Why do "quadrillions in derivatives" headlines mislead?
Every few years an article announces that the world sits on a derivatives time bomb of hundreds of trillions — sometimes a quadrillion — of dollars, a figure many times global GDP. The number is usually accurate. The conclusion drawn from it is usually wrong, and knowing exactly why is a genuine analytical skill.
Three different numbers, routinely confused
1. Notional amount. The reference quantity used to scale the payments. In the swap from the last lesson, $100 million — an amount that never moves.
2. Gross market value. What the contracts are actually worth today: the sum of what each one would cost to replace at current prices. If rates have moved so the swap is worth 1.5% of its notional, that is $1.5 million on a $100 million notional.
3. Gross credit exposure. What is genuinely at risk after netting — the amount two parties would owe each other once offsetting contracts between them are cancelled down to a single figure, before any collateral is counted.
Each step down is roughly an order of magnitude. Recent Bank for International Settlements surveys have put OTC derivatives notional outstanding in the region of $600–750 trillion, gross market value at around $20 trillion — roughly 3% of notional — and gross credit exposure after legally enforceable netting at around $3 trillion, under 0.5% of the headline figure.
So the "quadrillion" is measuring the size of the yardstick, not the size of the bet.
Work through why the gap is so large
A bank has 500 interest rate swaps facing another bank. Total notional: $50 billion. Of those, 300 are currently in its favour and worth $800 million; 200 are against it and worth −$760 million.
- Headline notional: $50,000 million
- Gross market value in its favour: $800 million (1.6% of notional)
- Net exposure after the master netting agreement: $40 million (0.08% of notional)
And then collateral is posted against most of that $40 million, shrinking it again. The $50 billion headline overstates the actual risk by a factor of over a thousand.
A familiar analogy: the total sum insured across a country's home insurance policies is a colossal number, vastly larger than the industry's capital. Nobody reads it as an impending catastrophe, because everyone understands it is a scaling figure, not an expected loss. Notional is the same kind of number.
Why notional still is not meaningless
Do not overcorrect. Notional determines sensitivity, and that is the whole reason leverage lives here. A 1% move against a $100 million notional is $1 million of real cash — on a contract that required no purchase price. Notional tells you how violently a position responds to the underlying, which is exactly the number you want when asking how much can go wrong in a week.
The correct reading is therefore: notional measures exposure to movement; market value measures what is owed today; credit exposure measures what is at risk if someone fails. Three questions, three numbers. A headline that uses the first to answer the third is not lying — it is answering the wrong question loudly.
In the data
The table below is the whole market in one kind of derivative, credit default swaps, as the CFTC, the US derivatives regulator, counts it each week: gross notional outstanding, split into trades that sit behind a clearing house and trades that do not.
Add the two lines and you have a figure in the trillions of dollars. That is the number a "size of the market" headline quotes, and it is notional: no market value and no exposure after netting sits beside it, so nothing in the table tells you how much money is actually at risk.
Try it now
- Find the most recent BIS OTC derivatives statistics and note the three published figures: notional outstanding, gross market value, and gross credit exposure.
- Compute market value as a percentage of notional, and credit exposure as a percentage of notional. Keep both numbers; they are the antidote to a whole category of scary headline.
- Now do the same arithmetic on something you can see. Read the current price off the chart below, take a futures-style position of 1,000 units, and compute the notional. Then Measure a 1% adverse move and convert it to cash. The ratio of that cash to a plausible margin deposit is the subject of the next lesson.