In 1929 it was margin debt.

In 1987 it was junk bonds and program trading.
In 2000 it was dotcoms.
In 2007?

First let's ask: what moves global markets right now? The answer comes easily: an excess "liquidity" tide that raises all ships. But what is this "liquidity" and where does it come from? Since all money is debt, increased liquidity naturally comes from more debt. When the price of this debt is cheap, i.e. when interest rates are low, more "money" is created and it chases assets like a tomcat in heat. For as long as the cost of money is cheaper than the expected marginal returns, the chase goes on.

At some point asset prices reach such a high level that no matter how cheap borrowing is, there is no significant extra return to be had: diminishing returns set in. But what if the price of money is zero? Then the marginal return on assets bought with such "free" financing is infinity (anything divided by zero results in infinity). But who would be foolish enough to give money away for nothing - except in Dire Straits lyrics? Keep reading, because it seems there are financial alchemists that act as if they have discovered the monetary equivalent of the philosopher's stone and the perpetual motion machine, rolled into one.

For professional money managers competing ferociously for that extra bit of return that passes for genius in their trade, a tad of infinity goes a long way in boosting performance and year-end performance bonuses. So there is, indeed, lots of demand for free money - if you can get it - for as long as markets are going up, of course. Because if the direction is down, the process works in reverse: no one wants to borrow to invest and speculators rush to liquidate positions to repay their loans even if they got them at zero interest - after all, they still have to return the principal amount in the whole.

All right, we have the demand side identified: market speculators of all sorts. So how is the free money supply created? The Japanese yen carry trade has been already identified as one source of ultra cheap money. Indeed, it has been pumping liquidity onto markets for at least five years. But it still carries an interest rate cost, however modest, plus the day-to-day FX volatility that requires constant mark-to-market adjustments. And, crucially, even when you borrow yen at 0.1% the whole process is still a loan, subject to credit qualification criteria. The price may be almost free, but those pesky bankers won't give it you unless you can prove that you have the ability to pay it back. Plus 0.1% interest, of course. The loan also goes in the banks' books and is subject to reserve requirements. Though it comes rather close, this is not the "free" money capable of producing infinite returns and $200 million performance bonuses.

But Credit Default Swaps (CDS) is free money to those that sell them. The premium payment money keeps coming in free and clear for as long as no defaults occur. Money you get against assuming someone else's default risk does not carry a clearly visible x% interest cost per year. What is happening, then, is that the cost of money is shifting away from paying an interest rate and onto risk assumption - and this is where trouble lies.

Let's look at an example:

Ace Diceroll, manager of Hedge Fund ABC, wants to boost his annual return by 1% on the $1 billion he manages - he has to come up with an extra $10 million in bookable profit. He calls Bob Lencalot, the account manager at his friendly prime lender Bank XYZ and arranges to sell as much credit protection as it takes to generate $10 million in income. Bob informs him that he has two choices: he can provide insurance for $3.3 billion of investment grade bonds, or for $435 million of risky "high yield", or "junk", bonds. Ace is mildly perturbed - just a few months ago, when he had also pondered the same strategy, he had to insure "only" $2.4 billion and $312 million respectively. Be that as it may, Ace swallows hard and sells the insurance. Construction of his new ski chalet in Aspen is costing more than originally planned.

Seems lots of "Aces" have been in the market recently for just such free "performance boosts", helping to drive CDS prices down. Indeed, one can see this plainly in the performance of the CDX Index

The CDX index tracks investment grade CDS - the red line is the annual cost of insurance expressed in basis points, i.e. Ace can only get $3.1 per year for insuring $1,000 versus $4.1 last September. The equivalent for insuring junk bonds has gone from $32 per $1,000 to $23.

Bank XYZ is acting both as principal and intermediary here: It is unlikely that Ace's ABC hedge fund could get its own name accepted as an insurer in the market, so it sells the CDS to XYZ who either keeps it in its own books or turns around and markets it under its own name and credit rating. Why would XYZ do that? Two reasons: a) it makes an immediate fat spread profit on the transaction and b) it gets lots of Ace's other business such as commissions on securities' transactions, margin loans, custody fees, etc. (thus the term "prime bank"). Plus, all this stuff is strictly "off balance sheet", so it does not swell the bank's asset/liability ledger.

The money ABC receives from selling the CDS doesn't just sit around collecting dust, of course. Ace will put it to use to further enhance performance - if he gets just a 10% return on the CDS money he will boost his return by another 0.1% for a total 1.1%. If he really loves a crapshoot (and most hedge fund managers do), he will further leverage this money, say 2 times, and boost his performance even further, to 1.3%. Hey, every little bit helps when it comes to ski chalets - have you priced one lately?

And when you add it all up it is certainly not "a little bit". A notional $35 trillion is outstanding in CDS (see previous post): 75% of them cover investment grade bonds (generating 0.3%) and 25% cover junk (generating 2.3%) resulting in a blended insurance income of approx. 0.80%, or a stunning $280 billion annually (before potential credit losses, of course). Just five years ago this source of "free money" liquidity did not provide more than approx. $5-8 billion as CDS's were essentially non-existent. If this money is leveraged just once, the total purchase power impact on markets doubles to $560 billion. More debt piled on top of "free" money that came about through leveraging other people's debt!

Think about it: people totally unrelated to GE - for example - can make a bundle by leveraging its AAA credit rating to the sky! Can you blame financial alchemists for believing in magic?

OK, then, let's assume this is where CDS supply is coming from these days. But, who is buying the stuff that ABC is selling through XYZ? - who is the ultimate provider of all this "free liquidity" that buys protection to the tune of $280 billion?

a) Hedgers
b) Speculators

Hedgers: In a previous post I explained that the only way a rational hedger (i.e. a bond holder) would buy CDS is if it were mis-priced vs. real risk, i.e. if he could buy risk protection on the cheap. For example, if a pension fund holds a BBB rated bond he would gladly buy insurance from a AA rated CDS issuer if it cost significantly less than the difference in yield between the two. If the BBB bond yields 6% right now and the AA's bonds yield 5%, the holder of the BBB bond would gladly buy CDS insurance at, say, 0.50% per annum. He would have effectively turned his BBB bond into a synthetic AA bond and still collect 5.50% (6.00% - 0.50%), i.e. more than a straight AA bond. Indeed, this is happening right now: the rush to sell CDS and collect "free money" has resulted in just such seemingly impossible imbalances (though not at these hypothetical levels, of course).

Speculators: sell anything for cheap and smart money will eventually show up. Quietly and slowly. And patiently...

Where's the bottom line? Where it always is, of course: RISK. Just think of it this way: if anything happens to raise risk premiums by 100 basis points the mark to market loss for CDS alone will be $350 billion. Throw in leverage and the impact across all markets could easily double or even triple. If there are major defaults, too, the potential losses could be in the trillions, since CDS issuance is so concentrated amongst just a few hundred bond issuers: GFI, the major interdealer broker, reports that it regularly sees just 650 unique names in the CDS market. Do the math: $35 trillion divided by 650 means there is an average of $54 billion in CDS outstanding for every issuer. Averages being what they are, it is certain that some names have hundreds of billions of CDS outstanding against their credit.

The double bottom line: Such derivatives do not serve to moderate risk but to concentrate it, both at the original credit default level and across other, unrelated markets, since "free" money is undoubtedly finding its way into them as well.

Stay tuned - Tomorrow I'll try to provide some perspective on how I think the Fed is viewing all of this and what it means for interest rates going forward.