Where you are. Lesson 1 put two doors on a balance sheet and lessons 2 to 4 walked through the equity one, ending with an event pushed down four levels of custody. This lesson goes back to the debt door. A loan between two parties is a private arrangement; a bond is that same arrangement written in a standard shape so it can be sold to somebody else entirely, which is the whole reason bond markets exist.
An IOU with a resale value
Lend a friend a hundred for five years at four percent and you have an asset: a claim on four a year, plus the hundred back at the end. It is also completely illiquid. If you need the money in year two, your only options are to ask your friend nicely or find somebody willing to take over a private, unstandardised, unverifiable arrangement with a stranger.
Now write the same loan in a shape everybody recognises. Who owes the money. How much will be repaid at the end. What rate is paid in the meantime. How often. When it ends. Five numbers, standard fields, verifiable by anyone.
That version can be sold. Not because the borrower is any better, but because the claim can now be described completely to a buyer who has never met them. Standardisation is what turns a loan into a security.
The idea in one paragraph
A bond is a tradeable loan, and it is defined completely by a handful of fields: the issuer who owes the money, the face value repaid at the end, the coupon rate paid periodically along the way, the frequency of those payments, and the maturity date when the last one lands. From those fields the entire future of the instrument follows as a schedule of dated payments, which you can generate as an array and which is the only thing that matters about the bond. Everything the next three lessons do - pricing it, converting between price and yield, measuring its sensitivity to rates - is arithmetic performed on that array. A bond is not a thing. It is a schedule.
The schedule, generated
Take the five fields: a 100 face value, a 4% annual coupon, five years to maturity, paid once a year, issued by a workshop you have been reading about all module.
Its whole future is nine words of arithmetic. Years one through four pay the coupon alone: four each. Year five pays the coupon and the face together: a hundred and four. Total promised, a hundred and twenty.
Wider than the screen; scroll it sideways.
What the fields do not tell you
The schedule says what is promised. It says nothing about what is likely.
A government bond and a struggling company’s bond can carry identical fields and identical schedules, and be worth wildly different amounts, because one issuer will almost certainly pay and the other might not. Nothing in the five fields captures that. It shows up in the price, and lesson 9 gives the difference its name.
Check yourself
1. Two bonds have identical fields but very different prices. What differs, given the schedules are the same?
The likelihood the schedule is honoured. The fields describe promises, not the issuer’s ability to keep them, so a riskier issuer’s identical promises are worth less. That difference in price is the market’s charge for taking on the possibility of default.
2. A 4% coupon bond trades at 80. What do you receive each year, and what is your return before maturity?
You receive 4, because the coupon is a rate on the 100 face value and has nothing to do with what you paid. On an outlay of 80 that is a 5% running return, and you also stand to receive 100 at maturity for the 80 you spent, so the full return is higher again.
3. Why does standardising a loan make it tradeable, when the borrower’s creditworthiness is unchanged?
Because a buyer can evaluate a standard set of fields without knowing anything about the original arrangement or the parties to it. The loan’s terms become describable, comparable and verifiable, so the pool of possible buyers grows from people who trust the original lender’s judgement to anyone who can read the fields.
4. Roughly seven-eighths of this bond’s promised money arrives on one day. What does that imply for pricing it?
That the rate used to discount the final payment matters far more than the rates used for the coupons, so the bond’s value is dominated by the far end of the curve. It also means the bond’s price reacts strongly to long-term rate changes, which is the sensitivity lesson 8 measures.
Do this
Generate the schedule from the fields, and check it totals what it should.
python3 code/bond_anatomy.pyThe starter leaves you cash_flows, which turns the five fields into the array of dated payments: the coupon in each year, and the coupon plus the face in the last. The assertions check the count, the first payment, the last, and the total promised. Success prints the schedule as a small bar chart and the line:
a bond is not a thing but a schedule: four small promises and one large one, and every question about it is a question about theseThe completed version is in solutions/bond_anatomy.py.
What you can now do. You can read any bond’s defining fields and generate the exact array of payments it promises, which is the input to everything that follows. What you cannot yet do is say what that array is worth today, because a payment in five years is plainly not worth its face amount now. The next lesson builds the machinery that answers it, and module 1 has owed you that derivation since its T-bill.