Adding all the offbalance-sheet liabilities together, I calculate total federal off-balance-sheet commitments came to $70.1 T as of 2012, or about 6 times the size of the on-balance-sheet debt. In other words, the budget impact associated with an aging population and other challenges could turn out to have much more significant fiscal consequences than even the mountain of on-balance-sheet debt already accumulated.
When Hamilton presented this paper a several weeks ago at Cato, Bob Hall and I had exactly the same reaction. The off-balance-sheet liabilities are contingent liabilities. They often take the form of out-of-the-money options. Think of the Pension Benefit Guaranty Corporation. In some states of the world, it will lose a lot of money, and in other states it will break even or make a profit. To report just one number seems uninformative. The same holds for the government’s portfolio and guarantees of mortgages and mortgage-backed securities. The problem cries out for scenario analysis, in which you present possible values for the key drivers (such as interest rates) and possible outcomes (for, say, the ten-year budget outlook).
This led to a testy exchange between me and Douglas Holtz-Eakin, who insisted that Congress wants a single number. It so happened that a couple of weeks ago I was scheduled to give an informal talk at the Congressional Budget Office (which Holtz-Eakin once headed) on a topic of my choice. I chose the topic of scenario analysis.
I said that for the purpose of my talk, we would assume that you could talk to Congress like adults. That is, anyone in a position of responsibility at a large financial corporation could understand scenario analysis. If our elected representatives, who oversee trillions of dollars, cannot handle it, then we have some really big problems. (I think, in fact, that this is the case. As an aside, I would love to have someone who thinks government is not too big explain to me why he is not bothered by the fact that you cannot have an adult conversation with the people who are in charge of it.)
So, assuming that you would not be thrown out of the room for engaging in scenario analysis, the question becomes how one should do it. I thought that the more outspoken people at CBO were a bit defensive. They said that in the case of macroeconomic forecasting, for example, they had white papers that considered many scenarios and that they reported a range of possibilities based on those scenarios. My reply was that this was not a particularly helpful way to communicate scenario analysis–it just creates a sort of smeared picture. Instead, for example, I suggested that in textbook macro terms you could look at the effect of fiscal stimulus under a scenario in which the Fed holds interest rates constant, a scenario in which the Fed uses a Taylor rule, and a scenario under which the Fed targets nominal GDP. Showing those three scenarios probably would be educational.
Returning to off-balance sheet liabilities, key drivers include interest rates, demographics, and the impact of medical technology and practice. I am particularly interested in seeing the effects of interest rates, because I suspect that a rise in interest rates would adversely affect the budget outlook for many of these off-balance-sheet items.
“I suspect that a rise in interest rates would adversely affect the budget outlook for many of these off-balance-sheet items”
Probably not, for the only reason for interest rates to rise is stronger growth. The real problem with these is they are already assuming higher interest rates than are likely. This is why our greatest fears are never realized, we are continually looking in the rear view mirror.
But doesn’t government face a “perfect storm” risk – the Great Recession turns into the Even Greater Depression, which results in lots of people on whatever assitance programs they can get on, AND lots of bankrupticies and related wipe-outs of pensions AND lots of housing defaults all at about the same time?
(Jo’s company goes bust, Jo’s pension is destroyed by market crash, Jo can’t pay for her house and defaults, Jo has to eat at a soup kitchen and go on SS early to avoid starving. If this happens to many millions of Jo’s at once, it’s a perfect storm.)
Dr. Kling: If off-balance-sheet liabilities can be analyzed as options, why not simply estimate their option value and report it, if Congress insists on just one number? You can use scenario analysis to project a wide range of paths and compute the option value using risk-neutral valuation? Then we would have a single number that fairly accurately reflects the wide range of possible outcomes.
I echo Jack’s PQ comment, having seen him beat me to the punch. As Dr Kling knows, there are various modeling techniques that could not only give you “a number”, but one that does not just rely in simplifying assumptions (such as constant volatilty, etc.). They as close as you can get to crystalizing myriad scenarios into a number but also complementing the number with all the “greeks” you need to gauge “risk” to the “number”.
More intriguinly, I recall watching the lecture Dr, Kling refers to, and remenber the comments made. But I remember the gents having an even more objectionable objection: the the tally of “liabilities” did not nert out the value of the assets the guarantees guaranteed. This is an even more basic problem. The exposure should be the simulated (or option-adjusted) value of the net of assets – liabilities (or value at which the assets are guaranteed or accounted for at). Was this not also true?
I do not think that the fair value of the option is the only number that matters. It is not even the number that matters the most. What probably matters the most is the cost of the options under other scenarios in which the government balance sheet is adversely effected, such as a scenario with higher interest rates.
I’m with Arnold. A fair value calculation would be most useful if the contingent liabilities were tradable or hedgeable. Since most of the government’s s are neither,
“As an aside, I would love to have someone who thinks government is not too big explain to me why he is not bothered by the fact that you cannot have an adult conversation with the people who are in charge of it.”
Because democracy relies more on the wisdom of crowds than the wisdom of leaders. Politics may select for people who are only good at politics, but behind the politicians stand huge coalitions of millions of Americans, some of them very smart, some more influential than others, all of them, when added together, very interested in America being a basically succeseful country. Giant coalitions work because, in a society that protects minorities, that much self interest actually does add up to something hazily resembling an interest in the common good. Public choice problems, public ignorance, and rent seeking all complicate that story and create inefficiencies (big ones, sometimes!), but the end result is still pretty good. And the end result is big government.
In other words, I am not worried if politicians are bad policy makers because their job isn’t actually to make polcy. Their job is to pass the policies the voters and lobbyiests push on them.