Where you are. You have an issuer, a legal cage and a business model. All three assume the reserve assets are what the issuer says they are. This lesson is about the document that tells you, and about how much less it says than people assume.
Every large issuer publishes a monthly breakdown of what it holds. It is usually a single page: a table of instrument types with amounts, signed off by an accounting firm.
The instinct is to check one number, the total, against the tokens in issue, and stop. That instinct is wrong, and the reason is the thing worth learning here. Two issuers can both be exactly one hundred percent reserved and be in completely different positions on the morning everybody asks for their money back.
The idea in one paragraph
An attestation is an accountant’s confirmation that a stated set of holdings existed at a stated moment. It is not an audit: it does not test the issuer’s controls, does not cover the days either side, and does not opine on whether the business is a going concern. What it does contain, if you read past the total, is composition, custodian and dates, and from those three you can compute the only number that matters in a run: how much of the portfolio becomes cash today, how much tomorrow, and how much only next week. Full backing is a statement about value. A run is a question about timing, and the two are answered by different columns of the same table.
The ladder
The exercise takes a synthetic breakdown and computes what fraction converts within each horizon.
The portfolio is entirely respectable: overnight repo, bills under thirty days, bills out to ninety days, some bank deposits, a little corporate paper. Nothing exotic and nothing that would raise an eyebrow. It is a hundred percent reserved and every instrument in it is investment grade.
The ladder still says that only 37 percent of it is available the same day. Everything else needs at least until tomorrow, and a slice of it needs a week.
| horizon | share of reserve assets available |
|---|---|
| same day | 37.0% |
| next day | 85.0% |
| within two days | 97.0% |
| within a week | 100.0% |
n = 5 trials · synthetic reserve portfolio, conversion to cash by horizon
That 37 percent is the number a run tests. If holders ask for more than that on day one, the issuer is not insolvent, it is illiquid, and lesson 8 shows that the market cannot easily tell the difference while it is happening.
What to actually read
Three things, in order.
Composition first, because it drives everything else: instrument types and maturities give you the ladder. Custodian second, since reserve assets held at a bank are a claim on that bank, and a portfolio described as safe can still be concentrated at one counterparty, which is the fact behind the worst week this sector has had. Timing third: how often the snapshots come, and whether the firm signing them does anything more than confirm the arithmetic.
Notice that none of the three is the headline percentage.
Check yourself
1. What does an attestation confirm, and what are the three things it does not?
It confirms that stated holdings existed at a stated moment. It does not test the issuer’s internal controls, does not cover any period other than that instant, and does not opine on the business as a going concern. It answers a narrow question narrowly.
2. A portfolio is one hundred percent reserved and only 37 percent of it is available same-day. Is the issuer in trouble?
Not necessarily, and that is the point. It is fully backed by value and constrained by timing. If redemptions on day one stay under 37 percent of the book it is fine; above that it must sell or wait, and it becomes illiquid rather than insolvent. Those are different conditions with different remedies, and from outside they look identical.
3. Why does the custodian line matter as much as the composition line?
Because reserve assets held as a bank deposit are a claim on that bank rather than on a government. A portfolio can be entirely investment grade and still be concentrated at one counterparty, so the issuer inherits that counterparty’s failure. Concentration is invisible in a total and visible in a custodian breakdown.
4. Why prefer frequent attestations over a single impressive one?
Because a snapshot at a date the issuer knew in advance constrains nothing about the days either side. More frequent snapshots make it progressively harder for the reported position to be unrepresentative of the ordinary one, which is a weaker guarantee than an audit and a considerably stronger one than an annual photograph.
Do this
Compute the ladder from the breakdown.
python3 code/attestation.pyThe starter provides a synthetic attestation as instrument, amount and days to cash, and leaves you liquidity_ladder: for each horizon in the table, what fraction of the total converts by then. The assertions check that the ladder increases, that everything converts eventually, and that same-day availability is under half the book.
The completed version is in solutions/attestation.py.
What you can now do. You can read a reserve disclosure for composition, custodian and timing rather than for its headline, compute the liquidity ladder that a run actually tests, and explain why fully reserved and instantly redeemable are different claims. The next lesson follows the tokens themselves in and out of existence.