The rate that makes a deal balance
Everything in this rung has used a rate that was *handed to you*: discount at 5%, value the annuity at 4%, build the amortization schedule at the loan rate. But where does that rate come from when you are the one being offered a deal? Turn the question inside out. Write down every cash flow — money out as you pay, money in as you receive — and demand that, at some rate, the present value of what you put in exactly equals the present value of what you get back. That single balancing statement is the **equation of value**, the most important sentence in all of interest theory.
The rate that makes it balance has a name: the **yield rate**, or *internal rate of return* (IRR). "Internal" because it is squeezed out of the deal's own cash flows — no external benchmark, just the numbers in front of you. Pay $1,000 today for a bond that returns $1,080 in a year, and the yield solves 1000 = 1080 ÷ (1 + i), giving i = 8%. The yield is the deal speaking for itself, telling you the constant rate at which it grows your money.
One rate is a fiction — meet the yield curve
So far we have pretended a single rate applies to every future date. Reality is richer. Lending money for one year and lending it for thirty are different bargains, and the market prices them differently — usually demanding more for tying your money up longer. Collect the rate the market charges for each maturity and plot it, and you get the **term structure of interest rates**, better known as the yield curve. It usually slopes gently upward; when it slopes *down* — short rates above long rates — markets are signalling something, and an inverted curve has a long history of preceding recessions.
The curve speaks in two voices. A **spot rate** is the rate you can lock in *today* for money parked from now until a single future date — the honest annual rate for, say, a three-year zero-coupon investment. A **forward rate** is the rate for a future *segment* — the rate from year two to year three, agreed today but not starting until later. The two are tightly bound: rolling at the one-year spot and then at next year's one-year forward must give the same growth as locking in the two-year spot outright, because if it did not, you could borrow one way and lend the other for free money. No-arbitrage stitches the whole curve together.
1-yr spot s1 = 4.0% grow 1 yr: (1.040) 2-yr spot s2 = 5.0% grow 2 yr: (1.050)^2 = 1.1025 implied 1-yr forward, year 1 to 2: (1.050)^2 = (1.040) x (1 + f) 1 + f = 1.1025 / 1.040 -> f = 6.0%
How much does value move when rates move?
Now the question that keeps actuaries awake. A bundle of future cash flows has a present value today, but that value is hostage to the rate used to discount it. Push rates up and every future dollar is discounted harder, so the value falls; pull rates down and it rises. The pressing practical question is *how fast*. If rates jump by 1%, does your portfolio barely flinch, or does it lose a tenth of its worth? The answer is **duration**, and despite its name it is not really about time — it is the sensitivity of value to a change in the rate.
There is a lovely intuition hiding inside. Duration turns out to be the *weighted-average time* at which you receive your cash, where each payment date is weighted by how much of today's value it carries. A 30-year zero-coupon bond pays everything at the very end, so its duration is a full 30 years and its price whips around violently when rates twitch. A bond paying fat coupons every year gets much of its value back early, shortening its duration and steadying its price. The rule of thumb is clean: percentage change in value is roughly minus duration times the change in rate. A duration of 8 means a 1% rate rise costs you about 8% of value.
Why an insurer cares: matching assets to promises
Here is where this stops being arithmetic and becomes the actuary's daily work. An insurer holds two piles of cash flows: assets (the bonds it owns) and liabilities (the claims and benefits it has promised to pay). Both are just future cash flows, so both have a present value and both have a duration. The danger is that they react *differently* when rates move. If liabilities are long-dated promises but the company invested in short bonds, a fall in rates inflates the value of the promises far faster than the assets — and a hole opens in the balance sheet. This mismatch is the core of **interest-rate risk**.
The fix is to make the two piles move *together*. Arrange the asset portfolio so its duration matches the liabilities' duration, and a small rate move pushes both values by the same percentage — the gap stays roughly flat. That deliberate balancing act is the heart of **asset-liability management** (ALM): not chasing the highest yield, but choosing investments whose *interest-rate behaviour* mirrors the promises. Duration matching is its blunt first tool; the next rung will sharpen it into immunization, where you also line up convexity so the surplus is protected against moves in either direction.
Honest limits — a yield is not a guarantee
Three quiet assumptions hide inside a quoted yield, and a careful actuary names all three. First, the IRR silently assumes every interim payment is *reinvested at that same yield* until the end — but the coupons you collect this year must be put back to work at *whatever* the market offers then, which is rarely the original rate. That gap is **reinvestment risk**, and it means a bond's realised return can fall short of its advertised yield even if the issuer never misses a payment.
Second, duration assumes the *whole curve shifts in parallel* — every maturity moving up or down by the same amount. Real curves twist: the short end can leap while the long end barely stirs, and a portfolio that looked duration-matched can still spring a leak. Third, all of this prices only the *time value of money*; it says nothing about whether the issuer will actually pay. A 9% yield that looks generous beside a 4% government bond is usually just the market pricing in the chance you will not be paid at all. The extra yield is compensation for risk, not a free lunch.
With that, the financial engine of interest theory is fully assembled. You can take any stream of cash flows, discount it to a present value, distil it into a single yield, lay the whole market out as a curve of spot and forward rates, and measure exactly how much a value will move when that curve shifts. That last skill — sensitivity — is what carries you out of pure interest theory and into the mortality, valuation, and risk-management rungs ahead, where the cash flows are no longer certain but contingent on whether people live, die, or fall ill.