Week 11: Counterparty risk

The week where the market is fine, your judgement was fine, and the money is gone anyway — because you were diversified in the wrong dimension.

17 May 2027rationedhanded to you, never assessed

Ten weeks of this course have been about your position. This week is about the phrase “your position”, which is doing more work than it looks.

What you own is a promise

You do not, in most arrangements, hold the thing you bought. A broker holds it. Often the broker does not hold it either — a custodian holds it for them, which is a firm whose whole business is keeping records of who owns what. And if the thing you bought is itself a promise to pay you later, then an issuer owes you the money and everything above is just bookkeeping about the promise.

So there is a chain, and at every link what you own is an entry on somebody else’s books saying you are owed something. That entry is excellent while the somebody is solvent. It is worth what the courts eventually say it is worth when they are not.

Counterparty risk is the chance that one of those parties fails to do their part. It is unrelated to whether you were right about the instrument. You can pick well, size sensibly, pay little, hold patiently, and still finish with nothing, because the failure happened in the plumbing rather than in the market.

You are more concentrated than you think

Here is the part that makes this week its own rather than a footnote to week 3, and it is the reason it comes after week 8 rather than before.

Suppose you did everything the course asked. You split $10,980 across five positions, in five different businesses, in different industries, deliberately chosen to fail for different reasons. In instrument terms you are as diversified as this course can make you.

Now ask where those five positions are held. For almost everyone the answer is one account at one broker. So:

instrument    $2,196 / $10,980 =   20%
counterparty $10,980 / $10,980 =  100%

Twenty percent in any one business, and every dollar of it with one firm. Splitting the instruments did nothing to the second number. And unlike week 8, where the fix at least exists in principle — hold things that move apart — this one is not fixed by any amount of care about what you buy, because it is not a fact about what you buy.

That is the shape of it. Week 3 was one number too large. Week 8 was five numbers that turned out to be one number. This is five numbers that were never separate to begin with, in a dimension nobody was looking at.

Why nobody in this room will experience it

This is the only week on the syllabus that is guaranteed not to happen to you, and it is worth being straight about why.

It is rare. It cannot be committed on purpose — there is no way to deliberately cause your broker to fail. It cannot happen on a paper account, because there is no counterparty. And a $20 real position is far too small to be worth anyone’s fraud or to be affected by anyone’s insolvency.

So it is taught and never assessed, and if the semester hands it to nobody, nobody’s mark changes. It earns its week anyway, for one reason: knowing the category exists changes exactly one thing you do, and that thing is this week’s work.

This week’s work involves no trading

Find out, in writing, who holds what you own.

Bring the document, not a summary of it. Most of the room will be reading it for the first time, eleven weeks after opening a position on the strength of it, which is the finding rather than an embarrassment — nobody reads it, and this course has one week spare to make thirty people read it once.

Teaching team