Where you are. You know what a wrapper is and where its links break. This lesson asks the empirical question: with all of this possible, what did people actually tokenise?
The answer is not what the early enthusiasm predicted. Not property, not art, not private company shares. The largest category by a wide margin is short-dated government debt, which is the most boring asset in this course.
That is not a disappointment. It is a clue about what tokenisation is actually good for.
The idea in one paragraph
An asset tokenises well when the three links from lesson 12 are cheap: custody is straightforward, the legal claim is standard, and servicing is nearly nothing. Treasuries score perfectly on all three, and they are also liquid and easy to price, which means the wrapper’s value can be checked continuously against a real market. Property scores badly on all three, which is why the tokenised-property projects that got the most attention produced the least volume. The pattern is that tokenisation adds most where the asset was already easy and the distribution was the constraint, and least where the asset itself was the problem.
Wider than the screen; scroll it sideways.
Why gold is on the list and buildings are not
Gold is the one physical thing in the table with real size behind it, and it is worth asking why, because the answer confirms the rule rather than breaking it.
Score gold on the three links. Custody is a vault with a number on a bar, which is about as clean as physical custody gets. The legal claim is a standard warehouse arrangement that has existed for centuries. Servicing is nothing at all: gold pays no coupon, has no tenant, needs no maintenance, and does not default. It is a physical asset with the servicing profile of a bearer instrument.
Now score a building. Custody means somebody must actually control the property. The claim is jurisdiction-specific and the transfer usually touches a land registry that has never heard of a shared ledger. Servicing means tenants, repairs, insurance, and occasionally eviction proceedings.
Same physicality, opposite scores, and the volumes match the scores rather than the enthusiasm.
The weekend problem
A tokenised bill trades every hour of every day. The bill market does not.
The exercise runs sixty hours of weekend token trading against an underlying that never opened, and measures the gap at Monday’s morning. The token drifted; the bill’s actual value did not move, because there was no market in which it could.
That gap is not an error. It is the market’s collective guess about where the bill will open, expressed in the only instrument that was trading. But it means the round-the-clock claim needs qualifying: the wrapper trades continuously, and it is trading a forecast for two days a week rather than an asset with a current price.
What this predicts
If the pattern holds, the next categories to grow are the ones that score well on the same three links: repo, money market fund shares, and other short instruments held by institutions that already have custody arrangements.
The categories that have been promised for a decade and not arrived, property most of all, are the ones where the asset itself is the problem. A tokenised building still needs somebody to value it, insure it, maintain it, and evict a tenant who stops paying. The wrapper does not help with any of those, which is exactly what lesson 12 predicted.
Check yourself
1. Why do Treasuries lead the table rather than property?
Because they score well on all three links from lesson 12: custody is straightforward, the legal claim is standard, and servicing is almost nothing. They are also liquid and easy to price. Property scores badly on every one of those, and a wrapper does not improve any of them.
2. Two credible sources give totals that differ by nearly half. Why, and what should you do about it?
Because they count different categories: one includes private credit and real estate, the other does not. Neither is wrong. The response is to quote any figure here with its scope, its source and its date, because in a market this young the measurement is part of the subject.
3. A tokenised bill trades all weekend and its price moves. Has something gone wrong?
No. The underlying market is shut, so the token’s price is the market’s guess about where the bill opens on Monday. It is a forecast rather than an error, but it does mean the round-the-clock claim needs qualifying: the wrapper trades continuously, and for part of the week it trades an opinion.
4. Which categories does this pattern predict will grow next?
Instruments that score well on the same three links and are already held by institutions with custody arrangements: repo, money market fund shares and similar short paper. The long-promised categories, property above all, are the ones where the asset itself rather than its distribution is the problem.
Do this
Measure the weekend gap.
python3 code/weekend_gap.pyThe starter runs sixty hours of seeded token trading from Friday’s close and leaves you the summary: the underlying is still at Friday’s price because its market never opened, and you report the token’s close, the gap and the gap as a fraction. The assertions check that the underlying did not move, that the token did, and that a weekend of drift opens a gap of real size.
The completed version is in solutions/weekend_gap.py.
What you can now do. You can say which assets tokenise well and why, quote the current figures with their scope and date, and measure the gap a continuous wrapper opens over a closed market. The next lesson issues a security natively rather than wrapping one.