An Insurance Fund Looks Like a Cap on Loss

An insurance fund looks like a cap on loss. A dashboard prints a reserve and treats open risk as covered. The reserve is the first story. The size of the positions that can break it is not.

A fund can pay a hole after a liquidation fails. It can also run out in one window. I read an insurance line the way a fire bucket is read: the bucket is real. It is not the building.

• A bucket is the first reading
• Auto-deleveraging is the next room
• A fund in one unit, risk in another
• Sheet I kept for a fund called CUSHION
• A bucket that actually covers a window
• Measure open risk before you treat the reserve as a cap

A bucket is the first reading

The problem is easy to name. A perp or options venue keeps a reserve. The first reading treats that reserve as the maximum anyone can lose beyond margin. The reserve is a balance. Open interest and the worst remaining book are different objects.

Is the first reading always false? No. Some venues publish the fund, the rules for when it pays, and a hard stop that pauses new risk when the fund is thin. The first reading fails when the fund is small next to concentrated positions, when the same token is both the fund and the thing being traded, or when auto-deleveraging is hidden behind the word insurance.

My claim is narrow. An insurance fund is a pot that may pay. A cap on loss is a later fact: after the pot is empty, the rules still name who pays. Mixing them turns a bucket into a roof.

Opinion, not a law for every venue: the reserve number is a costume. The ADL clause is the room.

Auto-deleveraging is the next room

When a liquidation cannot fill, and the fund cannot pay, many designs take from winning accounts. That is not insurance. That is a second payer.

Question I keep on the page: after the fund hits zero, who is next in line? If the answer is “profitable positions, ranked by leverage,” write that rank. Do not stop at the reserve chart.

Exception: a venue that only pauses and does not haircut winners is a different object. Pause is a stop. It is not a payment.

A fund in one unit, risk in another

A fund held in the venue’s own token is a pot that moves with that token. Risk measured in an outside unit does not shrink because the pot’s ticker looks large.

Observation from public pages: I have lined a fund balance next to open interest in the same window and found the interest several times the pot. That ratio is a fact on two numbers. It is not a forecast.

Condition: if the fund is in an outside unit and cannot be minted by the venue, write that. If the venue can mint the fund token, the bucket can look full on paper.

Sheet I kept for a fund called CUSHION

I keep a sheet for a made-up venue I call CUSHION. Fund: 8 units of an outside coin. Open interest that window: 90. Docs: fund pays first, then ADL. A dashboard badge said “protected.” One large account held 30 of the 90.

Question in the margin: protected from what? Answer I could defend: from a hole smaller than 8, if the fill works. Not from a 30-unit account that fails to fill.

Was I looking at a live book? No. CUSHION is a page. The experience was putting pot and open interest on one row. A reader can repeat that row on any public venue page without taking a position.

Exception I left beside the sheet: if CUSHION later caps a single account far below the fund, the 30 cannot exist. Until that cap exists, the badge is ahead of the book.

A bucket that actually covers a window

The first reading works when the fund is in an outside unit, when it is large next to the biggest accounts, and when the next payer after the fund is written in public. It works for that window.

It fails when a reserve badge is treated as a roof over all open risk. It fails when ADL is counted as insurance because the same page shows both words.

I do not treat a thin fund as a command to act. I treat it as a reason to keep the bucket and the book on two lines.

Measure open risk before you treat the reserve as a cap

The solution that holds under the conditions above is a short measure, not a slogan.

Write the fund balance and its unit. Write open interest in the same window. Write the largest accounts if the venue shows them. Write who pays after the fund is empty. If a page will not show that next payer, the “cap on loss” sentence is not ready to stand.

If two dashboards disagree on the fund, say so and stop before the roof becomes certain.

The spare thought on the desk is small. A bucket can be real water. The floor can still be larger than the bucket. I measure the open book first. I do not call the reserve a cap because the first reading stopped at the badge.

The articles on this site are not investment recommendations or financial advice. They are structural analysis based on on-chain data and project documents.