There is a story about a science professor giving a public lecture on the solar system. An elderly lady interrupts to claim that, contrary to his assertions about gravity, the world travels through the universe on the back of a giant turtle. “But what supports the turtle?” retorts the professor. “You can’t trick me,” says the woman. “It’s turtles all the way down.”
The American financial system has started to look as logical as “turtles all the way down” this week. Only six months ago, politicians were counting on Fannie Mae and Freddie Mac, the country’s mortgage giants, to bolster the housing market by buying more mortgages. Now the rescuers themselves have needed rescuing.
After a headlong plunge in the two firms’ share prices (see chart 1), Hank Paulson, the treasury secretary, felt obliged to make an emergency announcement on July 13th. He will seek Congress’s approval for extending the Treasury’s credit lines to the pair and even buying their shares if necessary. Separately, the Federal Reserve said Fannie and Freddie could get financing at its discount window, a privilege previously available only to banks.
The absurdity of this situation was highlighted by the way the discount window works. The Fed does not just accept any old assets as collateral; it wants assets that are “safe”. As well as Treasury bonds, it is willing to accept paper issued by “government-sponsored enterprises” (GSEs). But the two most prominent GSEs are Fannie Mae and Freddie Mac. In theory, therefore, the two companies could issue their own debt and exchange it for loans from the government—the equivalent of having access to the printing press.
Absurd or not, the rescue package notched up one immediate success. Freddie Mac was able to raise $3 billion in short-term finance on July 14th. But the deal did little to help the share price of either company or indeed of banks, where sentiment was dented by the collapse of IndyMac, a mortgage lender. The next day Moody’s, a rating agency, downgraded both the financial strength and the preferred stock of Fannie and Freddie, making a capital-raising exercise look even more difficult. As a sign of its concern, the Securities and Exchange Commission, America’s leading financial regulator, weighed in with rules restricting the short-selling of shares in Fannie and Freddie.
The whole affair has raised questions about the giant twins. They were set up to provide liquidity for the housing market by buying mortgages from the banks. They repackaged these loans and used them as collateral for bonds called mortgage-backed securities; they guaranteed buyers of those securities against default.
This model was based on the ability of investors to see through one illusion and boosted by their willingness to believe in another. The illusion that investors saw through was the official line that debt issued by Fannie and Freddie was not backed by the government. No one believed this. Investors felt that the government would not let Fannie and Freddie fail; they have just been proved right.
The belief in the implicit government guarantee allowed the pair to borrow cheaply. This made their model work. They could earn more on the mortgages they bought than they paid to raise money in the markets. Had Fannie and Freddie been hedge funds, this strategy would have been known as a “carry trade”.
It also allowed Fannie and Freddie to operate with tiny amounts of capital. The two groups had core capital (as defined by their regulator) of $83.2 billion at the end of 2007 (see chart 2); this supported around $5.2 trillion of debt and guarantees, a gearing ratio of 65 to one. According to CreditSights, a research group, Fannie and Freddie were counterparties in $2.3 trillion-worth of derivative transactions, related to their hedging activities.
There is no way a private bank would be allowed to have such a highly geared balance sheet, nor would it qualify for the highest AAA credit rating. In a speech to Congress in 2004, Alan Greenspan, then the chairman of the Fed, said: “Without the expectation of government support in a crisis, such leverage would not be possible without a significantly higher cost of debt.” The likelihood of “extraordinary support” from the government is cited by Standard & Poor’s (S&P), a rating agency, in explaining its rating of the firms’ debt.
The illusion investors fell for was the idea that American house prices would not fall across the country. This bolstered the twins’ creditworthiness. Although the two organisations have suffered from regional busts in the past, house prices have not fallen nationally on an annual basis since Fannie was founded in 1938.
Investors have got quite a bit of protection against a housing bust because of the type of deals that Fannie and Freddie guaranteed. The duo focused on mortgages to borrowers with good credit scores and the wherewithal to put down a deposit. This was not subprime lending. Howard Shapiro, an analyst at Fox-Pitt, an investment bank, says the pair’s average loan-to-value ratio at the end of 2007 was 68%; in other words, they could survive a 30% fall in house prices. So far, declared losses on their core portfolios have indeed been small by the standards of many others; in 2008, they are likely to be between 0.1% and 0.2% of assets, according to S&P.
Of course, this strategy only raises another question. Why does America need government-sponsored bodies to back the type of mortgages that were most likely to be repaid? It looks as if their core business is a solution to a non-existent problem.
However, Fannie and Freddie did not stick to their knitting. In the late 1990s they moved heavily into another area: buying mortgage-backed securities issued by others (see chart 3). Again, this was a version of the carry trade: they used their cheap financing to buy higher-yielding assets. In 1998 Freddie owned $25 billion of other securities, according to a report by its regulator, the Office of Federal Housing Enterprise Oversight (OFHEO); by the end of 2007 it had $267 billion. Fannie’s outside portfolio grew from $18.5 billion in 1997 to $127.8 billion at the end of 2007. Although they tended to buy AAA-rated paper, that designation is not as reliable as it used to be, as the credit crunch has shown.
Sometimes the mortgage companies were buying each other’s debt: turtles propping each other up. Although this boosted short-term profits, it did not seem to be part of the duo’s original mission. As Mr Greenspan remarked, these purchases “do not appear needed to supply mortgage market liquidity or to enhance capital markets in the United States”.
Joshua Rosner, an analyst at Graham Fisher, a research firm, who was one of the first to identify the problems in the mortgage market in early 2007, reckons Fannie and Freddie were buying 50% of all “private-label” mortgage-backed securities in some years—that is, those issued by conventional mortgage lenders. This left them exposed to the very subprime assets they were meant to avoid. Although that exposure was small compared with their portfolios, it could have a big impact because they have so little equity as a cushion.
Both companies make a distinction between losses on trading assets (which they take as a hit against profits) and on “available-for-sale” securities which they hold for the longer term and disregard, if they think the losses are temporary. At the end of 2007, according to OFHEO, Fannie had pre-tax losses of this type of $4.8 billion; Freddie’s amounted to $15 billion.
The companies have also been unwilling to accept the pain of market prices in acknowledging delinquent loans. When borrowers fail to keep up payments on mortgages in the pool that supports asset-backed loans, Fannie and Freddie must buy back the loan. But that requires an immediate write-off at a time when the market prices of asset-backed loans are depressed. Instead, the twins sometimes pay the interest into the pool to keep the loans afloat. In Mr Rosner’s view, this merely pushes the losses into the future.
Adding to the complexity is the need for both Fannie and Freddie to insure their portfolios against interest-rate risk—in particular, the danger that borrowers may pay back their loans early, if interest rates fall, leaving the companies with money to reinvest at a lower rate. This risk caused the duo to take huge positions in the derivatives market, and was at the centre of an accounting scandal earlier this decade.
In addition, Fannie and Freddie have bought insurance against borrower defaults when the homebuyer lacks a 20% deposit. But the finances of the mortgage insurers do not look that healthy, which may mean the risk ends up back with the siblings. Just as the rescuers need rescuing, so the insurers may need insuring.
None of these practices seemed to dent the confidence of OFHEO in its charges. The regulator said as recently as July 10th that both Fannie and Freddie had enough capital. Indeed, their capital-adequacy requirement was reduced earlier this year so that they could make more of an effort to bolster the housing market.
By its own measure, OFHEO was right. At the end of the first quarter, the two companies exceeded their minimum capital requirements by $11 billion apiece, according to CreditSights. To fall to the “critical level”, which would require OFHEO to take the agencies into “conservatorship” (a fancy word for nationalisation), CreditSights says Fannie would have to lose $16 billion of capital and Freddie $14 billion. And because neither Fannie nor Freddie has depositors, there is no danger of their suffering a run, as Northern Rock, a British bank, did last year.
So why the crisis? Given the gearing in the businesses, things only need to go slightly wrong for there to be a big problem. Freddie lost $3.5 billion in 2007; Fannie reported a $2.2 billion loss in the first quarter, having lost $2.05 billion last year. Each had credit-related write-downs of between $5 billion and $6 billion last year. On a fair-value basis, which assumes that all assets and liabilities are realised immediately, Freddie had negative net worth of $5.2 billion at the end of the first quarter.
Clearly, if the pair continue to lose money for much longer, their capital base will be eroded. And, of course, Congress wanted their businesses to expand—meaning that more, not less, capital would be needed. That would require shareholders to stump up more money. But investors tend to anticipate a big equity-raising by selling the shares, and a falling share price makes an equity issue less likely. The fall was sufficiently speedy in mid-July to prompt Mr Paulson to step in. The stockmarket had called the government’s bluff.
The rescue package may have reassured the creditors but it did not stop the share price of either Fannie or Freddie from falling. After all, the government is likely to extract a heavy penalty from shareholders in return for its support (creditors are another matter, especially as a lot of GSE paper is held by foreign central banks).
Nevertheless the hope is that, if confidence can be restored, Fannie and Freddie can survive without raising capital until market conditions improve. In the short term, as the success of the debt issue on July 14th showed, they should be able to go about their business.
The authorities are keen to avoid nationalisation, which would bring the whole of Fannie’s and Freddie’s debt onto the federal government’s balance sheet. In terms of book-keeping this would almost double the public debt, but that is rather misleading. It would hardly be like issuing $5.2 trillion of new Treasury bonds, because Fannie’s and Freddie’s debt is backed by real assets. Nevertheless, the fear that the taxpayer may have to absorb the GSEs’ debt pushed Treasury bond yields higher. That suggests yet another irony; the debt of the GSEs has been trading as if it were guaranteed by the American government, but the debt of the government was not trading as if Uncle Sam had guaranteed that of the GSEs.
If Congress approves this package, the Fed will have more authority over the agencies. But that will give the central bank another headache. If an institution is struggling, the normal answer is to shrink its activities and wind it down slowly. But that is the last thing that the housing market needs right now.
With the credit crunch, Fannie and Freddie have become more important than ever, financing some 80% of mortgages in January. So they will need to keep lending. Nor is there scope to offload their portfolios of mortgage-backed securities, given that there are scarcely any buyers of such debt. And if the Fed has to worry about safeguarding Fannie and Freddie, can it afford to raise interest rates to combat inflation? American monetary policy may be constrained.
The GSEs are not the only liability for the government. IndyMac’s recent collapse is the latest call on the Federal Deposit Insurance Corporation (FDIC). The FDIC has some $53 billion of assets, so it is better funded than most deposit-insurance schemes. But if enough banks got into trouble, the government would be on the hook for any shortfall. The same is true of the Pension Benefit Guaranty Corporation, which insures private sector benefits, but is already $14 billion in deficit.
In the end, the turtle at the bottom of the pile is the American taxpayer. But that suggests that, if Americans are losing money on their houses, pensions or bank accounts, the right answer is to tax them to pay for it. Perhaps it is no surprise that traders in the credit-default swaps market have recently made bets on the unthinkable: that America may default on its debt.