Risk, Return and Performance — measuring both sides

Return is only half the answer

Investors pick funds on returns. The workbook is blunt that this is not enough: the return number alone is not adequate to decide whether to invest in or exit a scheme. Suitability must also weigh the risk — chiefly the volatility in returns over time, because the extent of volatility indicates the riskiness of the scheme.

This chapter gives you the tools for both sides.

Four measures of return

1. Simple return

NAV moves from ₹12 to ₹15 → (15 − 12) ÷ 12 = 25%.

It is just the change in value over a period, with no adjustment for time.

2. Annualised return

Two schemes report 5% and 3% since inception. Not comparable — the first ran 6 months, the second 4 months.

• 5% over 6 months → 10%
• 3% over 4 months → 9%

Annualisation makes different periods comparable. Note the first still wins, but by far less than the headline suggested.

3. Compounded return

Annualising works for periods under a year. Beyond that it breaks, because it ignores compounding.

The workbook's demonstration: ₹10,000 at 10% for 3 years, compounded annually, becomes ₹13,310 — not the ₹13,000 that simple interest would give. That ₹310 is compounding, and the gap widens the longer you hold.

₹1,000 growing to ₹4,000 in 2 years: (4000/1000)^(1/2) − 1 = 1, i.e. 100%. Check it: ₹1,000 doubles to ₹2,000 in year one, doubles again to ₹4,000 in year two. Correct.

For a scheme, use the NAV at the start as Initial Value, the NAV at the end as Later Value, and exact days ÷ 365 as n.

4. CAGR — when a dividend has been paid

Here is the limitation that makes the fourth measure necessary.

All three formulas above capture only the change in NAV. But NAV goes down after a dividend is paid — so if the scheme distributed anything, those formulas understate the return.

Where a dividend was paid and compounding matters, SEBI prescribes the CAGR technique, also called the reinvestment method. It assumes the dividend is re-invested in the same scheme at the ex-dividend NAV.

The workbook's worked example:

• ₹10,000 invested at ₹10 on 30 June 2019 → 1,000 units
• 1 January 2020: dividend ₹1 per unit = ₹1,000, reinvested at ex-dividend NAV ₹12.50 → 80 units. Holding: 1,080 units
• 1 January 2021: dividend ₹1 per unit on 1,080 units = ₹1,080, reinvested at ex-dividend NAV ₹15 → 72 units. Holding: 1,152 units
• At ₹15, that is ₹17,280
• Period: 30 June 2019 to 1 January 2021 = 551 days ÷ 365 = 1.51 years

CAGR = (17,280 ÷ 10,000)^(1/1.51) − 1 = 43.65%

The impact of the dividend has been captured as an increase in the number of units. Note that this calculation does not take taxation into account.

Scheme returns versus investor returns

A published return is the scheme's return. Yours will differ, for three reasons:

Loads. With a 1% exit load, a closing NAV of ₹15 pays you only ₹14.85. With a (now banned) 2% entry load, ₹10 would have cost you ₹10.20. Loads drag the investor's return below the scheme's return, and taxes pull post-tax returns down further.

To compute investor returns, substitute:

• Amount actually paid (NAV plus entry load, if any) for the initial NAV
• Amount actually received (NAV minus exit load, if any) for the later NAV

Your own schedule. Additional investments or partial redemptions during the period mean your return differs from the published scheme return even over the identical period.

And note: returns published in a mutual fund advertisement are without factoring entry or exit load.

Holding period returns and rolling returns

Holding period returns are computed for a fixed period — one month, three months, one year, three years, since inception. CAGR is used where the holding period is over one year, and simple absolute returns where it is less.

The weakness: if the initial or end value used was unusually high or low, the picture is distorted. A fund measured from a market bottom looks brilliant; the same fund measured from a peak looks poor. Neither tells you about the manager.

Rolling returns fix this. They are the average annualised return calculated for multiple consecutive holding periods within an evaluation period — for example, every consecutive one-year return within a three-year window, rolled daily, weekly or monthly, then averaged.

Because rolling returns use every starting point rather than one convenient one, they are far harder to flatter. When comparing funds for a client, this is the more honest number.

What else to look at: the workbook asks for consistency of performance, and performance relative to the benchmark and peer group. Its standard for an actively managed fund is a good one to remember: perform well in rising markets and fall less than the benchmark in declining markets.

No promised returns

Mutual funds are not permitted to promise any returns, unless it is an assured returns scheme — which requires a guarantor named in the SID who must write a cheque if the scheme cannot pay.

And the reality: there has not been any assured return or guaranteed product in India. What exists is capital protection-oriented funds — oriented, not guaranteed.

Five measures of risk

Fluctuation in returns is the measure of risk. Periodic returns are computed, then their variation against the average is measured. Variation on the higher side counts as risk just as much as variation on the lower side — a fund that lurches upward unpredictably is still unpredictable.

Variance

Measures fluctuation in a scheme's periodic returns against its own average return. Excel: =var(range).

The workbook's illustration makes the point instantly:

Scheme 2 has a lower average return and nearly sixty times the variance. Worse returns, far more anxiety.

Relevant for both debt and equity schemes.

Standard deviation

The square root of variance — the same idea in the same units as the returns themselves. Excel: =stdev(range).

Standard deviation is a measure of TOTAL risk. A high standard deviation means greater volatility and greater risk. Comparing a scheme's standard deviation with its benchmark and peer group gives perspective, and combined with the average return it estimates the range of returns to expect.

Its limitation, stated plainly: being calculated from historic numbers, it has limited use in predicting future performance.

Annualising standard deviation — multiply by the square root of the number of periods in a year:

So a weekly standard deviation of 0.65% annualises to 0.65 × √52 = 4.70%.

At least 30 observations are needed for a statistically accurate value.

Relevant for both debt and equity schemes.

Beta

Beta comes from the Capital Asset Pricing Model (CAPM), which divides equity risk into systematic and non-systematic. Since non-systematic risk can be diversified away, investors need compensation only for systematic risk — and beta measures that systematic risk.

Beta measures fluctuation in a scheme's periodic returns against the fluctuation of a diversified stock index over the same period.

• The index has a beta of 1 by definition
• Beta above 1 → riskier than the market
• Beta below 1 → less risky than the market

A beta of 0.8 moves 8% when the market moves 10%. A beta of 1.2 moves 12%. Both on the upside and the downside — beta is not a measure of skill, only of amplification.

Beta is relevant only for equity schemes.

Modified duration

Measures the sensitivity of a debt security's value to changes in interest rates. Higher modified duration → higher interest rate risk.

A professional debt fund manager relies on modified duration as the better measure of interest rate sensitivity.

Weighted average maturity

Longer balance tenor → greater fluctuation in value for the same change in interest rates.

A scheme with 70% in a 4-year security and 30% in a 1-year security has a weighted average maturity of (0.70 × 4) + (0.30 × 1) = 3.1 years. Its NAV will fluctuate more than a scheme at 1.5 years.

Being simpler to comprehend, weighted average maturity is widely used in discussions with lay investors — while professionals prefer modified duration.

Credit rating

Indicates the credit or default risk in a scheme.

• Government securities have no credit risk. Cash and cash equivalents have no credit risk. Corporate issuances do
• Higher credit rating → lower default risk
• Better credit rating → lower spread. Poorly rated issuers must offer higher yields to attract investors
• Ratings change. A AAA downgraded to AA raises the yield expected from it, so its market value falls. Conversely, an investor who correctly anticipates an upgrade benefits from the rise

Which measure for which fund

Risks by fund type

Equity funds. Price fluctuation risk rises from large-cap to mid-cap to small-cap, and so does business risk — the risk of a company's business failing. Liquidity risk matters too, and all these risks increase in a focused fund because of portfolio concentration.

The practical advice is good: invest in line with your risk profile, and keep adequate liquidity through liquid funds so you never have to sell equity funds when markets are down.

Debt funds. Many Indians hold non-marketable debt — fixed deposits, small savings — and are simply not used to seeing prices move. But debt fund NAVs fluctuate with interest rate changes and credit migration, so the portfolio may not be as stable as expected. IDCW income is not guaranteed — schemes have skipped dividends for want of distributable surplus.

And the sentence worth carrying into your client conversations: "low risk" does not mean "zero risk". In 2018 and 2019, credit events hit even liquid, ultra-short-term and low-duration funds — categories safe on interest rate risk, but not immune to credit risk.

Hybrid funds. Arbitrage funds are widely believed to be very safe, but some may invest in debt securities and employ paired arbitrage, alpha hedging or merger arbitrage — strategies that carry real risk.

Gold funds. Gold prices are hard to manipulate as an international commodity, giving better pricing transparency, and gold does well when other markets are in turmoil or a currency weakens in war. But if gold prices fall the investor loses. Know whether it is a Gold Sector Fund or a Gold ETF.

Real estate funds. Every asset is different so valuation is highly subjective; the class is illiquid; the agent chain is unorganised; transaction costs (stamp duty, registration) are high; regulatory, litigation and encumbrance risks are high; transparency is low; and many groups are family-owned with poor corporate governance. Real estate funds are high risk relative to other scheme types — yet less risky than direct investment in real estate.

The credit event spiral

This is the most important mechanism in the chapter, because it explains both protections SEBI built.

A ₹10,000 crore scheme holds 8% in debenture M — ₹800 crore. M is downgraded. Investors take fright and redeem 20%, or ₹2,000 crore.

But nobody will buy the downgraded debenture. So the fund manager must sell other, healthier securities to fund those redemptions.

• Corpus falls to ₹8,000 crore
• Exposure to M is still ₹800 crore — now 10%

If more investors panic at the larger exposure and the corpus halves again, exposure to M doubles to 20%. And so on.

The investors who stayed are left holding progressively more of the bad paper. Loyalty is punished, and the maximum a scheme may hold in a single issuer — 10% — is breached not by any decision but by other people leaving.

There is a second version of the problem: market-wide liquidity dries up, and even a well-diversified portfolio of good securities cannot be liquidated to fund redemptions.

SEBI's two answers:

Gating — restriction on redemption

The philosophy is explicit: restrictions apply to market-wide crises, not entity-specific situations. The circumstances must cause illiquidity in almost all securities affecting the market at large, not in one issuer's paper.

Permitted only for:

• Liquidity issues — market at large faces illiquidity. Not for illiquidity in a particular security, and not as a tool to manage a scheme's liquidity or to cover a poor investment decision
• Market failure or exchange closure — unexpected political, economic, military, monetary or other emergencies
• Operational issues — force majeure, unpredictable operational problems, technical failures such as a blackout, occurring despite appropriate diligence and effective disaster recovery

The limits:

• Maximum 10 working days in any 90-day period
• Requires specific approval of the Boards of the AMC and the Trustees, and SEBI must be informed immediately
• No redemption request up to ₹2 lakh is subject to the restriction. Above ₹2 lakh, the first ₹2 lakh is redeemed without restriction and only the excess is restricted

That ₹2 lakh floor is deliberate: it protects small investors entirely, and applies the restriction only to the large redemptions that actually threaten the scheme.

The possibility must be disclosed prominently and extensively in the scheme related documents, along with the time limit.

Segregated portfolio — side pocketing

Permitted from December 2018, to ensure fair treatment to all investors in a credit event and to deal with liquidity risk.

• "Segregated portfolio" = the portion of debt or money market instruments affected by a credit event, separated out
• "Main portfolio" = the scheme portfolio excluding the segregated portfolio
• Triggered by a credit event at issuer level — a downgrade by a SEBI-registered Credit Rating Agency
• Creation is optional and at the discretion of the AMC
• For unrated debt of an issuer with no outstanding rated instruments, segregation is permitted only on actual default of interest or principal

Recall the pricing rules from Chapter 7: no investment and advisory fees on the segregated portfolio, other TER pro-rata only upon recovery, and NAV declared daily.

Every close-ended scheme other than ELSS, and units of a segregated portfolio, must be listed on a recognised stock exchange — and units of schemes in the process of winding up must also be listed, which gives investors an exit route.