Scheme Performance — judging the fund manager fairly

The question this chapter answers

A scheme returned 14%. Was that good?

You cannot say without something to compare it against. That comparator is the benchmark, and this chapter is about choosing one honestly and then measuring the manager against it.

What makes a benchmark credible

Two requirements:

It must be in sync with three things — the scheme's investment objective (the securities in the benchmark should represent the portfolio implied by the objective), its asset allocation pattern, and its investment strategy.

It must be calculated by an independent agency, in a transparent manner, and published regularly. Most are constructed by stock exchanges, credit rating agencies, securities research houses or financial publications.

For an index fund the choice is simple — the investment objective names the index it tracks, and that index is the benchmark. For every other scheme the choice is subjective, and is made by the AMC in consultation with the trustees.

The SID must state the benchmark, and the benchmark's performance over the same period must accompany the scheme's past performance wherever that is shown.

A benchmark can be changed — because the investment objective changed, the index construction changed, or a better index became available. The AMC does so in consultation with the trustees, and the change must be justified and documented.

Price Return versus Total Return — the honest comparison

Schemes were once benchmarked to the Price Return variant (PRI), which captures only the capital gains of the index constituents.

With effect from 1 February 2018, schemes are benchmarked to the Total Return variant (TRI), which also counts all dividends and interest generated by the constituents.

Why it had to change: scheme performance is calculated after adding the dividends the scheme earned — and even for the IDCW option, performance is computed by the reinvestment method. Comparing that against a price-only index gives a slight advantage to the scheme.

The consequence is worth stating carefully, because it is the exam's favourite framing:

The gap between PRI and TRI returns is the dividend. Historically, Indian index-level dividend yields have ranged between 1.5% and 2.5% most of the time. So roughly two percentage points of apparent outperformance simply vanished — outperformance that was never real.

Where TRI is not available for a period, funds use a composite CAGR of the PRI benchmark (until TRI becomes available) and the TRI thereafter.

Choosing benchmarks by scheme type

Equity schemes — four aspects of the objective drive the choice:

• Scheme type. Diversified funds use broad indices — Sensex, Nifty 50, S&P BSE 200, S&P BSE 500, Nifty 100, Nifty 500. Sectoral and thematic funds use sectoral indices — S&P BSE Bankex, S&P BSE FMCG, Nifty Infrastructure, Nifty Energy
• Investment universe. Sensex (30 companies) and Nifty 50 (50 companies) suit large-cap funds; midcap funds use Nifty Midcap 50 or S&P BSE Midcap
• Portfolio concentration. Funds holding few stocks suit narrow indices like Sensex and Nifty 50; funds holding many suit broader ones — Nifty 100, Nifty 200, Nifty 500
• Underlying exposure

Debt schemes — benchmarks must be developed by research and rating agencies recommended by AMFI. CRISIL, ICICI Securities and NSE have developed such indices.

Worth knowing by name: ICICI Securities' Sovereign Bond Index (I-Bex), with three maturity sub-indices — Si-Bex (1 to 3 years), Mi-Bex (3 to 7 years), Li-Bex (more than 7 years).

Liquid schemes invest in securities up to 91 days, so a short-term money market benchmark such as the CRISIL Liquid Fund Index is suitable. Gilt funds invest only in G-secs, so government securities indices are appropriate.

Hybrid funds use a blend of an equity and a debt index. A scheme at roughly 65% equity might use a synthetic index of 65% Sensex and 35% I-Bex.

Gold ETF — the gold price is the benchmark.

Real estate funds — some indices exist but have shorter histories and are yet to earn the acceptance equity indices enjoy.

International funds — depends on where the scheme invests. Shanghai Composite for China, S&P 500 for the US, or a synthetic blend across several countries.

The two-tier benchmark structure

From SEBI's circular of 27 October 2021, effective 1 January 2022, certain categories use two tiers:

• Tier 1 reflects the category of the scheme
• Tier 2 reflects the investment style or strategy of the fund manager within that category

All benchmarks must necessarily be Total Return Indices.

The logic is neat: where a scheme's mandate already pins it down — a sector fund, an index fund — one benchmark suffices. Where the category is broad enough for a manager to have a style, a second tier exposes that style.

Measuring the manager

Relative return comparison — did the scheme beat its benchmark or peer group? Beating it suggests outperformance; trailing it suggests underperformance. But these are premises that need validation through deeper performance reviews, and AMCs and trustees are expected to conduct such periodic reviews under SEBI guidelines.

The weakness of relative returns, stated precisely: it does not differentiate between two schemes that assumed different levels of risk in pursuit of the same objective. Two schemes may share a benchmark and carry quite different risk. Judging on relative returns alone may be unfair to the fund manager who took lower risk but generated the same return as a peer.

Hence risk-adjusted returns, whose underlying principle is simple: return ought to be commensurate with the risk taken. A manager taking higher risk ought to earn more to justify it; one earning less may justify that through the lower risk taken.

Sharpe Ratio

An investor can earn a risk-free rate (Rf) by lending to the government — the T-Bill index is a good measure of it. A scheme takes risk and earns Rs. The difference Rs − Rf is the risk premium.

Risk-free 5%, scheme return 7%, standard deviation 0.5% → (7 − 5) ÷ 0.5 = 4

Sharpe Ratio is the risk premium generated per unit of risk. Higher is better.

The constraint: comparisons can be undertaken only for comparable schemes. An equity scheme's Sharpe Ratio cannot be compared with a debt scheme's.

Treynor Ratio

Same risk premium, different denominator — Beta instead of standard deviation.

Risk-free 5%, scheme return 8%, beta 1.2 → (8 − 5) ÷ 1.2 = 2.5

Higher is better. And because beta is more relevant for diversified equity schemes, Treynor comparisons should ideally be restricted to such schemes.

The distinction between the two is exactly the distinction from Chapter 10: Sharpe uses standard deviation — total risk. Treynor uses beta — systematic risk.

Alpha

A non-index scheme has a level of return in line with its higher or lower beta relative to the market — call that the optimal return.

Positive alpha indicates outperformance by the fund manager; negative alpha may indicate underperformance. Because it rests on beta, Alpha should ideally be evaluated only for diversified equity schemes.

Information Ratio

Measures risk-adjusted return relative to a benchmark, and is used as a measure of a portfolio manager's level of skill and ability to generate excess returns. It attempts to identify the consistency of performance by incorporating standard deviation into the calculation.

And the fact that carries a mark:

It must be disclosed on the AMC's website daily, and is also available on AMFI's website in a comparable format.

The honest caveat

The workbook ends this section with a warning worth carrying into practice:

Tracking error

The market's beta is 1 by definition. An index fund mirrors the index, so it too has a beta of 1 and ought to earn the market's return. The difference between an index fund's return and the market return is the tracking error.

Its meaning has broadened. Earlier it measured how closely an index fund tracked its benchmark, and was expected to be zero. Now it measures how consistently a fund is able to outperform its benchmark — because it is not enough for a fund to generate a high excess return; it must do so consistently.

A consistently outperforming fund has a LOW tracking error.

That last point is counter-intuitive and heavily tested. Low tracking error is good, because it means the excess return is steady rather than the product of one lucky quarter.

Where performance is disclosed

• The SID must be updated twice every year, so scheme performance numbers are refreshed then
• The fund factsheet fills the gap, published monthly by all fund houses — though it is not a statutory requirement
• AMFI's website carries the performance data of all mutual fund schemes, across periods and categories
• Various portals and product literature, in physical and online form

Product literature is subject to SEBI's advertising guidelines, and typically covers three things:

Suitability — from the product labels, identifying the objective (wealth creation, regular income, liquidity), the asset class, the suitable holding period, and the risk-o-meter to match the investor's risk-taking ability to the scheme's.

Returns — cumulative returns over different holding periods to test whether the objective is being met; comparison with the benchmark to assess the manager; and discreet annual returns to assess consistency across different market scenarios.

Portfolio description — how assets are allocated and securities selected. For equity, diversification across sectors and securities and the strategy on market capitalisation (large/mid/small) and valuation (growth/value). For debt, the tenor and credit quality. For hybrid, the allocation across asset classes and how each is managed.

Factsheets also carry the fund manager's views on the economy and markets, and analyse index levels and yields, corporate results and earnings revisions, government spending and fiscal deficit, inflation, interest rates, domestic and foreign portfolio investor activity, GDP, overseas economic data, and commodity and input prices.