The Reserve's rules

Every rule in Operating policy v1 is a response to something in this page. The Reserve is not ours, we cannot change it, and its constraints are the reason the pool behaves the way it does.

Figures below are dated observations at a pinned block, not live values. The thesis ↗ derives these rules in full with its formulas numbered; this page states them as constraints.

The system

Standard Reserve is a closed onchain monetary economy: one currency, $STANDARD; one market where ETH trades against it; one policy signal, the net flow of ETH through that market; and a central bank written as immutable code. It cannot be paused, voted on or upgraded.

Its operators are bankers. A banker holds a charter, and a charter operates between one and ten branches. One thousand founding charters were sold at launch.

Issuance is one pie, divided by branch

Each day the central bank issues 700,000 $STANDARD multiplied by a policy rate m, and divides it equally among every branch in existence, credited second by second. One branch earns, per day:

y = 700,000 × m / N

where N is the total branch count across the whole Reserve. m is set from the market's net ETH flow and is bounded to the range 0.2 to 1.25.

Consequence for the pool: every new branch anywhere — including ours — dilutes every existing branch. Growth in the Reserve reduces per-branch earnings, and the pool's modelled returns are stated against an explicit assumption about how far that growth goes.

Earnings are a balance, not tokens

A branch's earnings accrue as a balance inside the Reserve, attributed to its charter. They are not tokens in a wallet and cannot be claimed as such.

$STANDARD is minted at one moment only: when a banker retires branches. Retiring k of a charter's n branches releases

R = B × (k / n) × (1 − φ)

where B is the charter's whole balance and φ is the exit fee. The retired branches are destroyed. Retiring a charter's last branch destroys the charter.

Consequences for the pool: there is no way to draw income while keeping the branch that earned it; a payout always costs capacity, which must then be bought back. A charter holding a single branch cannot take anything at all without destroying itself.

Branches are bought at a daily auction, and burned

A new branch needs a licence, bought at a daily Dutch auction, paid in $STANDARD which is burned entirely. The auction offers 100 licences a day across the whole Reserve, at most three per charter.

The price opens at twice the previous day's close and decays towards a floor of two days of one branch's issuance. The useful measure is τ, the number of days a branch takes to earn back its own licence:

τ = P / y

Consequences for the pool: a licence is a burned cost, not resaleable inventory, so its purchase price is never carried as pool value. Capacity cannot be added faster than the auction allows, and the pool sets a ceiling on τ rather than buying at any price.

Leaving is taxed, and so is silence

Event	Charge
Trading into $STANDARD	2%
Trading out of $STANDARD	3%
Retiring branches	The exit fee φ, which rises when many are exiting at once
Thirty days of inactivity	70% of the bank's balance, and anyone may close it

Half of each penalty is burned; half is paid to the banks that remain.

Consequences for the pool: contributed ETH loses the trade-in tax on its way to becoming branches, and that cost is counted as part of what members put in. The exit fee makes the timing of a retirement matter, which is why policy v1 refuses to retire above a stated fee ceiling. The dormancy rule makes liveness a hard operational requirement, not a nicety.

Why a payout has a cost, and why patience is the lever

Because a retirement releases a share of the balance but destroys the branches that earned it, the cost of re-opening — measured as a share of what the retirement releases — is

c = (τ / t) × (1 − φ)

where t is the charter's balance expressed in days of its current earnings. The number of branches retired cancels out entirely: c depends only on how long the balance has been allowed to build.

While t is below τ, a payout costs more than it releases. This single relation is why the pool holds rather than distributes early, and why policy v1 expresses its regular payout trigger as a threshold on c rather than as a schedule.

What this constrains, in short

• Income arrives in lumps, on the pool's timing, not on a member's request.
• Early payouts destroy value; the discipline to wait is the strategy.
• Branch capacity is burned cost, never a saleable asset.
• Reserve-wide growth dilutes the pool, and the pool cannot prevent it.
• The operator must stay live or the balance is forfeited.

Public contributions are not open. Continue to Operating policy v1 for the rules the pool sets within these constraints.