The notional of outstanding derivatives contracts is meaningless.
For example consider the following position:
Bank A pays fixed 5% $ 500 trillion vs 3M Libor for 10y
Bank B pays fixed 5.01% $ 500 trillion vs 3M Libor for 10y
Outstanding notional 500 trillion.
Time to panic? No, the position has a finite PV but zero market and counterparty risk.
90% of the deals in most derivatives book are offsetting in this way.
The problem arises because there are far more than two players involved and far more than two instruments. One bank’s position may be hedged with dozens of instruments, of differing types, with many institutions which are not involved in the deal that it wishes to hedge…and vice versa.
The problem manifests itself if any major institution fails. All the instruments held or offered by that institution become stuck. Assets held by the bank become impossible to reach, and the obligations of that institution may never be fulfilled. Even a short term inability to extract funds from that bank may impact another bank…and so on…in a classic domino effect.
Even if the total of trillions represents the same money all - theoretically - cancelling itself out, the practical complications are huge.
That’s interesting. Clearly bank B owes bank A money.
In other words Bank B’s deal with Bank A has a negative NPV from Bank Bs point of view (Work it out - its about
-$350 Billion assuming a duration of 7 for a 10y swap).
The details of what happens depends on how Bank A is wound down. Most likely the swap book probably moves to another institution.
If not, Bank Bs liabilitity to Bank A will be offset against Bank As liabilities to Bank B. For example, if Bank B owns some of Bank A’s bonds, it can offset bond notional less recovery value against its swap liability.
In the real world exposures are reduced further by collateral agreements.
(1) derivatives are a zero sum game. across the entire market the total derivatives exposure nets to zero.
(2) real exposures even in an individual derivatives book are usually many orders of magnitude smaller than notional values.
Derivatives are toxic, just not nearly as toxic as claimed by daft commentary based on outstanding notional amounts.
They are a zero sum game but a real issue out there is the value put on some of these bets, especially exotic products…
For example, I cld sell you some Conditional Knock-Out Variance Swap (just made that up, but am sure it exists!).
Then we both put the trade in our books. You book a profit and I book a profit…Surely we cant both simultaneously make money?!?
Well it happens…because we are using either different models to “mark” the trade or the “real” market value is so opaque we both become optimistic in they way we mark towards the side of the bet we have on.
Your model puts the value at 11, mine puts it at 9…We trade with each other at 10 and voila we both make an easy 1 !!!
This happens all the time…In fact when Variance Swaps first came to market it soon became apparent that 2 guys in 2 separate banks were trading with each other all day and both of them were booking a massive profit.
So, one has to wonder…by the time all these trades expire how much phantom wealth will evaporate from banks balance sheets?
I agree in theory
Some questions though.
How much commission was generated on these 0 sum deals and who pays it?
What’s the cost of unwinding them?
Given the massive sums involved and the difficulty banks are having in reconciling the numbers what was the transaction costs of this massive zero sum derivatives market and have those costs become so large as to turn the 0 sum game into something else?
True, but don’t exaggerate. Model wars tend to be short-lived and end in some hothead getting fired.
As regards exotics - they are typically traded with clueless clients at obscene margins. Banks
are supposed to hold reserves against residual model risk. No doubt these reserves have already been released to cover
up losses in other areas. Yep, I said there are major problems brewing there.
Btw, don’t knock model wars. There is nothing more enjoyable in derivatives business than a good model war.
