Where you are. You have an issuer whose tokens are backed one for one by reserve assets sitting in a bankruptcy-remote entity. Those assets are bills and repo. Bills and repo pay interest. Nobody has yet said where that interest goes.
Here is the shape of the business, and it is worth stating baldly because it is genuinely elegant.
You give the issuer a dollar. The issuer gives you a token. Your token pays you nothing, ever. The issuer takes your dollar, buys a three-month bill with it, and keeps the coupon. You can send your token to anyone, at any hour, for a fee that is often nothing at all.
Nobody is being defrauded and nothing is hidden. It is simply worth being clear that the free transfer is not free; it is paid for by the interest you are not receiving.
The idea in one paragraph
An issuer’s revenue is the short rate times the money it is holding, and its costs are largely fixed, so the business is enormously sensitive to interest rates and enormously scalable in size. The float is the customer money it holds at zero interest, and it is the whole asset. Paying holders a share of the yield would be the obvious competitive move, and issuers mostly do not, because a token that pays a return starts to look like an investment product to a securities regulator, with an entirely different and far heavier rulebook attached. So the yield stays with the issuer, and the visible price of using the token stays at nothing.
What the grid shows
The exercise computes the annual take across three supplies and three short rates, against a stylised fixed operating cost of forty million.
Two things fall out. The first is that rates dominate: the same hundred billion of tokens earns five billion a year at five percent and two hundred and fifty million at a quarter of a percent, a twentyfold difference in revenue from a business that is otherwise unchanged. The second is that scale is survival. At a quarter of a percent, an issuer needs sixteen billion of tokens outstanding merely to cover its costs, so a small issuer in a low-rate world simply loses money.
That combination explains a good deal of the sector’s behaviour: the push for scale at almost any cost, the sensitivity of the whole industry to a rate-cutting cycle, and the interest in adjacent businesses that do not depend on the short rate staying where it is.
A bank does this too
Hold the model next to module 1’s bank and the resemblance is uncomfortable.
A bank also takes money it does not own, also invests it, and also keeps a margin. The differences are the interesting part. A bank lends its deposits out, creating new money in the process, while an issuer is required to hold safe short assets and creates nothing. A bank pays its depositors something, at least in principle, and competes on that rate. A bank has deposit insurance behind it and a central bank that will lend to it in a crisis; an issuer has neither, which is why the quality of its reserve assets carries the whole weight.
So the issuer’s business is a narrower, safer version of banking that keeps more of the margin. Whether that is a better deal for the holder depends entirely on what the holder wanted: a claim that pays nothing but is backed by bills, or a claim that pays a little and is backed by loans with a government guarantee wrapped around it.
Where the money actually goes
One consequence is worth noting because it connects this module back to module 4.
The reserve assets have to be held in something safe and short. That means bills and repo, in size. Collectively, the issuers have become a meaningful buyer of short-dated government paper, which makes a payment-token business into a participant in the money market whether it thinks of itself that way or not. The money you put in to make payments ends up funding a government’s short-term borrowing, through a chain of two intermediaries and one legal structure.
Check yourself
1. Where does the return on the reserve assets go, and what do holders get?
To the issuer. Holders get nothing beyond the ability to redeem at par and to transfer the token. The free transfer is paid for by the interest the holder forgoes, which is a real price even though no fee is charged.
2. Why do issuers mostly not share the yield, when doing so would be an obvious way to compete?
Because a token paying a return looks like an investment product to a securities regulator, which brings a different and much heavier rulebook. The zero return is a deliberate legal design choice that keeps the instrument classified as a payment token.
3. At a quarter of a percent, an issuer needs sixteen billion outstanding to break even. Where does that number come from?
From dividing the fixed operating cost by the rate: forty million divided by 0.0025. Costs are largely fixed while revenue is the rate times the float, so the break-even size moves inversely with the rate, and a rate-cutting cycle raises the size at which the business is viable at all.
4. The issuer’s revenue is the product of two numbers. Why is that its strategic problem?
Because it chooses neither of them. The short rate is set by the central bank and the amount outstanding is set by market demand. That explains the sector’s push for scale, its exposure to a rate-cutting cycle, and its interest in revenue that does not depend on the short rate.
Do this
Compute the take across the grid.
python3 code/reserve_yield.pyThe starter leaves you annual_take: the reserve assets earn the short rate, holders receive nothing, and the operating cost comes off the top. The assertions check that a large issuer earns far more at five percent than at a quarter of a percent, and that a small issuer at a quarter of a percent loses money outright.
The completed version is in solutions/reserve_yield.py.
What you can now do. You can compute an issuer’s annual revenue from the float and the short rate, explain why the yield stays with the issuer as a legal choice rather than an oversight, and say who is actually paying for a free transfer. The next lesson looks at what the reserve assets are actually made of, and how fast they turn into cash on a bad day.