "good",
  "benchmark": "CDI",
  "observations": 252,
  "annualization_factor": 252
}

Benchmark suitability

CDI is appropriate for many Brazilian cash-like and low-duration strategies, but not for every fixed-income mandate. The design should therefore allow other benchmarks, including inflation-linked, nominal-duration, and credit indices, without changing the metric implementation.

Non-goals

This metric should not:

• imply investment advice;
• rank funds without disclosing benchmark, period, frequency, and fee basis;
• replace duration, credit, liquidity, drawdown, concentration, or stress analysis;
• use a static annual CDI rate in place of a properly aligned benchmark return series.

If the maintainers are interested, I would be happy to refine the scope around the preferred analytics or backtest service boundary before proposing an implementation.