The New York Times has an arresting chart on the government’s new financial commitments made during the financial crisis. According to the Times, the government has committed $3.1 trillion as an insurer, $3.0 trillion as an investor, and $1.7 trillion as a lender. Wow, you may think, that’s a lot of money. US GDP is about $14 trillion per year; the budget deficit in recent years has been running in the half-trillion range. But wait, there’s more: the Times omits roughly $5 trillion in guarantees made by Fannie Mae and Freddie Mac that are now officially on the government balance sheet (although they were always implicitly there).
All that said, though, there’s a big difference between these “commitments” and ordinary government spending. Ordinary government spending simply evaporates into the economy: for example, Medicare expenses go to pay for people’s health care, and the government will never get them back. Making financial commitments is what banks and other financial institutions do, and they do it because they expect to get their money back. What we are seeing is the growth of a massive financial institution within the government. This one’s primary goal is the public interest – in this case, the health of the economy – rather than getting its money back. But still, it should get most of the money back.
Let’s start with the insurance programs. Half of the Times’s $3.1 trillion number is the $1.5 trillion guarantee on new senior unsecured bank debt announced by the FDIC in October. Under that program, the FDIC is charging an insurance premium to banks of 0.75% of the debt issued. (I heard that banks are trying to negotiate this down, but I don’t know where that stands.) So the FDIC’s eventual losses will not be $1.5 trillion (the maximum amount of debt guaranteed), or even the fraction of that debt that defaults, but that fraction minus the insurance premium on all the debt, minus any return the FDIC gets on the premiums in the meantime. Who knows, the FDIC might even make money.
Of the $3.0 trillion in investments, the biggest chunk is the $1.6 trillion the Fed made available to buy commercial paper directly from issuers (big companies). Commercial paper pays interest, and it is historically among the safer of investments; this is a big part of what money market funds traditionally invest in. The rate of defaults on commercial paper will certainly go up during the recession, so there is a chance that the Fed will lose some money on this deal. But it should only be a small fraction of the $1.6 trillion. Another chunk of the investments is the capital injections that Treasury is making into banks; while there is some risk that some of that money will not be paid back, the money invested is earning 5% per year (8% for the latest Citigroup tranche).
Of the $1.7 trillion in loans, most of this is new facilities made available by the Fed to offer short-term loans against a wide variety of collateral. The vast majority of these loans will be paid back; for those that will default, the Fed will be able to sell its collateral, probably at a loss (since the whole point of this program was to give people a place to put their illiquid, impaired assets). So again, expected losses should be low.
I don’t have anything close to the data you would need to forecast the actual losses, but my wild guess would be the low hundreds of billions. (I am not particularly hopeful about the guarantee on $306 billion in toxic Citigroup assets, for example.) The ultimate magnitude of that loss depends, more than anything else, on the overall state of the economy over the next 3-5 years.
I wouldn’t say there is no reason to worry about the vast amounts of money the government is putting on the line. And you can have legitimate concerns about the influence this means the government has in the economy, or the ability of the government to manage this volume of assets and programs. But when you see a number like $7.8 trillion, it’s important to bear in mind what it means.
The Baseline Scenario 
Mr. Kwak – I have a couple of questions for you. If the government has the potential to recover portions of such bailout dollars, when will we begin to see that happen, 18 months, 24 months, 48 months? That being asked, I can assume that the returns will not be anytime soon, at which point I ask, when will the dollar reach an inflationary tipping point. I recognize we are experiencing deflationary pressures; however, in essence the national debt has doubled in 18 months, accelerating issues the former comptroller of the U.S. government warned of 15 to 20 years down the road. I am also beginning to hear numbers for a stimulus in excess of another 600 billion. Are we looking at hyper-inflation?
Let me summarize your post: assuming that losses are normal, very little of the government’s money is at risk.
Now let me remind you of some previous, similar precepts held by policy makers and market participants:
-Assuming that house prices never go down, there is very little risk of subprime losses.
-Assuming that volatility stays low, there is very little risk in derivatives contracts.
-Assuming that assets stay uncorrelated, there is very little risk in CDO’s.
-etc, etc.
The point is, we are not facing normal times. The market recognizes this by pricing in tail risk. You, Paulson and Bernanke think there is no tail risk. Fine, except that the latter two can express that view with my (the taxpayer’s) money. I’d rather they didn’t.
You say the government will get most of the money back. But beware the political process. Now that the money is out there, private interests will do their best to keep it. There will be lobbying to forgive the loans, for example, because they “can’t afford” to pay them back.
An accounting question – If the Fed is an insurer($3trillion) there is no cash outlay, but if the Fed is an investor($3trillion) or a lender ($1.7 trillion) then it seems there is a real cash outlay. But the Fed doesn’t have $4.7 trillion does it?
You’re correct, the Fed doesn’t have $4.7 trillion lying around. So the Fed will need to finance its investments – for example, its commercial paper purchases – somehow. The most direct means is simply to create money. This generally has an inflationary effect, although how significant that would be right now is anyone’s guess. Another route, which the Fed has used a lot recently, is to ask Treasury to issue new bonds (government debt) and deposit the proceeds in its Fed bank account. This does not create inflation (at least not directly), but does increase the debt.
Paroune: As far as timing is concerned, it depends on the type of commitment. For example, commercial paper purchases are short-term by definition, so the Fed will be constantly getting paid back and, presumably, rolling over its purchases. The real question is when the Fed can get out of the business of buying commercial paper, and there we could be talking about a couple of years. The preferred stock investments have dividends that go up after 3 years, so there is a good chance that banks will buy them out if at all possible at that point. Although Mark makes a good point that if banks are in trouble, there will be sure to be lobbying to relax their obligations. In any case, we are probably talking about years. As to whether or when we will hit an inflationary tipping point, that is anyone’s guess. Brad Setser has an often-cited post on China arguing that Chinese demand for US Treasuries will continue to be strong.