Yield Curve Bootstrapping

Every loan, bond and derivative carries an interest rate — but rates differ by how long the money is tied up. The yield curve is the map: a function from maturity to rate that summarizes what the market currently demands at each horizon.

The snag is that the market does not quote zero-coupon rates directly. What you see in the data are coupon-bearing bonds — instruments that pay periodic interest coupons plus a final principal at maturity. Each price is a blend of multiple maturities because each coupon arrives at a different date.

Bootstrapping is the algorithm that untangles this blend. Starting at the shortest maturity and working outward, it strips coupons away one by one until only a single, clean payment at a single date remains — a zero-coupon bond. That gives you one discount factor. The next instrument adds one more. Repeat until the whole curve is pinned down.

The method was formalized in modern finance through the work of practitioners in the 1970s and 1980s, and it remains the foundation of every interest-rate desk. There is no deep open question: bootstrapping is a solved, exact algorithm — but understanding why it works reveals the structure of present value, arbitrage and the entire machinery of fixed-income pricing.

Below are four market instruments: a 6-month deposit, a 1-year bond, a 2-year bond, and a 3-year bond. Each has a quoted price (or par yield). The algorithm solves for one new discount factor per instrument, using all previously found factors to strip prior coupons.

<!-- {{c_html_intro}} -->
<div class="controls">
  <p class="hint">{{hint_para}}</p>
  <table id="input-table">
    <thead>
      <tr>
        <th>{{th_maturity}}</th>
        <th>{{th_coupon}}</th>
        <th>{{th_price}}</th>
      </tr>
    </thead>
    <tbody>
      <tr>
        <td>0.5</td>
        <td class="coupon-cell"><input type="number" id="c0" value="0" min="0" max="20" step="0.1" aria-label="{{aria_coupon}} 0.5"></td>
        <td class="price-cell"><input type="number" id="p0" value="97.5" min="50" max="110" step="0.01" aria-label="{{aria_price}} 0.5"></td>
      </tr>
      <tr>
        <td>1.0</td>
        <td class="coupon-cell"><input type="number" id="c1" value="2" min="0" max="20" step="0.1" aria-label="{{aria_coupon}} 1.0"></td>
        <td class="price-cell"><input type="number" id="p1" value="99.0" min="50" max="110" step="0.01" aria-label="{{aria_price}} 1.0"></td>
      </tr>
      <tr>
        <td>2.0</td>
        <td class="coupon-cell"><input type="number" id="c2" value="3" min="0" max="20" step="0.1" aria-label="{{aria_coupon}} 2.0"></td>
        <td class="price-cell"><input type="number" id="p2" value="98.5" min="50" max="110" step="0.01" aria-label="{{aria_price}} 2.0"></td>
      </tr>
      <tr>
        <td>3.0</td>
        <td class="coupon-cell"><input type="number" id="c3" value="4" min="0" max="20" step="0.1" aria-label="{{aria_coupon}} 3.0"></td>
        <td class="price-cell"><input type="number" id="p3" value="98.0" min="50" max="110" step="0.01" aria-label="{{aria_price}} 3.0"></td>
      </tr>
    </tbody>
  </table>
  <div class="btns">
    <button id="run" type="button">{{btn_run}}</button>
    <button id="reset" type="button" class="ghost">{{btn_reset}}</button>
  </div>
</div>
<div id="results" class="results hidden">
  <table id="result-table">
    <thead>
      <tr>
        <th>{{th_maturity}}</th>
        <th>{{th_df}}</th>
        <th>{{th_zero}}</th>
        <th>{{th_step}}</th>
      </tr>
    </thead>
    <tbody id="result-body"></tbody>
  </table>
  <div class="chart-wrap">
    <canvas id="chart" width="340" height="160" aria-label="{{aria_chart}}"></canvas>
  </div>
  <p id="status" class="status"></p>
</div>

/* {{c_css_intro}} */
* { box-sizing: border-box; }
body { font-family: system-ui, sans-serif; color: #222; margin: 0; font-size: 14px; }
.hint { font-size: .85rem; color: #444; margin: 0 0 .6rem; line-height: 1.5; }
table { border-collapse: collapse; width: 100%; margin-bottom: .6rem; }
th, td { border: 1px solid #cdd9e3; padding: .3rem .5rem; text-align: right; }
th { background: #e8eef3; font-weight: 700; color: #1d3557; text-align: center; }
td:first-child { text-align: center; font-weight: 600; color: #1d3557; }
input[type=number] { width: 70px; border: none; background: transparent; font-size: 14px;
                     color: #222; text-align: right; outline: none; }
input[type=number]:focus { background: #f0f6ff; border-radius: 3px; }
.btns { display: flex; gap: .5rem; margin-bottom: .7rem; flex-wrap: wrap; }
button { font: 600 13px system-ui, sans-serif; padding: .4rem .8rem; border: 1px solid #1d3557;
         background: #1d3557; color: #fff; border-radius: 8px; cursor: pointer; }
button.ghost { background: #fff; color: #1d3557; }
.hidden { display: none; }
.results { margin-top: .4rem; }
#result-table td { font-variant-numeric: tabular-nums; }
.step-cell { font-size: .75rem; color: #555; text-align: left; max-width: 160px; }
.chart-wrap { margin: .6rem 0; }
canvas { display: block; background: #f8fafc; border: 1px solid #cdd9e3; border-radius: 6px; }
.status { font-size: .85rem; color: #0a7d33; font-weight: 600; margin: .3rem 0 0; min-height: 1.2em; }
.status.err { color: #c92f3c; }

// Code not found

Edit the prices or coupon rates and press Bootstrap to rebuild the curve. Notice that each new discount factor depends only on the current instrument and the already-known shorter-maturity factors — the algorithm never needs to re-solve earlier steps.

Bootstrapping is deceptively simple once you see its algebraic structure.

Present value identity. A coupon bond priced at $P$ with coupon rate $c$, face value $F$, and payment dates $t_1, t_2, \dots, t_n$ satisfies

$$P = c \cdot F \cdot \sum_{i=1}^{n} d(t_i) + F \cdot d(t_n)$$

where $d(t)$ is the discount factor — the present value of one unit of currency received at time $t$. For a zero-coupon bond maturing at $t_n$ this collapses to $P = F \cdot d(t_n)$, so $d(t_n) = P / F$.

The triangular structure. Instruments are ordered by maturity. The first instrument has a single cash flow, so $d(t_1)$ is found directly. The second instrument has two cash flows: the coupon at $t_1$ (already known) and the principal at $t_2$ (unknown). Rearranging,

$$d(t_2) = \frac{P_2 - c_2 \cdot F \cdot d(t_1)}{F \cdot (1 + c_2)}$$

Each new instrument adds exactly one unknown, and all prior factors are already pinned. The system is lower-triangular — no iteration, no least-squares, no approximation needed. Forward substitution solves it in $O(n^2)$ time.

Zero rates from discount factors. Once you have $d(t)$, the continuously-compounded zero rate is

$$r(t) = -\frac{\ln d(t)}{t}$$

or in annual compounding $r(t) = d(t)^{-1/t} - 1$. These are the rates the market would demand to lend for exactly $t$ years with no intermediate cash flows — the clean, unblended numbers that all derivative pricing formulas require.

Where it can break. Bootstrapping requires liquid, well-spaced instruments. If maturities overlap, the system becomes over-determined and a least-squares fit (or spline interpolation) replaces the exact solution. Real desks use parametric models — Nelson-Siegel, Svensson — that fit the whole curve simultaneously, trading exactness for smoothness.

The zero-coupon curve produced by bootstrapping sits at the center of nearly all fixed-income finance:

• Swap pricing: an interest-rate swap exchanges fixed coupon payments for floating ones. Its fair fixed rate is exactly the par yield read off the bootstrapped curve at the relevant maturity.
• Bond valuation: to check whether a bond is cheap or rich relative to the market, price it by discounting each cash flow at the zero rate for its payment date. Any deviation from par signals a spread.
• Options on rates: models for caps, floors and swaptions (such as Black-76 or the SABR model) require the forward rates implied by the zero curve as their starting point.
• Risk management — DV01 and duration: sensitivities of a portfolio to rate moves are computed by bumping the curve by one basis point and repricing. You cannot do this without a curve.
• Central bank analysis: policy rate decisions are informed by what the market has priced in at future meetings, read directly from the bootstrapped overnight-index-swap (OIS) curve.

The algorithm also connects to broader computational themes. Present-value decomposition is a close cousin of dynamic programming — breaking a complex pricing problem into independent sub-problems solved in order. And the condition number of the bootstrapped system rises with curve steepness, a numerical stability question shared with many triangular solvers.

Yield curve bootstrapping is one of the most elegant examples of a solved algorithm in quantitative finance: it extracts hidden information — zero-coupon rates — from observable prices using nothing more than the present-value identity applied repeatedly in the right order.

The triangular structure means no iteration is needed, no approximation is made, and no data is discarded. Each new instrument reveals one new rate, building the curve from the short end outward until the whole term structure is pinned down.

From that single object — the zero-coupon discount curve — the market can price any fixed cash flow, measure any interest-rate risk, and read the collective forecast embedded in bond prices. It is, in a very literal sense, the foundation of everything else in fixed-income finance.