One such figure came up yesterday. The total ‘value’ of global derivatives – financial instruments which are priced on the back of the underlying assets that they track – has now reached a breathtaking $596 trillion, after a mammoth rise over the previous twelve months.

That started the warning lights flashing…

So what, apart from containing more noughts than a normal human being can cope with, is this titanic number all about?

Let’s start by putting it into context. We can do this by checking out what the world actually made last year. The overall value of goods and services produced is measured by Gross Domestic Product (GDP). And for 2007, GDP for planet earth was reckoned by the International Monetary Fund to be just shy of $65 trillion. No less an organization than the CIA has come up with a similar estimate, at £65.8 trillion, so it must be about right.

So when the Bank for International Settlements (BIS) tells us that last year the total derivatives market grew by 44%, its fastest pace since the Basel-based bank started keeping records just over ten years ago, up go the antennae straightaway. And when that figure of $596 trillion crosses the radar screen, equivalent to more than nine times world GDP, the numbers are looking quite scary.

The money at risk is equivalent to a quarter of world output

Of course, the $596 trillion is a ‘notional’ amount. It’s the nominal value of all the underlying assets against which bets have been placed. But the actual amount of ‘real’ money at risk is still a massive $15 trillion, equal to almost a quarter of world output.

And within the individual areas there’s one even more eye-catching statistic. The value of contracts in credit default swaps (CDS) – a form of market insurance that investors can buy to protect themselves against corporate bond defaults – more than quadrupled last year to $2 trillion, covering a notional $58 trillion of loan debt.

The very size of all these numbers is just about enough to give the jitters to anyone, on the basis that when things can go wrong, they probably will.

When I wrote on this subject before, one respondent claimed that the topline numbers aren’t important because derivative markets are beautifully balanced. His theory was that if every derivatives position were hedging a risk relating to a specific transaction or asset, then derivatives would actually stabilise the world economy. All those noughts would be good news.

Sounds a bit too good to be true. And there are three reasons to be sceptical about this optimistic line of thinking.

Three reasons to be worried

Firstly, what we can call knowledge risk. That’s when derivatives players don’t know what they’re getting into.

A story on Bloomberg at the end of April summed this up pretty well. The chief finance officer of an Indian company was persuaded by his bank to start dabbling in the currency derivatives market. Although the CFO explained to the bankers that he didn’t understand how these products work, apparently they chauffeured him round and bombarded him with charts showing how his company could make a profit with a zero investment.

Too good to be true? Clearly it was. Three months later, two of the contracts had turned sour, incurring losses of $1.5 million and prompting the bank to issue a bankruptcy notice to recover the cash. Meanwhile, our poor CFO had no idea that these derivative bets could go so wrong. But he’s not alone. Indian companies could lose up to $4bn on derivatives, according to Hong Kong-based brokerage CLSA Ltd. Naïvety? Maybe. But we’re all good at repenting at leisure.

Which brings us onto the next potential problem, counterparty risk. That’s when the deal you’ve just done comes unstuck because the people on the other side of the trade can’t settle their side of the deal. A bit like backing the Derby winner, then finding the bookie can’t pay up because he’s run out of money.

Indian banks may lose up to $400m if they can’t enforce derivatives contracts they’ve set up with smaller companies, says CLSA. This is because 10% of these smaller companies may renege on their agreements because they haven’t the cash to settle the deals.

And this is just one country. BNP Paribas analyst Andera Cicione believes that total world CDS losses could hit $150bn. As the CDS market is unregulated, there are no public records showing whether sellers have the assets to pay out if a bond defaults. George Soros himself has warned this week that CDS counterparty risk is “a Damoclean sword waiting to fall.”

What’s worse – and here we come to the third problem – some buyers have now found out that the derivatives they’ve bought haven’t matched up to “what it said on the tin”.

The ratings agency Moody’s has just admitted awarding incorrect ratings to $4bn worth of debt instruments because of a bug in its computer models. Some ultra-complex derivative products, known as “constant proportion debt obligations” and thought up at the height of the credit bubble, incorrectly received over-optimistic triple A – i.e. top notch – ratings. And it took Moody’s nearly a year to find the problem.

As the derivatives market gets bigger and bigger, stories like these only make us ask: do the people who play around in it really know what they’re doing?

Turning to the wider markets:

The first true credit default swap was carried out as late as 1995, although various types of credit protection derivatives existed earlier. Its structure is similar to an ordinary interest rate or currency swap transaction, and the CDS market is covered by the International Swaps and Derivatives Association Inc.

Under a CDS, a bank originates loan to a company. A second bank (or other financial institution) can agree to cover the credit risk for the loan, by agreeing to make payment to originating bank if the company defaults on the original loan. The originating bank pays a small insurance premium to the second bank for assuming the risk of the loan.

Typically, payments under a CDS would only be triggered by the company’s failure to pay interest or principal on its debts due to bankruptcy or some other severe liquidity issue. But there are a host of intermediate or special cases that will doubtless provoke lawsuits when something goes wrong (CDS being a new market, it is by no means “recession-proof”).

Credit default swaps were sold to the world as hedging transactions. Investors were told that they were simply transfers of risk, so that banks that made loans could transfer credit risks to insurance companies, which did not make loans directly, or to foreign banks that could not easily make loans in the U.S. market.

And if an originating bank sells its loan exposure only once, and sells it to a financial firm of undoubtedly solid credit, the CDS does indeed act as a hedge for the originating bank; it transfers the company’s credit risk from the bank to the financial firm that bought its CDS.

But the product did not work as advertised.

Enter the Traders

Salesmen and traders took over, and expanded the volume far beyond what was required for hedging.

After all, bonuses depend on the volume of business. Therefore, bank traders sold the credit risk of a loan not just once, but as many as 10 times. And they sold it not to solid banks and insurance companies, but to three solid banks, one solid insurance company, three dodgy brokers and three hedge funds. Then the traders went out and sold other CDS products that were not even related to actual loans on the books, but to imaginary indices of credit quality in the “widget” industry.

The credit risk of the system was hugely multiplied.

Instead of one $10 million credit risk loan, there are now ten $10 million credit risks on just one loan.

• Three on solid banks – but will they stay solid?
• One on a solid insurance company – probably OK.
• Three on dodgy brokers – who knows?
• And three on hedge funds – probably not OK in a real downturn.

The total credit risk in the system has been increased from the original $10 million loan to somewhere between $160 million to 200 million, depending on whether the banks and insurance company are financially solid.

Of course, a lot of those credit risks offset each other, so that if the company that took the loan goes bust, the only risk to the bank that sold all those CDS is to the profits it expected to make. But since it probably hedged those positions against others, if the company does go bust, and dodgy brokers and hedge funds stop paying up, the total losses in the system from that company’s credit risk are likely to be a substantial multiple of the original $10 million loan.

But please don’t think I was exaggerating when I said as many as 10 credit default swaps got sold for each loan.

The U.S. commercial loan market is worth about $5 trillion, yet the volume of CDS outstanding is currently no less than $50 trillion. In other words, a huge number of traders, salesmen and quants have been making money off this product, without any real “hedging” rationale at all.

And it all worked fine while the volume of defaults remained low, which is why the market expanded from $2 trillion to $50 trillion between 2000 and 2007.