Risk Sharing Averted: Fannie and Freddie

Economists spend a considerable amount of time thinking about cross-country risk-sharing. It is rare to go through an international macroeconomics conference of more than 4 or 5 papers without seeing some ruminations on the question of whether the international financial system can deliver risk sharing across countries. The ability to diversify and share risks across borders is, after all, one of the most significant potential benefits of a global financial system.

The idea is that if something bad happens in your country, if you have sold a number of your assets to foreign investors and also bought a number of assets abroad, then the losses you face on a negative shock that hits just your economy are smaller. Foreign investors bear some of the losses and you own some assets out of your country that will not be affected by the downturn. In the classic model, better to own half the fruit bearing trees on both your island and the other island than just all the trees on your island.

In some ways, we have seen risk sharing in action during the US mortgage mess. A number of large investment banks whose primary ownership is from Europe have faced substantial losses in the US market. Further, a number of banks or pension funds abroad had some exposure to the US mortgage market. None of this is good for the European countries, but it certainly helped the US to have some of the afflicted assets owned outside the US rather than having them all reside on the US balance sheets.

The effective nationalization of the downside risk to Fannie and Freddie bonds, though, has cut off this process. Foreign investors – and notably foreign central banks – owned a lot of these bonds. The Wall Street Journal reported that Paulson listed the estimated 5 Trillion dollars of mortgage backed securities and Fannie and Freddie bonds held overseas as part of the reason the Treasury had to step in, but he did not seem to note that this meant the US taxpayers were taking on exposure to risks that foreign investors had taken (and profited from in some cases). The latest TIC data shows at least 1.3 Trillion of Agency (Fannie and Freddie) bonds held abroad, and more than half of that ($700B) is held by foreign Central Banks. Had the Treasury forced the bondholders, not just the shareholders, to take losses, some of these losses would have been felt abroad. The US taxpayer is effectively bailing out foreign banks and central banks to some extent.

Thus, when economists measure the potential for risk sharing based on the extent that assets are cross-held, one crucial necessary condition is that losses on assets not be taken on by the government. If they are, the risk is lifted from foreign investor and pushed onto taxpayers and hence stays in the local economy.

This is something of a special case. People have generally spoken of these bonds as having an implicit guarantee and hence they have often been purchased as near substitutes for US Treasury bonds, and Treasury officials have apparently been trying to soothe foreign banks’ fears in the last few months by reassuring them that these bonds are safe. Treasury officials may have worried that forcing large losses on central banks (and others) when those investors had plausible beliefs that the assets they bought had official backing would have weakened the credibility of all US bonds and possibly put serious downward pressure on the dollar.

Still, it wasn’t just “fat cats on Wall Street” who avoided taking a loss on the bonds. Many banks and governments around the world had large exposure. The US government just took on that exposure. It has been argued that in cases like this, profits were private, but losses were socialized. It may also be that profits were global and losses local – not the type of risk sharing we generally think of.

2 Responses to "Risk Sharing Averted: Fannie and Freddie"

  1. Stormy   September 12, 2008 at 8:35 pm

    Nice post.

  2. Anonymous   September 13, 2008 at 11:32 pm

    Good post. The upside is that housing values are helped within the USA, which is a valid governmental purpose in my view. (The “ambiguity” created by the “implied guarantee” is not important; since when could you ever get paid because of “ambiguity” or “implication” — I never could!)Now the new one is Lehman. There is absolutely no governmental purpose for this international behemoth, let it collapse. Boom! Risk sharing — bring it on!!And, hope more collapse before election. After election, there is infinite “moral hazard” from a government that no longer has a reason to please the people. Now that’s the real moral hazard!