Bond Valuation: What a Bond Is Actually Worth

Every bond price you see is the output of a single idea: a bond is a stream of future payments, and its value today is what those payments are worth once you account for the fact that money arriving later is worth less than money arriving now.

That’s the whole foundation. Everything technical in fixed income - duration, convexity, spread analysis, callable bond pricing - is an elaboration of that one sentence. This article builds the structure from the ground up, and shows where the standard tools break down.

The core mechanic: discounting

Money has a time cost. ₹100 today is worth more than ₹100 in three years, because today’s ₹100 can be invested and grow. To value a future payment, you reverse that logic: you discount it back to what it’s worth now.

A bond’s value is the sum of the present values of all its future cash flows - every coupon, plus the face value at maturity:

Price = Σ [ Coupon ÷ (1 + y)^t ] + [ Face Value ÷ (1 + y)^n ]

where y is the discount rate per period and t is each period until maturity n.

Work through an illustrative case: a ₹1,000 bond, 8% annual coupon, three years to maturity, discounted at 8%.

The price equals face value precisely because the discount rate matches the coupon. Now change the discount rate to 10% - market rates have risen - and the same cash flows are worth about ₹950. Change it to 6%, and they’re worth about ₹1,053.

Nothing about the bond changed. The issuer still owes exactly ₹80 a year and ₹1,000 at the end. Only the rate at which the market discounts those payments moved. This is the mechanical origin of the price-yield seesaw - not a market convention, but arithmetic.

And it reframes yield to maturity precisely: YTM is the discount rate that makes the present value of a bond’s cash flows equal its current market price. It’s not a forecast; it’s a solved variable.

Why price sensitivity isn’t uniform

If price is a discounted sum, then a change in the discount rate affects distant cash flows more than near ones, because compounding works harder over more periods. A payment 15 years out is far more sensitive to the discount rate than one arriving next year.

This produces the intuition behind duration: a bond’s sensitivity depends on how far away, on average, its cash flows are.

Macaulay duration formalises this as the weighted average time to receive the bond’s cash flows, weighted by each cash flow’s present value. It’s measured in years and has a genuine interpretation: it’s roughly the holding period at which price risk and reinvestment risk offset each other.

Modified duration converts that into a practical sensitivity measure:

Modified duration = Macaulay duration ÷ (1 + y)

and gives the approximate percentage price change for a 1% change in yield. A modified duration of 6 implies roughly a 6% price fall if yields rise 1%.

The word doing quiet work there is approximate.

Convexity: where the straight-line estimate fails

Duration describes a straight line. The actual relationship between price and yield is a curve. Over small yield changes the line is a fine approximation; over large ones it drifts meaningfully from reality.

Convexity measures that curvature - how much duration itself changes as yields move. For a standard bond, the price-yield curve is convex (bowed toward the origin), and this has a consequence that genuinely favours the bondholder:

When yields fall, prices rise by more than duration predicts. When yields rise, prices fall by less than duration predicts.

Convexity is therefore an asymmetry working in your favour. Duration alone overstates your losses and understates your gains.

An illustrative comparison for a bond with modified duration 7 and meaningful positive convexity:

(Figures are illustrative and rounded to demonstrate the asymmetry, not precise calculations for any specific bond.)

Two practical implications follow.

First, convexity is worth paying for - and you usually do. Between two bonds with identical yield and duration, the more convex one is superior, because it performs better whichever way yields move. In efficient markets that advantage is priced in, so higher convexity typically comes with slightly lower yield. The judgement is whether the yield given up is worth the improved asymmetry, and that depends on how volatile you expect yields to be. Convexity is most valuable when rate volatility is high - if yields barely move, you paid for an option you never used.

Second, convexity is why duration-based risk estimates understate resilience in large moves. In a sharp rate shock, a portfolio’s actual loss is typically smaller than a duration-only calculation suggests. Useful to know before panicking at a stress-test output.

When convexity turns against you

Standard bonds have positive convexity. Some structures don’t, and this is where valuation gets genuinely interesting.

A callable bond gives the issuer the right to redeem early. Consider what happens as yields fall: an ordinary bond’s price would rise steadily. But a callable bond’s price gets pinned near the call price, because everyone knows the issuer will call it and refinance more cheaply. The upside is capped.

Meanwhile, if yields rise, the call won’t be exercised, and the bond falls in price like any other. Cap the gains, keep the losses - that’s negative convexity, and it’s the opposite of the asymmetry you want.

Valuing a callable bond therefore requires treating it as two components: a plain bond, minus a call option you’ve effectively sold to the issuer.

Callable bond value = Straight bond value − Value of the embedded call option

Because the issuer holds an option against you, a callable bond must yield more than an otherwise-identical non-callable bond. If it doesn’t, it’s mispriced against you.

This gives rise to yield to call (YTC) - the return if the bond is redeemed at the earliest call date rather than at maturity. Prudent practice is to evaluate a callable bond on yield to worst: the lower of YTM and YTC, across all call dates. Assume the issuer exercises the option in whichever way suits them and disadvantages you, because that is precisely what a rational issuer does.

A puttable bond inverts the arrangement - you hold the option to sell back to the issuer - so its value is the straight bond plus the put’s value, and it should yield slightly less.

Reinvestment risk: the assumption inside YTM

YTM carries an assumption most investors never examine: that every coupon is reinvested at the YTM itself. In reality, you reinvest coupons at whatever rates prevail when they arrive.

If rates fall after you buy, your coupons get reinvested at lower rates, and your realised return falls short of the YTM you were quoted. If rates rise, you do better than advertised. This is reinvestment risk, and it’s the natural counterweight to price risk:

These forces offset, and Macaulay duration identifies where they balance. If your holding period equals the bond’s Macaulay duration, price and reinvestment effects broadly cancel, largely immunising your realised return against a parallel shift in rates. This is the foundation of immunisation strategies used to fund known future liabilities.

Two corollaries worth carrying:

• Zero-coupon bonds have no reinvestment risk, since there are no interim coupons to reinvest. Hold one to maturity and your return is locked at purchase - one reason zeros are favoured for funding a specific, dated obligation. The trade-off is maximum duration, hence maximum price volatility if sold early.

• High-coupon bonds carry more reinvestment risk than low-coupon bonds of the same maturity, since more of their return arrives early and needs redeploying.

Spread measures: decomposing the yield

For credit bonds, the yield decomposes into a risk-free component and compensation for everything else. Professionals use progressively more refined measures to isolate that compensation.

Nominal spread - the simple difference between a bond’s YTM and that of a government bond of similar maturity. Easy to compute, and crude: it compares against a single point on the government curve, ignoring the curve’s shape.

G-spread - the spread over an interpolated point on the government curve matching the bond’s maturity precisely. A refinement of the nominal spread that removes the mismatch from using the nearest available benchmark.

Z-spread (zero-volatility spread) - the constant spread that, when added to every point on the government spot curve, makes the discounted cash flows equal the bond’s market price. Rather than comparing against one benchmark yield, it discounts each cash flow at its own maturity-matched rate plus the spread. This handles a sloped yield curve properly and is the standard measure for comparing bonds with different cash flow profiles.

Option-adjusted spread (OAS) - the Z-spread with the value of any embedded option stripped out. For a callable bond, part of the apparent spread is compensation for the call option you’ve sold, not for credit risk. OAS removes that component to reveal the spread genuinely attributable to credit and liquidity.

The hierarchy matters practically: comparing a callable bond and a non-callable bond on Z-spread will systematically flatter the callable one, because its optionality masquerades as credit compensation. OAS is what makes them comparable. Any time a bond appears to offer unusual spread for its rating, the first question is whether an embedded option explains it.

Where models meet Indian market reality

The mathematics above is universal. Applying it in India requires some honesty about frictions.

Thin trading distorts observed prices. Valuation assumes a market price reflecting many participants’ views. Many Indian corporate bonds trade rarely, so the last traded price may be stale or reflect one motivated seller. A model calibrated to such a price inherits its noise.

Liquidity premium is embedded in the spread. A portion of the spread on an illiquid bond compensates for exit difficulty, not credit risk. Two identically-rated bonds with different Z-spreads may differ purely on tradability. Failing to separate these leads to systematically overestimating the credit compensation you’re receiving.

Benchmark curve construction matters. Spread measures depend on a government curve, and at maturities where government bonds trade thinly, the interpolated curve carries estimation error that propagates into every spread you calculate.

Day count and settlement conventions differ by segment. Government securities are typically quoted clean with accrued interest separate; exchange-traded corporate bonds commonly settle at the dirty price. Comparing prices across segments without normalising produces errors that are small but entirely avoidable.

The disciplined conclusion: use these measures as structured comparison tools, not as precision instruments. A Z-spread computed from a stale price on an illiquid bond is a number with false confidence attached. Knowing the model’s assumptions is what separates using it from being misled by it.

Common analytical mistakes

• Applying duration to large yield moves without convexity. For shocks beyond roughly 100 basis points, the linear estimate drifts materially.

• Evaluating callable bonds on YTM. Use yield to worst. Assume the issuer optimises against you.

• Treating YTM as a guaranteed return. It embeds a reinvestment assumption that rarely holds exactly.

• Comparing callable and non-callable bonds on Z-spread. Optionality inflates the callable bond’s apparent compensation. Use OAS.

• Ignoring the liquidity component of spread. Attributing an entire spread to credit risk overstates how well you’re paid for default risk.

• Trusting model outputs from stale prices. Illiquid bonds produce precise-looking numbers built on unreliable inputs.

• Forgetting that positive convexity is an asset. Between equal-yield, equal-duration bonds, the more convex one is genuinely better.

What this means for you

Bond valuation rewards a small number of durable habits. Understand that price is discounted cash flow, so yields and prices must move inversely. Use duration for a first-order estimate of risk and remember convexity improves the picture - asymmetrically in your favour, unless an embedded option reverses it. Evaluate anything callable on yield to worst. Recognise that your realised return depends on reinvestment, not just the quoted YTM. And when comparing credit bonds, use the most refined spread measure you can, while staying alert to how much of that spread is really compensation for illiquidity rather than default risk.

None of this requires building models yourself. It requires knowing what the numbers on your screen assume - which is exactly what separates an investor who uses analytics from one who is quietly misled by them.

Key takeaways

• A bond’s price is the present value of its future cash flows; YTM is the discount rate that equates those cash flows with the market price.

• Modified duration gives a first-order price sensitivity estimate; convexity corrects it, and positive convexity favours the holder asymmetrically.

• Callable bonds carry negative convexity - evaluate them on yield to worst, never on YTM alone.

• Reinvestment risk offsets price risk; they broadly cancel when your holding period matches Macaulay duration, which underpins immunisation strategies.

• Spread measures progress from nominal → G-spread → Z-spread → OAS; in Indian markets, remember that part of any spread compensates for illiquidity, not credit risk.