Excerpt from F.I.A.S.C.O.: Blood in the Water on Wall Street by Frank Partnoy. 1997.
[1/5] "The idea behind this special trade, presented in its most primitive form, is this: Suppose you pay $100 for a pot of gold. Half the pot is real gold worth $90 and the other half is fool's gold worth $10. If tomorrow the two halves still are worth $90 and $10, and you sell the real gold for $90, can you claim a profit? The answer must be no, right?"
"Wrong. At least if you're in Tokyo. Suppose you buy the pot of gold and claim that each half cost <i>'on average'</i> $50. Tomorrow, you sell the real gold for $90. Presto, you have booked a $40 profit. How? The real gold didn't cost you $90; it cost you $50 - <i>'on average'</i> - and you sold it the next day for $90, a gain of $40. Of course, you didn't really make a profit, but by Tokyo accounting standards, it was close enough, and such trading was quite common."
Obviously, if you recognized a $40 profit on the real gold, then eventually even under Tokyo accounting rules you would be required to recognize a $40 loss on the fool's gold. The key word is <i>'eventually.'</i> If you could avoid telling anyone that the fool's gold was actually worth only $10 when you were claiming it still was worth it's <i>'average'</i> cost of $50, you might not have to recognize that loss for a long time. In Japan that long time could be your entire career. In fact, if you could retain the fool's gold for long enough - say, until you retired from the company - who cared? At that point the loss would become somebody else's problem."
"I was amazed by this idea. It was financial alchemy at its best. The only problem with the pot of gold example is that it's too easy to distinguish fool's gold from real gold. The accounting and regulatory authorities, even in Japan, were sophisticated enough to catch an investor using this simple trick because they would be able to discover that the two halves were really worth $90 and $10, respectively, not $50 <i>'average.'</i> In other words, even they could tell what was fool's gold."
"Consequently, an investor needed to construct a more complicated way of doing the same thing, using <i>'halves'</i> so complex that accountants and regulators couldn't easily discover their actual worth. For this higher level of sophistication, Japanese buyers turned to Morgan Stanley, and to derivatives."
"As with many derivatives deals, Morgan Stanley employees debate who gets credit for the complex idea that allowed Japanese institutions to recognize billions of dollars of false profits. I certainly don't claim any credit for it, nor do I want credit. The idea predated my arrival at the firm, anyway, so it can't be blamed on me. Whoever invented the trade, one thing is clear: They gave it a horrible acronym." [1/5]
[2/5] "Not many people at Morgan Stanley, even in DPG, know about the AMIT's <i>(American Mortgage Investment Trust)</i>. The firm completed its first AMIT on February 14, 1992. That $100 million trade generated profits of $2 million, with minimal work. Numerous clients wanted similarly quick and easy profits, and by the end of the year the group had closed the 14th AMIT, although admittedly a few numbers had been skipped."
"It's also time to give you the inside scoop on the AMIT - which DPG <i>(Derivatives Product Group)</i> salesmen often called the 'Shamit' or 'Scamit' - described in the same way it was relayed to me. It's no different in principle from combining real and fool's gold. But I should say at the outset that these things are not easy to understand - one reason the AMIT succeeded was that its complexity masked its true nature. . . Whoever said one of the greatest frauds in the history of investment banking would be easy?"
"The MX trade was structured as a typical AMIT. Believe it or not, the recipe for an AMIT trade begins at home, with home mortgages. In fact, some of the billions of dollars actually paid to Japanese investors through Morgan Stanley's various AMIT trades, including MX, may originally have been from a check you wrote to make your mortgage payment."
"AMIT trades use mortgage derivatives, which split up the pieces of mortgage interest and principal payments. The ability to split mortgage payments into pieces was the spark that generated the idea for the AMIT trades. Remember, the goal of the AMIT trade is to create two "pieces" that appear to be the same size but actually aren't. In the example I described earlier, the real gold was worth $90, the fool's gold was worth $10, and each half was worth <i>'on average'</i> $50. In that example, the $90 of real gold would be called the "premium" piece, and the $10 of fool's gold would be called the "discount" piece. The AMIT trade requires both a premium and a discount piece."
"For the MX trade, as well as for many AMIT trades, the more valuable premium piece, the one worth $90 in my example, is called an IOette. The less valuable discount piece, the one worth $10 in my example, is called a zero coupon bond, also known as a Zero or a Strip. The Zero is the same type of bond that Joseph Jett allegedly lost $350 million on at Kidder, Peabody, that the Strips traders at First Boston supposedly made $50 million trading one year, and that we used in FP Trust. The zero coupon bond is simple. It's just a single payment, made by the U.S. government, at some specified future date."[2/5]
[3/5] "In contrast, the IOette is much more complicated. An IOette is a type of collateralized mortgage obligation, or CMO. CMO's may sound complex, but they are actually very simple. When you make your home mortgage payment, your check typically is passed through to a federal agency, such as Fannie Mae or Freddie Mac, that, as part of its day-to-day business, receives mortgage payments from various homeowners and collects them in mortgage pools. These pools form the basis for various mortgage securities, including mortgage derivatives such as CMO's, and your mortgage payments may be pooled through these securities and passed throughout the world. CMO's are simply different kinds of strips of home mortgages. They come in various shapes and sizes, of which the IOette is one of the most unusual."
"The most common way to strip apart mortgages is by interest and principal: the "interest only" pieces (IO's) have a claim on the home owner's payments of interest, and the "principal only" pieces (PO's) have a claim on only the payments of principal. IO's and PO's are the two most basic types of CMO's, and every monthly mortgage payment you make is part interest and part principal, that can be thought of as part IO and part PO. There are many more complex CMO derivatives, with exotic names like PAC's, TAC's, inverse floaters, and Z bonds."
"The special difficulty with all mortgages, including CMO's, is deciding what portion of the mortgage pool's principal will be prepaid by a given date. If interest rates decline by several percent, you might decide to refinance your current mortgage by prepaying the principal, then taking out a new mortgage at a lower rate. However, if every owner in a pool prepaid their mortgages, certain CMO owners would be devastated. For example, an owner of IO's wouldn't receive any more interest payments once the mortgages in a pool were prepaid, so his IO's would become worthless."
"What makes CMO's especially dangerous is that although they're extremely risky, they can appear quite safe. One deceiving and dangerous aspect of CMO's is their credit rating: AAA. Because payments on most CMO's are guaranteed by an agency of the federal government, the companies who rate bonds' credit quality, Standard and Poor's and Moody's, assign most CMO's their highest credit rating. But this AAA rating is misleading. Although an agency of the federal government is unlikely to default on its guarantee, default is only one of the risks associated with CMO's. Investors in CMO's can and do lose money for other reasons, including the risk of repayment of principal. These additional risks are not captured by the credit rating." [3/5]
[4/5] "But the most important yield curve to derivatives salesmen is one you won't find in the financial pages - the forward yield curve, or "forward curve." Actually, there are many forward curves, but all are based on the same idea. A forward curve is like a time machine: it tells you what the market is "predicting" the current yield curve will look like at some point forward in time. Embedded in the current yield curve are forward curves for various time forward times. For example, the "one-year forward curve" tells you what the current yield curve is predicting the same curve will look like in one year. The "two-year forward curve" tells what the current yield curve is predicting the same curve will look like in two years."
"The yield curve isn't really predicting changes in the way an astrologer or palm reader might, and as a time machine, a forward curve is not very accurate. If it were, derivatives traders would be even richer than they already are. Instead, the yield curve's predictions arise, almost like magic but not quite, out of arbitrage - so-called riskless trades to capture price differences between bonds - in an active, liquid bond market."
"Forward curves are an incredibly powerful and important concept in derivatives trading. If you don't understand forward curves and you work at an investment bank, the derivatives salesmen and traders are probably laughing at you. If you don't understand forward curves and you're a normal person investing in anything other than stocks and bank CD's, you're probably being screwed by someone who does understand them."
"The downside risk made longer maturity PERLS (Principal Exchange Rate Linked Security) especially attractive to sell. Selling a five-year PERLS to a widow or orphan meant you didn't have to worry about the repayment of principal for five years - an entire career on Wall Street - and even then there was a decent chance the buyer would have bet correctly and made money. Not even a widow or orphan will complain about receiving $200 instead of $100 at maturity."
"The beauty of this structured note was that it effectively hid the derivatives trade from public view. If you were a fund manager and you bought the sterling structured note, you wouldn't have to tell your boss or a regulatory agency, "Hey guys, look what I just bought - a Three Year Currency Protected Sterling Inverse Floater." The packaging hid all that. The note, too, looked innocuous, this time like a three-year bond issued by the AAA-rated Federal Home Loan Mortgage Corporation, popularly known as Freddie Mac. A buyer was like an underage kid who pays the local bum to buy booze for him, then pours the booze in a Coke can so his parents won't notice. Except in this case the local bum, an investment bank, was receiving more than just a few dollars for the effort." [4/5]
[5/5] "Freddie Mac, headquartered in McLean, Virginia, was the perfect safety package. Freddie Mac is a U.S. government agency that guarantees residential mortgages and since early 1993 has issued complex derivatives, too. The agency simply uses the derivatives the investment bank creates to borrow at a slightly lower cost. The bank structures the note issuance and parcels out the risk; Freddie Mac provides the credit guarantee. So never mind that the coupon payment is 16.050% - [2 x Two (2) Year Pound Sterling Swap Rate], payable semi-annually for three years in U.S. dollars. It's Freddie Mac, for God's sake. It's AAA and implicitly backed by the U.S. Treasury. Anyone can buy this."
"Why might unsophisticated state and local governments want to bet that the British two-year swap rates would decline? It's hundreds of miles from most of these buyers to the Mississippi River, thousands to the Thames. Why might the leaders of a community college, or a city council on this side of the Atlantic think British swap rates would decline?"
"Well, for one thing the bank's researchers said they would. And guess which derivatives trade those unbiased researchers were recommending that you should buy to take advantage of anticipated lower rates? Research suggested inflation was declining in Britain and that such a decline would lead to lower interest rates. At the same time, the sterling yield curve was very steep, which meant the 'magnified' forward curve also was even higher, and steeper, predicting that rates would increase. This seemed to be a perfect opportunity to bet against the forward curve. Could you afford not to?"
"The researchers at an investment bank had an amazing capacity for generating predictions that contradicted those of the forward curve. And every time they did, the bank's derivatives group would create a derivatives trade supported by the research. Or maybe it was the other way around." [5/5]
Sept. 9 (Bloomberg) -- When the history is written on the collapse of Fannie Mae and Freddie Mac, it will go down in the annals of corporate scandals as one of the greatest accounting scams committed in broad daylight.
All anyone had to do to know the government-guaranteed mortgage financiers were insolvent was read their financial statements. You didn't need a trained professional eye to discern this open secret, only a skeptical one.
Just last month, Fannie and Freddie said their regulatory capital was $47 billion and $37.1 billion, respectively, as of June 30. The Treasury Department now says it may have to inject as much as $200 billion of capital into the two companies. Nothing much changed at the companies in that span. They just couldn't get the government to keep up the ruse any longer.
To have believed those capital figures, you first needed to accept two key assertions by the companies. The first was that the mortgage-market meltdown was a temporary blip. The second was that Fannie and Freddie both would be wildly profitable for decades to come, once the blip was over.
These weren't the only fairy tales Fannie and Freddie told. They're just the ones that had the biggest impact on their calculations of regulatory capital. Had Fannie and Freddie ever backed off these predictions, they would have been officially insolvent, even under the government's lax standards.
Until late last week, though, nobody with any authority was willing to call them on it. That's why Fannie and Freddie were able to avoid a government takeover for so long.
By the time the government moved in, Freddie had accumulated $34.3 billion of paper losses on mortgage-related securities that it excluded from its calculations of regulatory capital. All Freddie had to do was say the losses were "temporary," and they could be kept out of the company's capital figure. It didn't seem to matter how ridiculous the claim was.
Fannie played the same game. As of June 30, it had $11.2 billion of supposedly temporary losses on mortgage-related securities, which it excluded from its calculations of core capital, as the government calls it. (A better name would be "kore kapital," like the imitation krab sticks on a sushi bar menu.)
The so-called temporary losses had the warped effect of inflating a line item on both companies' balance sheets called deferred-tax assets. The bigger the companies' losses got, the more these tax assets grew, based on the premise that someday the companies would be able to use the losses to offset future income-tax bills.
The catch is that if a company doesn't expect to have enough profits to use these assets, it's supposed to record a valuation allowance on its balance sheet to reduce their size. Freddie and Fannie didn't let this requirement get in the way. They never set up any allowances.
So, as of June 30, when Freddie said it had deferred-tax assets of $18.4 billion, it supposedly envisioned about $50 billion of future taxable income that, in its judgment, would probably materialize in the face of the worst financial crisis since the Great Depression.
Ditto for Fannie. It claimed to have $20.6 billion of deferred-tax assets as of June 30.
The companies' delusions didn't stop there. In their financial disclosures, both companies represented that the fair- market value of their tax assets was billions of dollars more than what they were allowed to show under generally accepted accounting principles.
All of this was malarkey, of course. And it was all disclosed, even if the companies' explanations weren't always clear.
There were lots of gatekeepers who could have stopped this sooner and chose not to. Freddie and Fannie had boards with outside directors and audit committees, though the evidence that they did their jobs is scant.
Freddie's auditor, PricewaterhouseCoopers LLP, could have stopped it, and didn't. The same is true of Fannie's auditor, Deloitte & Touche LLP.
The Securities and Exchange Commission's chairman, Christopher Cox, had other priorities. He was too busy this summer trying to prop up the companies' stock prices by chasing short sellers away from their shares.
James Lockhart, the director of the Federal Housing Finance Agency, kept telling the public this summer that Fannie and Freddie were adequately capitalized. He must have known this was a joke, assuming he had bothered to read their quarterly reports.
All the while, Ben Bernanke at the Federal Reserve and Hank Paulson at Treasury offered the same warm assurances about the companies' capital. They surely knew better, too.
Don't cry for investors who lost money on the companies' stocks either. If they didn't do any research or didn't understand what they owned, the people they should blame are themselves.
The reality is that investors should withhold their faith in the government officials who regulate our financial markets. That's not cynicism. In the parlance of securities law, it's a risk-factor disclosure.
The moral of this story: You're on your own, folks, and there's more where this came from.
(Jonathan Weil is a Bloomberg News columnist. The opinions expressed are his own.)
Last Updated: September 9, 2008 03:52 EDT
