Sunday, January 11, 2009

Should You Care if the Fed is Broke?

Are there any practical implications of central bank insolvency?  Should we even care?  This is a tough question for me.

In Buiter's magnum opus on all the ways a central bank can stuff an economy to the gills with currency, he discusses the balance sheet of a central bank.  Historically, the equity of the bank that issued the currency helped make the currency sound and valuable.  This was once very important, and the US has had many different currencies over the years, sometimes many at once, all with colorful histories.

It had become almost a historical relic before the Fed's new high-wire act.  Buiter details some of the risks the Fed is exposing itself to today, and mentions the need for the bank to recapitalize itself.  He doesn't go deep into why this recapitalization is necessary and how it is important, so we will.

Risks

Inflation risk and credit risk are the two really important risks the Fed is facing.

Inflation risk decribes the likelihood that inflation will rise too far during the term of the loan, and is greater for longer loans.  Should there be any resurgence of inflation over the next multiple years, the Fed will suffer at least a mark-to-market loss.

Credit risk describes the likelihood that the deadbeats will fail to pay you back.  Should it turn out that the Fed bought some toxic paper(I know, right?), then the Fed will suffer a capital loss.  More of this paper is likely to be bad if deflation persists.

A mark-to-market loss or a capital loss might seem like it doesn't really matter.  But as Sims pointed out at the beginning of this fiasco, it can actually be quite a big deal.  We will look at the Fed's three main tools for stopping inflation: open market operations, paying interest on excess reserves, and raising the required reserves ratio.  Suffering a either loss limits its capacity for open market operations, which is why Buiter discusses recapitalization as required, but there are other creative approaches.

Standard OMO

When the Fed wants to take dollars out of the system to curb inflation, it has to get its dollars back from people.  It traditionally sold off some of the Treasury securities on its balance sheet in exchange for these dollars.  This removed the dollars from circulation, hypothetically limiting the amount of credit that can be created, shadow banking system notwithstanding, controlling inflation.

But what happens if the assets it owns are no longer Treasury securities, but a hodgepodge of pier loans, bad credit card receivables, MBS IO strips, and more?  It would have to resort to selling those, and it's quite probable that those would be bought by the private sector only at a significant discount.  The Fed would be left with a negative equity position.

This negative equity position could be quite large, but only in severe inflation should it become a real problem: when the Fed needed to sell more assets than it actually had.  That's only when recapitalization with quality assets through outright taxation would be explicitly necessary.  If things didn't get that bad, the Fed might be able to simply sit there with the hole on its books, waiting eternally for seigniorage to fill its pockets once more.  The Treasury loses the small revenue from seigniorage, and that's the end of that.

Interest on Excess Reserves

Some of the clever readers in the blogosphere have suggested an alternative solution.  The Fed now pays interest on deposits.  If the Fed were able to increase that interest rate arbitrarily, it would quickly become more appealing to deposit funds at the Fed than loan them out.  This would give the Fed another way to effectively take dollars out of circulation and curb inflation.

A key problem with this idea is that the Fed's balance sheet grows gradually worse.  Just like anyone's favorite Ponzi -- or Madoff -- scheme, the Fed would be paying out more in interest on its good liabilities than it takes in on its bum assets.  It's fun to call the Fed a Ponzi scheme, but in this case it would actually be true: it would gradually grow more broke.

This might be a temporary state of affairs as it acquires more assets that yield more than it must pay out in the future, but that really depends on the state of the economy, and how much inflation it's willing to suffer.

Thus, even if the Fed ran out of assets to sell, it still isn't completely powerless.  In fact, the things that normally cause a Ponzi scheme to fail, like excessive redemption requests, are by definition not at play here.  Could it go on indefinitely?

Required Reserve Ratios

The Fed can also simply change by mandate how many dollars are deposited on its books by depository institutions.  It increases the required reserve ratio, which should mean the bank can create fewer dollars.  Unfortunately, banks are pretty talented at getting around regulations designed to limit their aggressive risk taking.  The shadow banking system and other leveraging mechanisms would probably flourish.

The RRR is currently roughly 10%.  It could easily be raised or lowered from there.  It might not have much tangible impact, and it would crimp bank profitability.  However, if enforced well, it would be a way to curb money creation with little risk to the Fed.

Conclusion

Even if the Fed is in the hole, there are still a lot of things the Fed can do.  I don't think the Fed will run out of options for controlling inflation.

Are there lesser risks than hyperinflation or collapse of the currency?

If either of the first two options were used, there would be more dollars outstanding than there would naturally be, which would be somewhat more inflationary, and require somewhat higher interest rates to combat it.  That's probably not a really big deal.

The one risk that I have sympathy for is a general loss of confidence in the currency.  Holders may anticipate the need for the bank to recapitalize itself, through inflation or taxation of dollars, and desire to escape this.  They may also not really enjoy Ponzi schemes.  This loss of confidence should lead to an increase in the level of interest rates or the amount of inflation that would occur.  It's a long-term drag either way.

The only way I could see a real hyperinflationary collapse happen is if the Fed were politically restricted from fighting aggressively, or there were simply a widespread collapse of confidence in the currency altogether.

Complex topic, and I'm sure I missed a lot of things.  What are your thoughts?

9 comments:

Mikkel said...

I hate to get all Austrian on you, but at some points it is confusing what you mean by inflation/deflation. While I don't agree that "price is always related to monetary supply" for a variety of reasons, there is some advantage to referring to inflation/deflation as related to monetary supply and price increase/decrease to prices.

For example in most of your post -- and your first one -- inflation and deflation obviously refer to trying to control monetary supply. However, as the final comment you allude to "collapse of confidence in the currency altogether."

I believe it is a mistake not to separate these two explicitly, and it is shared by both the inflation and deflationistas. With the rumblings already going on in the world, it is obvious that they are expecting us to try to enforce inflation and that's planting the seeds of a loss of faith. Ditto for Yves post about an increasing amount of trade in non-dollar denominated currencies, and if I may, even limited country to country bartering (i.e. Iran and um...who'd they trade oil with for some tangible thing?)

I'm open for arguments against this, but it seems to me that if the validity of currency itself starts to be challenged, then inflation and deflation of monetary supply become untethered from price. I would not be surprised at all for there to be net domestic deflation (monetary supply contracts once you account for velocity, asset and wage decreases) but also have the dollar plummet and huge import price increases. These dynamics would of course quickly cause a terrible positive feedback loop.

To me we need to just suck it up and tell our creditors that we're going to do our best to repay them with good money and are going to take the hit on the chin...but require cooperation (and even negotiation) with them to make sure that hit isn't a knockout blow.

ndk said...

I hate to get all Austrian on you, but at some points it is confusing what you mean by inflation/deflation. While I don't agree that "price is always related to monetary supply" for a variety of reasons, there is some advantage to referring to inflation/deflation as related to monetary supply and price increase/decrease to prices.

Mikkel, I really try hard to avoid this debate, because it's a religious war for a lot of people. I don't have strong feelings either way, but when I use "inflation" or "deflation", I'm generally using the terms as the government does: a sustained broad rise, or a sustained broad fall, in the aggregate price level of an economy.

Controlling money supply in my mind has empirically only weak effects on the aggregate price level, and in fact my first post was largely dedicated to that, trying to show that we could create as many dollars as we want, but something else entirely was constraining pricing.

With the rumblings already going on in the world, it is obvious that they are expecting us to try to enforce inflation and that's planting the seeds of a loss of faith.

I strongly agree with that, and furthermore, this is the Fed's explicit aim. Economists call it a credible commitment to inflate. It drives me nuts, because such a commitment in the face of structural pressures that prevent inflation can do no good. At best, you confuse people and cause them to undertake actions they wouldn't if it weren't for your jawboning. At worst, people freak out, panic, and there's a run.

Fix the structural issues, and then we can talk expectations.

I would not be surprised at all for there to be net domestic deflation (monetary supply contracts once you account for velocity, asset and wage decreases) but also have the dollar plummet and huge import price increases.

I can understand your thinking, but it might neglect a few feedbacks. A plummet in dollar prices and an increase of external prices would tend to cause domestic inflation directly, as domestic producers could charge higher prices and workers could demand higher wages. Exports from the US would soar and imports would drop.

Some things, such as the need to import large amounts of oil to run our economy, could lead to unexpected behavior here. But I generally agree with the party line here that a large drop in the dollar would be inflationary domestically.

To me we need to just suck it up and tell our creditors that we're going to do our best to repay them with good money and are going to take the hit on the chin...but require cooperation (and even negotiation) with them to make sure that hit isn't a knockout blow.

I'd like that. We've devoted an awful lot of effort towards avoiding making good on our commitment to the extent we can. We might devote more effort to making good on what we're able. I don't think our creditors are remotely interested in such negotiation yet, but good faith is a good first step.

Anonymous said...

ndk,

My response to your response at Buiter's:

"A central bank that seeks to unwind both policy rate and quantitative easing can pursue both jointly, at some co-ordinated “measured pace”. My point is that there shouldn’t be a desperate need to “front run” policy rate tightening with quantitative tightening, since both actions should reinforce each other through market expectations.

A government can recapitalize its central bank by issuing bonds in exchange for capital. This is a non-cash swap that leaves the consolidated resources of the government and its central bank unchanged. It provides the central bank liquid assets to sell or repo as necessary for monetary tightening; the point being that if the central bank can manufacture its own liabilities instead, or tighten by increasing the policy rate and resetting the reserve interest rate floor, insolvency is not necessarily a binding operational constraint - not that persistence despite insolvency is a particularly elegant or prudent public policy. (Insolvency imposes even less operational constraint for easing.) I think Professor Buiter’s prescription for credit limits backed by recapitalization “insurance” is much better as public policy. A further measure might be an official presentation of the consolidated solvency position of the government and its central bank as an annual reporting requirement. That would be quite entertaining, if not enlightening, and a veritable feast for the blogosphere."

Anonymous said...

Additional comments:

Given the operational capabilities of central banks, recapitalization is cosmetically interesting but substantially vacuous. The solvency of the consolidated entity is unchanged as a result of a government recapitalizing its central bank. The central bank solvency effect is purely internal accounting. Similarly, recapitalization by seigniorage is purely cosmetic when the assets of the bank are liabilities of the government.

The operational recapitalization of central banks provides them with government liabilities to use in sterilization operations, in addition to their own liabilities. This is convenient. But the liabilities of central banks and their governments constitute a seamless continuum of instruments for the sterilization of excess reserves. Currency, non-reserve interest paying liabilities of the central bank, government bonds, and taxes all serve to drain excess reserves. There is no purely operational reason why government bonds must first be positioned as central bank assets in order to achieve the reserve effect of selling them into the market. And paying interest on required and excess reserves provides essentially the same interest rate support as quantitative sterilization.

Solvency is only an economically meaningfully concept in the context of the consolidated government and central bank entity.

Mikkel said...

"A plummet in dollar prices and an increase of external prices would tend to cause domestic inflation directly, as domestic producers could charge higher prices and workers could demand higher wages. Exports from the US would soar and imports would drop."

On this point I am pessimistic. I've read that a huge component of US exports are infrastructure and manufacturing related. I might be wrong, but I'm under the general impression that we primarily export components that are then installed to make stuff to sell back to us. And then of course a lot of the rest is highly technical related.

With extreme specialization in trade that the world current sees, I have little faith in generalized models about the relationship to currency strength. In fact, I think this is the huge blindspot that almost no one sees. Just look at Brad Setzer's most recent data on Taiwan and Korea for an example of how quickly trade collapses where there is little diversification.

If anything, food production is our greatest strength, but if oil goes back up to $200 a barrel because of dollar declines...

I disagree that charging higher prices means that workers can demand higher wages. It's all about margins. In my proposed scenario, margins are going to be squeezed like never before and I don't see how there will be any wage pricing power. I guess one outcome is wage inflation (or stagnation at least) but just all the price increases go into higher unemployment. Now that I think about it that might be the most likely scenario.

ndk said...

Largely agree with your comments, JKH, and thank you for coming by.

A further measure might be an official presentation of the consolidated solvency position of the government and its central bank as an annual reporting requirement. That would be quite entertaining, if not enlightening, and a veritable feast for the blogosphere.

This would be fun, I agree. The estimation would be difficult and sensitive because you'd need a good read on the NPV of seigniorage, and that alone would open eyes to aspects of our currency system that might be best kept under the blankets.

And paying interest on required and excess reserves provides essentially the same interest rate support as quantitative sterilization.

Solvency is only an economically meaningfully concept in the context of the consolidated government and central bank entity.


I agree, and in further defense of this idea, you'll notice that the practical negative effects I find of a bankrupt central bank are very similar to the practical negative effects of having lots of government debt outstanding. That's a good indication for both theories.

Viewed against the volume of treasuries outstanding, any Fed insolvency is relatively small by comparison. That feeds back into my suspicion that a bankrupt central bank, as a practical matter, might not be such a big deal after all.

ndk said...

Just look at Brad Setzer's most recent data on Taiwan and Korea for an example of how quickly trade collapses where there is little diversification.

That was ghastly. But at least as far as diversity in exports goes, I think the US sits a little better than Taiwan and Korea, as you can see from perusing exhibits 3 and 7 in our FT900 international trade report. I'm not well-versed in this, though.

If anything, food production is our greatest strength, but if oil goes back up to $200 a barrel because of dollar declines...

Oil is a big problem, I agree. A generalized commodity bubble probably helps rather than hinders the US' position, but if it were only oil or energy that soared, it wouldn't be a good situation.

I disagree that charging higher prices means that workers can demand higher wages. It's all about margins. In my proposed scenario, margins are going to be squeezed like never before and I don't see how there will be any wage pricing power.

There is an extraordinary surplus of skilled and unskilled labor, synchronization of the business cycle, and margin compression or destruction of business model that has been unleashed by globalization and the Internet. We are all the small retailer and Wal-Mart has just moved into town. I also wouldn't be surprised at all globally persistent wage and margin pressures for the foreseeable future, and that is indeed a strong counter to any inflationary force.

Mikkel said...

Thanks, I was looking for that trade data! My concern isn't that we aren't diversified in products, but how much of the "Industrial supplies and materials" and "Capital goods" categories are purchased simply to create end products that are then sold back to us (or in building up extravagant tourism related industries that of course will collapse). Without an appreciable rise in international consumption they won't have any reason to buy things from us...I guess I'd say we can have diversity of goods but not diversity of intent.

David said...

If the Fed bought every garbage instrument and they all proved worthless, they indeed would have MTM losses. If inflation peaked up, they would likely have more trouble reigning it in. They couldn't take all of those dollars out of circulation. But lets look at the worth case scenario. If they try to sell their assets to take out the dollars and get to the point where they have sold everything (even the marble columns), we still never get hyperinflation. We just get inflation. Even if the Fed goes out of business, there will only be a slightly higher amount of money in the world. As long as they don't go printing more of them, you can't have hyperinflation. The dollar would become, like gold. That is, there would be a fixed supply even if it is somewhat larger than before. The market would recognize that and discover the right value.

Eventually, a growing economy with a fixed money supply would lead to deflation. So you would get a one-time devaluation not hyperinflation. Hyperinflation is only possible when a government wants hyperinflation. Printing a few trillion dollars can't result in anymore than say a 40% depreciation. That might not be great but would be something American's could live with. That is, it would have some beneficial effects.