Thursday, October 25, 2012

What would destroying a central bank's assets do?


Gavyn Davies's post Will central banks cancel government debt? dovetails nicely with the recent fundamental value of fiat money debate. [For commentary on this debate, see Nick Rowe, Paul Krugman, David Glasner, Stephen Williamson here, here, and here, David Andolfatto, Brad DeLong, and Noah Smith]

Let recap the debate first before turning to Gavyn's post. Noah Smith pointed out that since fiat money is fundamentally worth nothing (its future value = 0), then all financial assets are worth zero. Financial assets, after all, are mere promises to receive fiat money. Now back up a second. As I pointed out here, modern central bank money is not fundamentally worthless. Were it to fall to a small discount to its fundamental value, Warren Buffet would buy every bit of money up. Central bank money has a fundamental value because even if it can no longer be passed off to shopkeepers, there are assets in the central bank's kitty. Modern central bank money provides a conditional claim on those assets. David Andolfatto and Stephen Williamson also note the importance of central bank assets.

I got this idea from Mike Sproul. Back in the day, Mike used to start huge comment wars on the Mises blog when he brought up the topic of central bank assets "backing" its liabilities. In fact, here's the post where I first ran into Mike talking about this interesting feature of central bank money. Geez, I sound pretty ornery.

Nick Rowe also brings up central bank assets in his contribution. It's a rendition of his old classic, From gold standard to CPI standard (which I commented on here). In his new post, Nick explains how a modern central bank holds hypothetical CPI baskets in its basement, promising to redeem the liabilities it issues with those CPI baskets at a declining rate 2% every year.

Bill Woolsey gives a similiar story to Nick's in this comment on David Glasner's blog. Bill's point is that if a central bank provides a credible commitment to repurchase and cancel its liabilities should the demand for them evaporate, then central bank money will have value. As he points out, this commitment is only credible as long as the central bank holds (or can get a hold of) the necessary assets to commit to those buy backs.

The point of all this is that central bank assets are important. They are the key for understanding why modern central bank money is different from pure fiat money, why central bank money's future value > 0, and why financial assets (like corporate bonds) that pay out central bank money are not fundamentally worthless. Which brings us back to Gavyn's post.

Gavyn describes a radical idea to reduce UK sovereign debt whereby the Bank of England, the nation's central bank, would cancel part of the government bonds that they've acquired via quantitative easing. By canceling debt held at the BoE, the government's debt-to-GDP ratio comes down. No one in the private sector loses out, since they don't hold any of the canceled debt. The central bank loses out but its loss is counterbalanced by the government's gain such that if you aggregate both under the title "public sector", nothing has happened.

Let's look at this with our previous discussion in mind. With less assets on the BoE's balance sheet, the fundamental value of BoE money would have deteriorated. Why? Should the monetary demand for pounds disappear so that all that remains is fundamental value, the BoE will have fewer assets remaining to commit to repurchases so as to prop up the value of the pound.

Now, it could be that the debt cancellation really means that a formal debt on the asset side of the central bank's books has been replaced by an implicit promise that the government will come to the aid of the central bank during a run on a central bank's liabilities. In that case, holders of BoE money will quickly realize that there is an unwritten and unrecognized asset on the central bank's balance sheet. As a result, the fundamental value ascribed to the central bank liabilities would be damaged somewhat less than a scenario in which the debt was canceled outright. But damaged they would be since implicit guarantees are not as good as real assets.

One reason to make a central bank independent is to cordon off a fixed set of assets that can be used to provide a permanent basis for the fundamental value of central bank money. In this respect, a central bank is like a special purpose vehicle. SPVs are subsidiaries to which a parent company has transferred specific assets. The vehicle has been structured to prevent the parent from tampering with the assets after the fact. The SPV issues its own liabilities to other investors using these ring-fenced assets as backing. A central bank, much like an SPV, has been hived off from its parent, the government, and as a result the holders of its liabilities, the public, can be sure that they have claim to a secure set of assets. If an SPV suddenly had its assets removed by its parent with no guarantee of replacement, the liabilities issued by that SPV would suffer. Same with the liabilities of a central bank.

Gavyn worries that the destruction of central bank assets would unleash inflation. He also points out that there are people who are worried about deflation, and they would welcome a destruction of central bank assets. Whichever way you stand, the point here is that central bank money has a fundamental value. The proof of this would be what Gavyn describes: a scenario in which the value of central bank money declines as central bank assets are destroyed.

Update: Britmouse and Nick Rowe have blog posts on these issues too.

13 comments:

  1. JP, I'd just posted on this before I saw your comments, I also find this interesting.

    The BoE actually already has an explicit agreement that the Treasury will recapitalise it if there are losses on the QE portfolio. And they even put this promise on their balance sheet and treat it as a financial derivative.

    See the discussion of the "Indemnity" in the annual report of the BoE Asset Purchase Facility Fund Ltd.:

    here

    As a matter of accounting, writing down the value of the gilts would write up the value of the Indemnity by an exactly equal value. So the "gilt cancellation" policy would have to be a bit more subtle that Davies imagines.

    I agree that the "implicit recap promise" (even if not explicit) would undermine the effectiveness of "voluntary asset destruction" as a signalling tool. Even without that, the Bank would retain the ability to contract the monetary base by issuing non-monetary liabilities; I think the BoE themselves mooted the idea of issuing BoE Bills to contract the base without selling gilts if that proved necessary.

    And Nick Rowe's point here is surely correct; it is the nominal anchor which defines the fundamental value of CB money. The discussion of policy tools is somewhat secondary. If they started using "voluntary asset destruction" as a policy tool and kept targeting 2% inflation I see no reason why that would be more or less effective at boosting NGDP than current practice.

    - Britmouse

    ReplyDelete
    Replies
    1. Britmouse. Thanks, those are interesting details regarding the BoE Asset Purchase Facility Fund Ltd. Sounds very similiar to what the Fed has done.

      Regarding sterilization, you make a good point that the BoE could do it in principle. But in the hypothetical example I've used for the last two posts (demonetization of central bank liabilities), the only reason people willingly buy sterilization bills (which promise to pay out those same central bank liabilities) is because they are in the final analysis a claim on real central bank assets.

      I'm not really sure about your last paragraph because I don't know what definition of "fundamental" Nick is working with. Using the definition of the term from my last two posts, a central bank can peg the nominal value of its liabilities to gold (for instance), but do so at a price above the fundamental price.

      Delete
  2. JP: in my view, central bank assets do two things:

    1. If the demand for their money is expected to fall at some future date (that is likely true for many central banks right now) the assets will let them buy back that money and prevent inflation rising above target. But then central banks could also borrow against their future stream of seigniorage revenue to do the same thing. (This could be done by the CB issuing its own Tbills, or simply by paying high enough interest on reserves to make reserves big enough.)

    2. They rule out the second equilibrium in which paper money has zero value, and because it has zero value cannot be used as a medium of exchange, and so there would be a zero demand to hold it. (But even the value of paper money as paper might be enough to rule out this second equilibrium.)

    Apart from 1 and 2, I think a central bank with or without assets are two observationally equivalent worlds. If the central bank's assets had been destroyed by a fire, flooded basement, or consumed in a secret party by the central bankers, but nobody ever looked in the central bank's basement, nobody would ever notice any difference.

    In one world, the central bank has 100% assets in its basement, and money trades at its fundamental value.

    In the second world, the central bank has 0% assets in its basement, and money is a stable/sustainable (because the real rate of interest paid on money is less than the growth rate) bubble/ponzi/chain letter. Those two worlds are observationally equivalent.

    ReplyDelete
    Replies
    1. [1] Makes sense to me. You've mentioned before that the CB's monopoly is one of its biggest assets, though unrecognized. I buy that. My last two posts have played around with this by asking, what happens if central bank liabilities are no longer demanded as mediums-of-exchange? - which is equivalent to asking, what happens if you destroy the central bank's monopoly asset, leaving only its regular assets?

      [2] Makes sense to me.

      "Apart from 1 and 2, I think a central bank with or without assets are two observationally equivalent worlds. If the central bank's assets had been destroyed by a fire, flooded basement, or consumed in a secret party by the central bankers, but nobody ever looked in the central bank's basement, nobody would ever notice any difference."

      Agreed. To me, this is equivalent to the recent Sino-Forest scandal. The company literally had no assets in its basement, yet its shares continued to be valued, year after year after year. Only when people dug into the issue did they realize that the company was a sham. In the case of a central bank with no assets, it could probably keep the show going, until the market notices that the central bank has problems doing open market operations.

      "In the second world, the central bank has 0% assets in its basement, and money is a stable/sustainable (because the real rate of interest paid on money is less than the growth rate) bubble/ponzi/chain letter. Those two worlds are observationally equivalent."

      I'll have to read more of your posts on this. Still trying to grasp the concept.

      Delete
  3. I acknowledge that the view I am putting forward here is slightly different from the view I have put forward in my posts. (My views have changed over time, as a result of reading other posts, arguing with commenters, and thinking about it.)

    Here is my revised view:

    Provided the rate of interest paid on central bank money is lass than the long run growth rate in the demand for central bank money, these 3 worlds are observationally equivalent:

    1. The central bank holds 100% real assets, like CPI baskets of goods.

    2. The central bank holds 100% nominal assets, like government bonds that promise to pay central bank money.

    3. The central bank holds 0% assets.

    If I told you there was a hoard of gold in my basement, but I'm never ever going to sell any of it, because I have plenty of income from my salary, and always will, and always give my excess income to charity, and can always borrow against my future income if I need a loan, why would it matter whether you believed me or not?

    ReplyDelete
  4. JP: You make a very important point here, on the link between Gavyn Davies policy proposal and the apparently "angels on pins" debate about whether money is a bubble. They are indeed the same question.

    So I have reformulated my thoughts in a new post: http://worthwhile.typepad.com/worthwhile_canadian_initi/2012/10/the-observational-equivalence-of-fundamental-and-bubble-money.html

    ReplyDelete
  5. "Provided the rate of interest paid on central bank money is lass than the long run growth rate in the demand for central bank money, these 3 worlds are observationally equivalent..."

    I must confess that I'm confused by what "the growth rate" refers to in your first comment. The wording is the same as the posts you did on debt/generations, right? You had a lot of different growth rates in those posts. In your second comment growth refers to demand for central bank money. I'm not sure why the rate of interest paid on central bank money would have a connection to the demand for central bank money.

    ReplyDelete
  6. JP:

    I just reviewed that 2007 blog post where we first argued over the backing theory. Good times. These days I'm much more inclined to say that the dollar is backed by the combined assets of the central bank and the government, not just by the central bank's assets. Not only would the government probably bail out the central bank if the central bank lost assets, but even if no bailout happened, the government would continue to accept its own dollars for taxes, and that alone could maintain the dollar's value.

    In my paper about the Fiscal Theory of the Price Level (downloadable on REPEC) I use an analogy of a landlord who collects rents of 50 oz. of silver per year, so that if R=5%, his land is worth 1000 oz. The landlord then goes around town buying stuff by writing out 1 oz rent certificates (dollars) that he agrees to accept for rent. Once he issues about 700 of these dollars, people will start to worry about their backing, so let's say that he prudently issues just $400 and spends it on candy. No problem. The 400 dollars are more than covered by the 1000 oz of land. His net worth is 600 oz.

    Next, he buys another 2000 oz worth of land, paying for it by issuing a 2000 oz. bond. The seller is barely willing to accept this, as long as the landlord has that cushion of 600 oz of net worth. But now the landlord has extended himself as far as he can. Each of his dollars is still worth 1 oz, but any further borrowing or money-issuance will get his creditors worried.

    Next, the landlord designates his basement as his central bank. From his basement, he issues another $1000 of money, and uses it to buy 1000 oz of the bonds that he previously paid to the land seller. He keeps the bonds in his basement, and the money he issued goes first to the bondholder, then to the rest of the economy. Now he's in Noah's world, and it should be obvious that his net worth is unaffected. Each of his dollars is still worth 1 oz. He could even sell 200 oz worth of his bonds for 200 oz of silver. Now he can tell people that there's a hoard of silver in his basement, and he can even offer silver convertibility of his dollars. The dollar is still worth 1 oz.

    At some point the landlord might stop offering silver convertibility. Then quantity theorists will go crazy and say that his dollars have no backing, that they are fiat money. The folks at mises.org will call him a fraud and a counterfeiter. Nick will say that the value of the dollars is held up by nothing but money demand, network effects, etc., and he'll add that it wouldn't matter if the landlord lost 70% of the assets in his basement.

    Finally, the landlord does what Noah suggested and issues another $500 of money from his basement, buys another 500 oz of his own bonds back from the public, and burns the bonds. He just swapped one liability (dollars) for another (bonds). His net worth is still unaffected, the dollars and bonds hold their value, and nothing important happens.

    ReplyDelete
  7. "...but even if no bailout happened, the government would continue to accept its own dollars for taxes, and that alone could maintain the dollar's value."

    Oh oh, looks like you and Nick have another thing to disagree over. See Nick here.

    ReplyDelete
  8. As a holder of fiat money I have no conditional claim on anything but fiat money from a central bank. I can't walk in there and ask for an asset for my fiat money. I stopped reading after that point.

    ReplyDelete
    Replies
    1. You're right. You can't walk into the bank and ask for an asset. It's would be the same if you were a shareholder in Google. You can't walk into Google and ask for an asset.

      The key word here is "conditional". You have a claim conditional on certain events. Conditional on what? Windup, bankruptcy, or liquidation. Or maybe the central bank (or Google) decides to buy up all its liabilities.

      Delete
  9. In that case, this kind of question should not be taken from
    any show telecasted in specific country. Popular prizes include sports tickets,
    cash and vouchers for drinks, food - and dollars off of tabs.
    Her father, Bruce Paltrow, produced the critically acclaimed TV series that is considered the
    precursor to many medical shows today, St.

    Feel free to visit my webpage - best pub quiz team names

    ReplyDelete