« The optimal destruction of wolves | Main | L’Europe et l’alchimie monétaire »

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

But only a central bank can act as an unlimited lender of last resort, because only a central bank can print unlimited quantities of irredeemable money.

No CB under (gold) convertibility?

Some other observations...

The 20s Fed was exclusively fixated on the penalty-rate. There was a lot of debate about what 'last resort' meant. For instance, there should be a lender to any liquidity issue at some price. The penalty-rate is clearly an instance of a soft usury law.

In the 80s the Fed acted as a lender of last resort routinely because OMO we used to create a structural scarcity of reserves, the balance of which was met via the discount window at the penalty rate.

A question: I thought that the ECB was obliged by treaty to discount the public debt of any euro-zone member. This is one reason yields on euro-zone debt were so compressed for so long. So why precisely is there a discretionary question here at all. The normal mechanisms of the institution are such that the Greek CB can raise unlimited funds in Euros by fronting an unlimited supply of zero-coupon debt to the discount window of the ECB.

Sure I know there is some sort of 'no bailout' clause, but I'm a afraid I don't understand the legal strictures by which it disables the normal mechanisms of the institution. Somehow it seems like a framing thing.

Jon: Under gold convertibility, central banks weren't really central banks, at least not as we understand the term today. Gold miners fulfilled part of that function. And during a liquidity crisis central banks were often forced to suspend convertibility. I was thinking about saying this in the post, but decided it would be too much of a digression.

JP's the resident expert here on the rules about what central banks may and may not do. As I understand it the ECB can act as lender of last resort to a commercial bank that has A-rated government bonds it can leave as collateral. And then it imposes a haircut, I think. It cannot lend against any asset of a solvent bank. A bank could be solvent, but unable to borrow from the ECB. Plus, Greek bonds ratings have dropped (but it's ignoring the rules). And it can't help a bank whose solvency is questionable. And it can't help governments. Etc.

I agree with Jon's point about the Greek CB being able to raise unlimited Euros by supplying zero-coupon debt. The ECB has collaterol requirements, such as rating requirements for the bonds. These requirements have been suspended for Greece (I imagine the suspension will be extended Euro-land wide soon). So what's to stop Greece from funding itself by just supplying an unlimited amount of zero-coupon bonds to the ECB?

"The normal mechanisms of the institution are such that the Greek CB can raise unlimited funds in Euros by fronting an unlimited supply of zero-coupon debt to the discount window of the ECB."

Do you have any links for this?

I'm fairly sure it can't be done:

See Article 123 TEC – The Prohibition on Monetary Financing Rule

1. Overdraft facilities or any other type of credit facility with the ECB or with the central banks of the Member States (hereinafter referred to as ‘national central banks’) in favour of Community institutions or bodies, central governments, regional, local or other public authorities, other bodies governed by public law, or public undertakings of Member States shall be prohibited, as shall the purchase directly from them by the ECB or national central banks of debt instruments.

"But only a central bank can act as an unlimited lender of last resort, because only a central bank can print unlimited quantities of irredeemable money."

1. If a private bank is insolvent, then a loan from the central bank is not really a loan. It is a gift. The central bank is just transferring wealth to the private bank. If the private bank is solvent, then it doesn't need a lender of last resort.

2. Consider a free banking world. ANY bank can print unlimited quantities of irredeemable paper money. If the bank has enough assets to buy back its money at par, then that money will hold its value. If not, the money will lose value.

Mike: " If a private bank is insolvent, then a loan from the central bank is not really a loan. It is a gift."

A partial gift, yes. Depending on the degree to which it is insolvent.

"If the private bank is solvent, then it doesn't need a lender of last resort."

I disagree. If everyone wants to withdraw at once, and the bank's assets are illiquid, so it can't sell them, or can only sell them at a discount, a liquidity crisis can turn into a solvency crisis. This is especially true if it happens on an economy-wide scale.

2. Why does such a bank need assets, if the people holding its money can't redeem it? To pay interest on the money, to persuade them to hold it? Or to redeem it in case its value falls, if ever there were a run on that bank?

JP, Kosta, Nick. I did not intend to make a statement of fact, everything I stated should be framed by the remark: "Sure I know there is some sort of 'no bailout' clause, but I'm a afraid I don't understand the legal strictures by which it disables the normal mechanisms of the institution. Somehow it seems like a framing thing."

So JP, perhaps you can explain what some of the phrases in your passage mean.

For instance, what does it mean for "facility " to be "in favour"? What I'm wondering is why couldn't bank capitalized by the Greek government make use of the facilities open to any bank generally? Is that "in favour"? And if such a bank could make use of the facilities of any bank generally, then the Greeks could 'sell' an unlimited number of bonds to such a conduit which could then post them as collateral per the normal rules.

Nick:
1. Consider a bank that has issued 300 checking account dollars, against which it holds 100 paper dollars (issued by the central bank) and various IOU's worth $200. If all 300 checking account (ck.) dollars want to be redeemed at once, the bank can sell its $200 of IOU's for 200 of its own ck. dollars, and then redeem the remaining 100 ck. dollars with its 100 paper dollars. This bank is solvent and does not need a lender of last resort.

But what if the IOU's have fallen in value to the equivalent of 199 paper dollars, maybe because of liquidity problems? Well, whatever the cause, the private bank is now insolvent. If the central bank lends 200 paper dollars to the private bank, with $199 of IOU's as collateral, then the central bank has just given a $1 gift to the private bank. The private banks is out of trouble, but the central bank does not have unlimited wealth. Of course if the central bank only lent 199 paper dollars against the $199 IOU, then the bank is still insolvent, and the first 299 ck. dollars will be redeemed at par, while the last ck. dollar is worthless.

Somebody has to lose $1. That $1 loss can be transferred from the private bank to customers or to the central bank, but it doesn't just disappear because the central bank made a $1 gift disguised as a $200 loan.

2. When I say 'irredeemable', I mean that the above bank announces that it will no longer pay paper dollars for its ck. dollars, but it will exchange its IOU's for its ck. dollars. In my lingo, the bank suspends physical convertibility, but maintains financial convertibility. Such a bank must have assets (paper + IOU's) sufficient to buy back all its ck. dollars at par, or its ck. dollars will lose value. Note that once 200 ck. dollars have refluxed to the bank in exchange for the bank's $200 of IOU's, people will start to care about physical inconvertibility. If there are dollars still wanting to reflux to the bank, the bank must resume physical convertibility or the ck. dollars will lose value.

Mike: If I tried to sell my house in the next 10 minutes, I don't think I could get very much for it. If I tried to sell it in the next moth, I could get a lot more.

Also, my problems in trying to sell my house quickly would be exacerbated if a lot of other people were trying to raise cash because there were a sudden increased demand for money. Given 10 years, the price level might fall to increase the real supply of money to match the demand. But the role of central banks is to increase the supply of money in such circumstances, so we don't have to face the disasterous consequences of needing the price level to adjust instead.

" But the role of central banks is to increase the supply of money in such circumstances, so we don't have to face the disasterous consequences of needing the price level to adjust instead."

What do you mean by "money" when you say "a sudden increased demand for money"?

In a crisis, I panic, call up my broker and say "sell! sell!".

All he does is sell the stock I held and immediately roll the proceeds over to a money market fund, or deposit with a bank, or some other cash-equivalent.

I do not hold more currency than I had before, so I do not have a desire for currency per se.

I have a desire for very safe, liquid assets -- government guaranteed, risk-free assets.

In that case, when the central bank buys one government backed asset and exchanges it for currency, it *does* not increase the supply of "money" -- what I want to hold --all it does is change the portfolio composition of the pool of assets that I want to hold.

It's like saying that when faced with a sudden desire for food, the government will replace bread with rice. That doesn't help at all!

The only way for the government to increase the supply of the assets that I want to hold is to either start issuing more debt or more currency -- fiscal deficit spending, which is too slow -- or to start extending government guarantees to private sector debt, so that I am willing to hold it. Or equivalently, to buy up the junk no one wants to hold and replace with cash or government guaranteed debt. This is why you heard the push for TARP and related attempts to have government step in and guarantee or buy risky assets.

But replacing a 100% safe and liquid treasury bill with newly printed currency does not add to the demand of the stock of things I want to hold. It's important to realize the impotence of that type of operation, and misunderstandings about what people do and do not want to hold in a crisis are at the heart of this.

Nick:

One bank might get robbed of $1. Another might suffer a $1 default by a borrower, and another might be forced to sell its assets into an illiquid market for $1 less than what they could get if they had 30 days. A loss is a loss, and it is no less real when it's caused by the bank's maturity mismatching. If that bank ends up with $300 ck. dollars backed by assets worth 299 paper dollars, then a loan from the central bank must include a $1 gift to the private bank, or the private bank will face a run.

That $1 gift leaves the central bank with less assets per paper dollar, so the paper dollars will lose value. This is especially clear if we imagine that the 'central' bank was actually a private bank whose paper dollars happened to be widely used as reserves by private banks. I imagine you'd claim that the central bank doesn't need assets the way a private bank does, but of course you know we'll disagree on that one.

"That $1 gift leaves the central bank with less assets per paper dollar, so the paper dollars will lose value. "

Huh? The value of the paper dollars is a function of the supply and demand for goods. Not any "assets" backing those paper dollars. Those assets are more paper dollar assets.

According to your view, if tomorrow, someone "robbed" the central bank and "stole" the electronic records of the paper dollar assets "backing" the paper dollar liabilities, then we must have hyper-inflation, independent of what happens to people's incomes or how much it costs to produce real goods.

A strange view.

I say that unless people's incomes rise or it becomes more expensive to produce goods, then we could just create some new electronic entries for more paper assets to back the paper liabilities and no one would care about the software update.

RSJ:

Then can you name a single example of a central bank that has ever issued money (of positive value) without holding assets against that money? Nick couldn't, last time I asked him.

Which view is strange: The one that says that paper money has value because it is backed by the assets of its issuer? Or the one that says that money, alone among all financial instruments, has value in spite of having no backing, even as the backing is sitting right there in the vaults of every central bank that has ever existed?

Mike,

Central banks *create* money and *buy* some financial asset with it. As a result, the private sector has less of whatever the central bank bought and more currency, and the central bank has a liability -- the currency it issued, matched by the asset -- what it bought with the currency. If tomorrow, the central bank wanted more assets, it would create more currency and buy them. If it wanted less, it would sell the assets. So the CB has a lot of flexibility in holding whatever level of assets it wants, and for this reason we put some legal limits on what the CB can buy -- we do not want it to own everything, for example.

The currency has "utility" because it can be used to extinguish debts, and if you want to produce something -- or even be a citizen -- you will incur debts and will need to extinguish them. Once you have something with utility, you only need scarcity to ensure value, and the fiscal arm of government does not supply currency to the private sector in unlimited amounts, neither does the CB arm of government purchase unlimited amounts of assets -- the big variable is in the private sector's credit granting capacity, which is responsible for most of the screw-ups.

"alone among all financial instruments, has value in spite of having no backing,"

I am not sure what you mean by "backing". The value of a financial claim comes from an expectation of future return -- a discounted set of cash-flows. This makes cash unique, in that there is no time shift so the "flow" is pretty boring.

Jon, sorry to dump all the legaleeze on you with no further explanation.

You've probably found a loophole. Article 123 prevents the direct monetary financing of governments, but this can still be carried out through the public institutions (ie banks) owned by those governments ie through a conduit, as you describe it.

Which reminds me of a paper I recently read. A number of West African countries (Cote D'Ivoire being one of them) used banks owned by their government to get around a law similar to Article 123 in order to get monetary financing from the Central Bank of the West African States. The end result was a devaluation of the currency.

Anyways, I guess the question is... does Greece have such a bank, and could it submit enough Greek debt to the ECB to adequately finance the Greek government.

RSJ @10.37:

OK. Suppose there is a "run" on some asset. What doe the sellers of that asset want to hold instead?

It might be currency, as when there is a run on a bank, and people withdraw their demand deposits. In which case the CB prints currency.

But it might not be currency, it might be (say) TBills, as you say. This is perhaps likely if it's a run on (say) commercial bonds. The CB prints currency to buy the commercial bonds, then might subsequently want to sell TBills to buy back the currency.

It all depends. As a lender of last resort, the CB can buy illiquid assets for unlimited amounts of monetary base - the most liquid asset. There may or may not be a subsequent desire by the central bank to readjust its asset holdings subsequently, if people switch from the monetary base into TBills, pushing down the yields on TBills.

Money, as medium of exchange, is special. It is what people always accept when they want to sell something. Whether they want subsequently to hold it is another question.

On the RSJ vs Mike argument over "backing": I'm just going to sit back and watch, since I have already had my argument with Mike on this question. If anyone is interested, it's here:
http://worthwhile.typepad.com/worthwhile_canadian_initi/2009/10/why-do-central-banks-have-assets.html

JP and Jon: another interesting exchange I'm just going to watch.

Oh, RSJ, one more argument to "back" you up: closed-end mutual funds often sell at a significant discount or premium to Net Asset Value (the value of their "backing"). Would it be any surprise if money, the most liquid and most special of all assets, just because it is the medium of exchange, also had a market value independent of its backing?

JP, I'm on Jon's side here, both with thinking that a conduit exists for the ECB to fund Greece, and in not knowing ECB regulations well enough to be sure. But you wrote:

"[Jon] you've probably found a loophole. Article 123 prevents the direct monetary financing of governments, but this can still be carried out through the public institutions (ie banks) owned by those governments ie through a conduit, as you describe it ... the question is... does Greece have such a bank, and could it submit enough Greek debt to the ECB to adequately finance the Greek government. "

What about Greece issuing zero-coupon bonds, which are then bought by Greek banks (public or private), and then repo-ed to the ECB with some small haircut. The funds received by the Greek banks can then be used to buy more zero-coupon bonds issued by the Greek government, which then repo-ed to the ECB (lather-rinse-repeat). If the repo operation is long enough (say 1 year), the bonds will mature while the ECB still holds them and the repo operation will automatically unwind.

As long as the ECB maintains a credible commitment to accepting Greek zero-coupon bonds indefinitely irrespective of credit rating, there is no credit risk, as the ECB is guaranteeing that it will effectively fund Greece. Depending on the price the bonds are sold for, and the haircut, the Greek banks can even make a profit.

I suggested this before, and you've replied before, but I didn't understand why you think the above won't work.

RSJ (and Nick):
You are clearly not in the proper frame of mind to consider what must seem a fantastically unlikely proposition: that the quantity theory is wrong and the backing theory is right.

Here's how my mind was changed: I started graduate school at UCLA in 1978, learning micro from prof's. Hirshleifer, Alchian, and Demsetz, and macro from Prof's Thompson, Darby, and Leijunhufvud. I liked micro enough that I was able to hold my nose and get through the macro. The only part of macro that even approached a coherent theory was the quantity theory. I taught economics (including the QT) at CSUN starting in 1980, and at UCLA starting in 1993.

The QT bothered me more and more over the years. How to define money? Do private banks have the same effect on the price level as counterfeiters? Does the central bank really get a free lunch? Why do central banks hold assets? How about when foreign currencies cross borders?

In 1989 it occurred to me that a checking account dollar is a call option on a paper dollar, and since calls don't affect the value of the base security, it follows that the issue of checking account dollars does not affect the value of base dollars. After 5 years of pondering, plus reading economists like Tooke, Fullarton, Bosanquet, Viner, Mints, Hayek, etc., I had realized that the QT was not just a little wrong, but completely wrong, while the backing theory was a completely satisfactory theory of money. I was finally able to put the theory down on paper in "Backed Money, Fiat Money, and the Real Bills Doctrine", which you can find by clicking my name above. (I have also found, unfortunately, that this type of writing is beyond the pale for academic journals.)

To put it mildly, I look at money in a different way than you do. So when you say things like:

" If tomorrow, the central bank wanted more assets, it would create more currency and buy them. If it wanted less, it would sell the assets. So the CB has a lot of flexibility in holding whatever level of assets it wants"
I'm a little stunned. Do you not realize that as the central bank buys more assets, it also takes on more liabilities in the form of the money it has issued?

When you say "we put some legal limits on what the CB can buy -- we do not want it to own everything", I have to ask: Have you never heard of the Law of the Reflux?

When you say "if tomorrow, someone "robbed" the central bank and "stole" the electronic records of the paper dollar assets "backing" the paper dollar liabilities, then we must have hyper-inflation", I think that you sound unaware that the central bank is just a branch of the central government, and if the central bank were robbed, the government would bail it out.

And when Nick says "closed-end mutual funds often sell at a significant discount or premium to Net Asset Value " I wonder if he has never heard that there is no such thing as a free lunch.

You even say whoppers like this: "I am not sure what you mean by "backing", as if you have never looked at the assets side of the central bank's balance sheet.

I'd bet a lot of money that neither of you have ever read Fullarton et. al., or anything else by any reputable defender of the real bills doctrine or backing theory. It is worth serious study and you have not studied it.


Kosta, I guess the main question I would have is... why would Greek banks buy 0% bonds from the government when Greek debt yields 16% (or so?) in the secondary market? A private bank has a fiduciary obligation to its shareholders to be a good steward of the firm's capital, I can't see it ever taking this trade.

You could make the argument that only a public bank run by bureaucrats would be daft enough (and unaccountable enough) to buy 0% bonds when 16% beckons. In that case your Greek financing scheme would have to be run through the Greek public banks, if they exist and are large enough - this is Jon's loophole.

"closed-end mutual funds often sell at a significant discount or premium to Net Asset Value (the value of their "backing"). Would it be any surprise if money, the most liquid and most special of all assets, just because it is the medium of exchange, also had a market value independent of its backing?"

Yes, the price of CEF diverges from the net asset value, just as the price of firms can be lower than their book value. This is because there is no way for the investor to force liquidation, and there is a belief that those prices will fall even lower, due to the management of those of assets. Or perhaps the act of forcing liquidation (e.g. via a shareholder lawsuit) would lower the collateral value even more. That is rational. But it is impossible for the price of a dollar to deviate from its dollar value. Only claims on future dollars can fluctuate in dollar terms.

Look, this is something pretty simple.

In Mathus' day, in Ricardo's day, and to a lesser extent, even in the era of Keynes, there was indeed a demand to hold currency as a store of value -- people with mattresses stuffed with gold and socks buried in the backyard. Over time, that phenomena became less significant due to things like deposit insurance and the changing architecture of our payments system. Perhaps in rural China or other places there is still a strong enough demand to hold the medium of exchange that it is economically significant enough to be an explanation of gluts. But in a modern economy you can no longer view the demand to hold the medium of exchange as a source of general gluts, because this demand is effectively zero.

So it's important when reading Malthus or Keynes that you separate out the timeless trade-offs that they were describing from the actual form that those trade-offs take.

That will be specific to the financial architecture of the economy, and you will get into trouble talking about "an increased desire to hold money balances" if you mean that literally. Depending on your personal integrity, you will either get into a lot of confusion trying to redefine "money" so that the statement is true, or you will believe the statement even though it is not true. I think, to your credit, that is why you spilled a lot of ink trying to decide what "money" was -- getting into all sorts of talmudic disputations about whether credit card balances counted as money, etc. You were trying to re-define money so that the statement remained true.

Instead, *change the statement* to "increase demand to hold safe financial assets", and, if in the future, we have a different architecture, you will need to re-interpret the statement again. Unfortunately you have to keep up with both the economic theory as well as knowledge of the financial system in order to interpret policy effects.

In the specific case of the modern industrialized world, there is never an increased desire to hold the medium of exchange. That ended with the second world war.

Now, there is only an increased desire to hold government backed securities, of which cash is just a special case, in which case CB OMO operations do not affect this. That is why we got TARP, FDIC bonds, TALF, TAF, and PDCF. The common thread behind all of those programs was to either have the government extend guarantees to private sector assets (increasing the supply of government backed assets), or for the government to increase the credit-worthiness of private sector issuers by lowering their funding costs, indirectly making the issuers' obligations safer (thereby increasing the supply of safe assets).

In both cases, it was recognized that swapping treasuries for cash will not help anything, and something else needed to be done.

"As a lender of last resort, the CB can buy illiquid assets for unlimited amounts of monetary base"

The CB cannot do this -- at least not in the U.S.

The Fed has "emergency" powers to *lend*, but can only purchase outright government guaranteed assets -- all other asset are not purchased, but held as collateral for the CB loan.

As those loans are repaid, the assets held for TALF, PDCF, and TAF shrink and are put back in the private sector. The difference in ownership is meaningful, because at the end of the day the losses are not born by the Fed if the security does not perform. Suppose you have a bond that will most likely pay out 10 cents on the dollar. The mafia (the Fed) comes along and agrees to pretend with you that the bond is worth $90. You get a $90 loan from the Fed with the bond as collateral, but must repay the loan within a year, and take the collateral back. No one walks away from the Fed -- another operational reality to be aware of -- so you will still bear the losses, and the Fed has not bought your bond.

As an aside, the MBS purchase program would probably fail a court challenge, as the agency assets are not officially government guaranteed. In any case, the emergency powers would not allow it to buy "unlimited amounts". Perhaps in Canada it is different -- again the architecture of the financial system makes a big difference. Whether or not there is deposit insurance makes a big difference. You can't just read Malthus and have a literal model of gluts, independent of the financial architecture.

"Money, as medium of exchange, is special. It is what people always accept when they want to sell something. Whether they want subsequently to hold it is another question."

Exactly, but if they don't hold it, then your economic analysis changes. If they don't hold it, then OMOs do not help and something else is needed. If they don't hold it, then reserves become unimportant, as everyone immediately redeposits their funds. If they don't hold it, then you can ignore currency entirely with the assumption that assets are liquid, and focus solely on financial assets of differing safety.

And it's important for the macro people to keep up with the times and know whether we live in a world in which people hold the currency or not.

JP,

You wrote: "I guess the main question I would have is... why would Greek banks buy 0% bonds from the government when Greek debt yields 16% (or so?) in the secondary market? ... You could make the argument that only a public bank run by bureaucrats would be daft enough (and unaccountable enough) to buy 0% bonds when 16% beckons..."

I see your point, but in this case, isn't it the ECB which is the public bank run by bureaucrats which is daft enough and unaccountable enough? You could run the system by having Greece issue zero coupon 1 year bonds with a face of 100 and the ECB make a credible commitment to accept these special-issue bonds in 1 year repo operations with a haircut of 5%, i.e., will lend 95 against the bond. Private banks, knowing that they can immediately unload whatever bonds they purchase straight to the ECB, would be happy to buy Greece's special issue for 94.5, repo to the ECB for 95, and collect the 0.5 for their troubles.

This is essentially a debt Ponzi scheme with Greece being forced to issue more bonds to cover past interest (and the ECB being willing to accept those bonds indefinitely), but as long as the ECB is willing to run repo operations at a given haircut to the face value of the bond, shouldn't the interest rates on the second market quickly converge to that haircut? And since the ECB is relatively unaccountable, couldn't they pull it off?

I found this via FT Alphaville, but the ECB already has a program to purchase covered bonds. Couldn't they expand the program to purchase covered bonds backed by gov't debt?

4 June 2009 - Purchase programme for covered bonds
Following-up on its decision of 7 May 2009 to purchase euro-denominated covered bonds issued in the euro area, the Governing Council of the European Central Bank (ECB) decided upon the technical modalities today. These modalities are as follows:

■The purchases, for an amount of EUR 60 billion, will be distributed across the euro area and will be carried out by means of direct purchases.

■The purchases will be conducted in both the primary and the secondary markets.

■In order to be eligible for purchase under the programme, covered bonds must:

•be eligible for use as collateral for Eurosystem credit operations;

•comply with the criteria set out in Article 22(4) of the Directive on undertakings for collective investment in transferable securities (UCITS) or similar safeguards for non-UCITS-compliant covered bonds;

•have, as a rule, an issue volume of about EUR 500 million or more and, in any case, not lower than EUR 100 million;

•have, as a rule, been given a minimum rating of AA or equivalent by at least one of the major rating agencies (Fitch, Moody’s, S&P or DBRS) and, in any case, not lower than BBB-/Baa3; and

•have underlying assets that include exposure to private and/or public entities.

■The counterparties eligible to the purchase programme are those eligible for the Eurosystem’s credit operations, as well as euro area-based counterparties used by the Eurosystem for the investment of its euro denominated portfolios.

■The purchases will start in July 2009 and are expected to be fully implemented by the end of June 2010 at the latest.
ECB Website

There's a ratings requirement, but that could easily be suspended. What about gearing up this program to buy gov't debt?

Greece has public banks. Article 123 does not prohibit them from using the repo facility for bonds issued in a private offering. However, you are forgetting the voluntary constraints that the system operates in for ideological reasons. Furthermore, there is jawboning going on! Last Novemebr the Greek CB ordered Greek banks to refrain from using the 1 year facility (supposedely out of a requst by the ECB). It was from that moment that spreads begun to jump, leading to the rating downgrades, the spreads rose agai and a spiral took hold!

Nick Rowe, RSJ and Mike Sproul,

It is time to review the liquidity definition of the Radcliff Report. Is it money you are debating or liquidity?

Food for thought.

There is some empirical evidence that CB short term interest policy is inversely and risk management regulation is positively related with risk taking by banks. If correct, given that the money multiplier concept is suspect by many, the CB can influence credit and private financial expansion via interest policy and risk regulation.

My God! Someone who remembers the Radcliffe Report! http://www.google.ca/url?sa=t&source=web&ct=res&cd=2&ved=0CBkQFjAB&url=http%3A%2F%2Fwww.jstor.org%2Fstable%2F1926090&rct=j&q=radcliffe+report&ei=F5HmS6e5FYKclgfc75yNBw&usg=AFQjCNEsqIk56v9TAkab7fN_lmqdsCGm3w
I can't find an ungated copy. I don't remember its definition of liquidity. But I agree that "liquidity may be at the root of the 3-way argument here between me, RSJ, and Mike.

My own (crap) definition of (il)liquidity is the ease/cost/delay of converting an asset into other goods. (It's a crap definition because "ease/cost/delay" is really loose and undefined). Nicky Kaldor (author of Radcliffe) would say it's only a difference in degree (there's a continuum of liquidity across assets, so there's nothing special about money). I say that a difference in degree can become a difference in kind. By my definition, in a monetary exchange economy the medium of exchange is the most liquid of all assets, because all other assets must first be converted into the medium of exchange before being converted into other goods. That makes money, defined as medium of exchange, special.

If I didn't know otherwise, I would guess that you were an old Brit, probably from Cambridge! ;-)

Mike: I can't remember if I have read Fullarton et al. (Actually, I can't remember much of anything nowadays). But I did read lots of Real Bills stuff maybe 15 years or so ago. Don't ask me what. It's all one big Irish Stew in my mind.

If a firm has a monopoly, the value of the firm's equity may exceed the break-up value of its assets. The greater the liquidity of an asset, the lower the required rate of return at which people will willingly hold it. Money is the most liquid of all assets, so people will hold it even at a very negative (real) rate of return. Central banks have a de facto or de jure monopoly. And that explains why the value of money is unrelated to the value of the backing.

RSJ: Take a snapshot of the economy. All currency is held by someone. People do "hold" currency, just as they hold bonds, and houses. They just hold currency for shorter periods, on average, than bonds and houses. Currency is more liquid than bonds, which are more liquid than houses. I never bought into the idea that one could meaningfully identify separate "transactions" and "store of value" demands for money. Even Keynes, IIRC, describes transactions/precautionary/speculative as three "motives" for "retaining" money.

I can't keep up with and properly respond to all aspects of all comments. Trying to write another post, on Julie Dixon, on the bank tax.

Mike: a few months ago I did click on your name and read some of your writings on backing. What you write is good, and I found it a lot clearer than previous stuff I had read on Real Bills. But I found that my previous understanding of the Real Bills/backing doctrine (I think of it as the "Equity Theory of Money") was reflected in your writings (in other words, my previous understanding was correct).

I can imagine a world in which the Equity Theory of Money is true. It is a world where either the demand function or the supply function of money is perfectly elastic at the competitive supply-price of financial assets. (Opportunity cost of holding money on the vertical axis). But it isn't true in a world where the money demand function slopes down, or there is a monopoly supplier.

Yes. Like you, many years ago I stumbled across the Equity Theory of Money, realised that it gave a very different perspective, plus a serious challenge, to the QT, and spent some time reading about it, and a lot more time thinking about it. I decided I disagreed with it. Though I think it contains some truth, but that truth can be captured within a sophisticated QT, which recognises that "backing" may, under some circumstances, affect P via its effect on expected future money supply.

I certainly have not ignored the Equity Theory.

The best thing, for me, about reading and thinking about the Real Bills/backing/equity theory is that it forced me to think more deeply about the QT. Why was money so different from other financial assets? Why couldn't we just apply the same theory to money as we do to shares? And that, in turn, forced me to re-think my theory of other financial assets (and non-financial assets). "If the standard theory of the value of financial assets doesn't work for money, maybe it doesn't work exactly for other financial assets, either?" I think it doesn't. Just for that reason alone, thinking about the alternative theory was very valuable to me.

Here's a way to think about the QT, starting from the Backing Theory.

Start with BT. The value of shares is determined by the PV of the stream of returns created by the assets backing those shares.

That PV has an r in the formula, reflecting the demand-price (demand rate of interest?) of those shares. (The 'r' at which those shares are willingly held).

Now, suppose that the shares of one company are special in some way, and that the 'r' is positively related to the total value of the shares outstanding (the company's market cap?).

And suppose the company has a de facto or de jure monopoly on whatever it is that makes its shares special.

There is then a downward-sloping demand function for the total value of that company's shares, as a function of the difference between the market rate of return and the rate of return on that company's shares.

That's the modern QT.

What's the thing that's special? Money is the medium of exchange, the most liquid of all assets.

Put it another way. If you wanted to say that QT and BT emphasise different aspects of the same underlying more general theory, I could live with that.

BT teaches QT that backing may matter, and will matter if future M depends on backing. And that backing may be endogenous with respect to changes in current M.

QT teaches BT that the "demand-rate of return" on M, relative to the rate of return on other assets, may [EDIT will] be endogenous with respect to current M (money demand functions may slope down).

"Private banks, knowing that they can immediately unload whatever bonds they purchase straight to the ECB, would be happy to buy Greece's special issue for 94.5, repo to the ECB for 95, and collect the 0.5 for their troubles."

But again, why would a private bank buy the issue for 94.5, repo it for 95, and collect 0.5, when it could buy existing bonds at something like 80 cents on the dollar, repo them for 95, and collect 15. The debt issue will fall flat.

"Couldn't they expand the program to purchase covered bonds backed by gov't debt?"

Yes, that is the nuclear option. But Trichet has refused to turn to it, rightly so I think.

Which is what Panyotis brought up; the importance of voluntary constraints (vs legal constraints). If banks buying newly issued Greek debt and repoing are perceived by the ECB to be engaging in monetary financing, their access to repos may be cut off.

Most of the loopholes you are trying to find would never be allowed (in significant quanitities) because the ECB is run German-style, and Germans care about good money.

Nick:

Sorry for getting cranky back there and accusing you of not having studied the backing theory. Obviously you have, and I have to admire your good manners for not responding in kind. (But I still recommend reading John Fullarton, Regulation of Currencies, 1844.)

Prior to Feb 26, 1797, the British pound was convertible into about 1/4 oz of gold, and the Bank's assets (gold + bonds and other securities) were approximately enough to buy back all the pounds it had issued. (Actually assets were a little short, as evidenced by a continuing run and gold drain on the bank.) On Feb. 26, the Bank suspended physical convertibility, meaning that it would no longer buy back its pounds with gold, though it still stood ready to buy them back with bonds. That is, the Bank maintained financial convertibility.

Clearly, the Bank's assets were not much different on Feb 27 than they had been on Feb 26. But the bank no longer paid out gold for its pounds. Did the pound suddenly become fiat money overnight? Obviously not. After all, the Bank had always suspended convertibility every weekend, and people had always valued the pound anyway because they knew the assets were still there, and they expected the bank to re-open on Monday. This suspension was just longer than usual, but people knew that the assets were still there.

Suppose people expected the suspension to last for 1 year. Everyone expects the pound to be redeemable for .25 oz in 1 year, so the most the pound would have fallen is about 5% (the interest rate). But this ignores the cost of issuing and handling pounds. We all know that if an ordinary bond is issued in a R=5% world, while the costs of issuing and handling that bond are 3%, then that bond will yield only 2% over the year. Suppose, for the sake of argument, that the pound costs 5% to maintain in circulation. Then the pound would sell for .25 oz throughout the year, and it would appear to bear no interest, though the fact is that the interest it bears is just burned up by the cost of printing and handling.

Now as a matter of fact the pound did not drop at all on Feb 27, and it was pretty stable for the next few years. But by 1810 some inflation (3-5%) had set in, and people (e.g., Ricardo and Bosanquet) started arguing about the cause. This era, as far as I can tell, was when the idea of fiat money took hold. But why? Simply because the pound was physically inconvertible, and people wrongly equated 'inconvertible' with 'unbacked'. (Note that if you were right about monopoly of note issue causing paper money to sell for more than its backing, the pound should have acquired some premium over and above its asset backing. It didn't.) The Bank actually restored convertibility in 1821, something it could not have done if it had not kept all its assets.

Do we have any better reason to believe that the modern paper dollar is fiat money? Absolutely not. The dollar is now physically inconvertible but financially convertible, and the issuing central bank still maintains enough assets to buy back all the dollars it has issued. Exactly the same as the old Bank of England.

Now throw in the paradoxes raised by fiat money: (1) Its issuer gets a free lunch (2) Private banks have the same effect on the price level as counterfeiters. (3) When US dollars invade Canada, the US dollar gains value and the Canada dollar loses value. (4) There has never been a bank that issued paper money without holding assets against it. This list could get a lot longer.

Your argument for believing in fiat money seems to hinge on your belief that people have a downward-sloping demand for paper dollars. But think of the substitutes for those paper dollars: checking accounts, credit cards, foreign currencies, book credit offsets. It's a long list.

Finally, you might consider the empirical support for the Backing Theory offered by Sargent, Smith, Cunningham, Calomiris, Bomberger, Makinen, Siklos, etc. Whenever someone does an empirical study of the BT vs. the QT, it's the BT that wins.

Mike: I vaguely remember reading about that suspension of convertibility episode. Maybe David Laidler had us read Fullerton?

Here's my QT-theoretic interpretation. There's a downward-sloping demand curve for currency. But the BoE prices pounds at marginal cost (you get 0% nominal return on pounds, and 5% on bonds, but the costs of producing and maintaining pounds in circulation are 5%, so the BoE is not exploiting its monopoly power. (Or maybe other banks could also produce currency then, so even though the demand curve for currency sloped down, the demand curve for BoE notes was roughly horizontal?).

The demand curve for bonds was roughly horizontal at 5%.

The BoE suspends gold convertibility, but still allows notes to be converted into bonds on demand. So even if the downward-sloping demand curve were to shift left or right, the quantity of notes supplied would adjust accordingly, and the value of pound notes would neither fall nor rise, and we should never observe its downward slope.

(Think about what it would take for an econometrician to be able to estimate the slope of a demand curve. You need an upward-sloping supply curve (or, at least, a supply curve that is not horizontal). When you refer to "Smith" above, you mean the late Bruce Smith? I remember reading one of his empirical pieces on BT vs QT. But he was looking at an episode with fixed exchange rates, so the quantity of money was demand-determined. This is why his econometrics could not identify the QT effect.)

If the BoE had forced an increase in the quantity of notes into circulation during this period, and had not allowed people to convert them into bonds, the QT would predict the value of those notes would fall against gold, but only a limited amount, until people expected them to appreciate and so earn the same rate of return as bonds going forward.

"Your argument for believing in fiat money seems to hinge on your belief that people have a downward-sloping demand for paper dollars. But think of the substitutes for those paper dollars: checking accounts, credit cards, foreign currencies, book credit offsets. It's a long list."

Yep. At one extreme, if the money demand curve is vertical, we get the old version of the QT, MV=PY, with V exogenous with respect to expected future money creation, so any notion of backing is irrelevant.

At the other extreme, if the money demand curve is horizontal (at the competitive rate of return), then it's pure BT. The current supply (as long as backing adjusts in proportion) is irrelevant.

Just in my lifetime, in Canada and the UK, I have seen the opportunity cost of holding currency vary between 0.25% and 21% (Canada, 1982, IIRC). And yet people still held the stuff, even at 21%. And what is so amazing about Zimbabwe is not that the Zim dollar eventually got wiped out by foreign competition, but that it took so long, and such incredibly high negative real rates of return, for that to happen. People were still accepting and holding (albeit briefly) Zim notes, even during hyperinflation. The money demand curve is not everywhere horizontal.

The BoC earns about $2 billion monopoly profits per year, because it has moved up along the downward-sloping demand curve for its currency. There's the free lunch. Partly government laws, partly network externalities, creating that monopoly profit. If private banks issued money that were a perfect substitute for central bank money, paid the same rate of return, and if the central bank did not reduce the supply of its own money by the same amount, they would have the same effect on P as counterfeiters. Same with US dollars. I gave my own explanation of why CBs hold assets in my old post. My guess is that there's probably some weird case where a CB has disappeared (e.g. war), but its notes still hold value, until a new CB gets invented.

Nick: The suspension period was intensively covered in econ classes up to the mid 1960's. The book grad students used the most was Ashton & Sayers. That whole book was anti-backing theory.

Yes; Bruce Smith. He wrote about colonial American currency and concluded that the currencies were valued like equity claims against the issuing colony, where the colony's 'taxes receivable' was the main asset backing the currency.

"And yet people still held the stuff, even at 21%."

If the average person carries $100 in currency, then he's paying just $21 for a year's worth of liquidity. It's not so amazing that people would continue using currency. And most other forms of money bear interest in rough proportion to inflation, so they're protected.

Assume you're right about monopoly enabling the central bank's money to trade at a premium over asset backing. How much of a premium? 10%? Then wouldn't that make you 90% backing theorist and 10% quantity theorist?

I think that an arbitrage argument shoots down the idea of currency trading at a premium. Let's say that some firm has issued 1000 shares of stock, and its only asset is a $60,000 bank account. Each share must trade at $60. If those shares traded at a premium, say $61, then a short seller could make an arbitrage profit.

Short sellers should be able to make the same kind of arbitrage profit by shorting an over-valued currency. Or even easier to think about, suppose the Zim$ trades at a premium relative to its asset value. Any foreign country could circulate its own (supposedly fiat) money in Zimbabwe, thus reducing the value of the Zim$ and getting a free lunch for itself. That free lunch would be a huge incentive for the foreign country to make its own currency circulate in Zimbabwe, and of course Zim traders would be eager to get that foreign currency.

@Nick: " My guess is that there's probably some weird case where a CB has disappeared (e.g. war), but its notes still hold value, until a new CB gets invented."

http://en.wikipedia.org/wiki/Iraqi_Swiss_dinar

Perhaps an even weirder case: the CB did not disappear, the ability to print the existing notes did - and so the existing notes retained value, despite having explicitly no backing (one could argue they had negative backing, since holding/using them was technically illegal, but this argument probably wouldn't hold).

JP,

"But again, why would a private bank buy the issue for 94.5, repo it for 95, and collect 0.5, when it could buy existing bonds at something like 80 cents on the dollar, repo them for 95, and collect 15. The debt issue will fall flat. "

I would think that if the ECB announced a haircut ahead of time, the price would quickly converge to that haircut. Why would some third party sell their existing bonds to a Greek bank for 80, when they know that the ECB will repo for 95? Instead the third parties would all be demanding prices close to the haircut, and as long as there are sufficient banks competing to purchase the debt, they would eventually get that price. Since the ECB has deeper pockets than everyone else, wouldn't the ECB be setting the price? There undoubtedly would be some volatility at initiation of the program, but that would pass.

I agree with your point that it would take a change of character by the ECB for any of these loopholes to be instituted. Thank you for your responses.

GA: good find!

Mike: "Assume you're right about monopoly enabling the central bank's money to trade at a premium over asset backing. How much of a premium? 10%? Then wouldn't that make you 90% backing theorist and 10% quantity theorist?"

That's a damn good question. It's the right question to ask in general about theories, or about 2 competing theories. Not "Which theory is true?" but "what percentage of the truth does each of the two theories capture?"

It's also a hard question (for me) to get my head around. I've been thinking about it. Here's my best shot:

Let's ask your question *at the margin*. Suppose we start in equilibrium, then backing increases exogenously by 1% of the real stock of money, M/P, (the BoC finds some bonds behind the sofa). And all the revenues from that extra backing flow to the holders of money (they use the interest on the bonds they have just found to withdraw money from circulation, which is equivalent to paying interest on money in terms of its effect on M/P).

Assume bonds pay 5% interest. The newly-found bonds are 1% of M/P, so that enables the BoC to withdraw M from circulation at a rate of 0.05% per year, which would lower inflation by 0.05%. Equivalently, they pay 0.05% extra interest on M. Assume the interest semi-elasticity of the demand for M is 0.7 (MVLS Minimum Variance Lucky Seven estimator). That means a 1 percentage point increase in the yield on M causes Md to rise by 0.7%. 0.05 x 0.7 = 0.035, so M/P would rise by 0.035%. Multiply by 100 (since we only increased backing by 1%) means the Backing Theory is 3.5% true? Depending on elasticity of course, and the more interest-elastic is Md, the more the BT is true, which makes sense.

Not sure if that makes any sense at all. Or even if my arithmetic is right.

"I think that an arbitrage argument shoots down the idea of currency trading at a premium. Let's say that some firm has issued 1000 shares of stock, and its only asset is a $60,000 bank account. Each share must trade at $60. If those shares traded at a premium, say $61, then a short seller could make an arbitrage profit. "

This is THE classic argument, it's also false. I suppose the closed-end fund puzzle is the most immediate counter-example. The palm/3COM story another.

However, more deeply, institutional and legal structure matters here. Suppose the share just trades at $61 forever, where's the arbitrage? If the share sells at $59 then perhaps someone with enough leverage could just buy the company and close it down to complete the trade, but the short seller? How does he force the issue. And if the company is worth $60bln then you may not be able to buy it all.

Isn't commercial banking just one big arbitrage trade? They sell dollars naked short (i.e. they accept deposits) and go long bonds and stuff.

And in a bank run, they all try to cover their short positions at once.

no, it's not an arbitrage in the strict sense since the trade is not risk-free. the proper definition if an arbitrage is a trade that has cost less than or equal zero, payoff that with probaility one is greater than zero and either the inequality is strict for the cost (you get money up front) or the inequality is strict in some positive probability subset (you get a strictly positive payoff with positive probability).

borrowing short and lending long is collecting a risk and/or liquidity premium. it is not an arbitrage.

In the example given, suppose the money in the bank was earning interest and the company paid out the interest in dividends. Then shorting the stock at 61 and investing the proceeds at the same interest rate looks like an arbitrage, you pay the interest on 60 but receive the interst on 61.

However, in a realistic institutional set-up there is still no arbitrage. After all, the stock has no maturity date but you can't hold the short forever and the stock may go up in value and force you to cover the short at a loss or provide more margin, the possibility of having to come up with external financing means the trade already fails to satisfy the definition of an arbitrage. Moreover, the stock doesn't mature but short positions have to be covered by some fixed time, if the stock price rises by that date you can lose money. Again, no arbitrage here.

After all, this sort of "arbitrage" strategy has been tried A LOT. It's known as the carry trade, it is very risky.

In the last sentence I should have said:

this sort of "arbitrage" strategy has been tried A LOT with currencies...

I found this in a Reuter's Factbox


* National central banks do not appear to be limited on which bonds they can buy. The Bank of Italy said it has already started to buy Italian and Greek government bonds, adding that other sovereign securities could follow.

The Bank of Italy was already buying Greek bonds? Could individual central banks of the Eurozone provide material support for any one nation?

"Since the ECB has deeper pockets than everyone else, wouldn't the ECB be setting the price?"

Yes, now that I look back you're probably right on that one.

One other theoretical constraint on unlimited financing through the ECB might be that imposed by private issuers coming to market the day that the Greek government does. Seeing that investors get a free 0.5% by buying Greek debt, private issuers may sweeten the terms of their own bond offerings to compensate investors and tease them away from buying the government issue.

GA: The Iraqi Swiss Dinar was ultimately paid off in another currency, so its positive value can be explained by peoples' expectation of redemption. That's consistent with the backing theory (but also with the quantity theory).

Adam P:

Suppose shares in that company were trading at $59, when they should be trading at $60. You buy and buy and the stock refuses to rise. What then? You force the issue by going LONG and TAKING delivery of the firm (i.e., the bank account with $60,000 in it), like what you described. Now, if the stock is trading at $61 and refuses to drop, you force the issue by going SHORT and MAKING delivery of the firm. Borrow all 1000 shares. Promise the share lender that you will return either the shares themselves or the amount of money in the firm's bank account. Now sell the shares for $61,000, with the specification that you will deliver either the shares or the contents of the bank account. Deliver the contents of the bank account ($60,000) to the buyer of the shares. Now take the $61,000 and deliver $60,000 of it to the share lender. Arbitrage profit =$1000.

"Now sell the shares for $61,000, with the specification that you will deliver either the shares or the contents of the bank account."

Mike, you can't necessarily do that. Why does anyone take the other side of the transaction? You are saying to your counterparty at the outset "give ME a cheapest-to-deliver option for free". The person agrees to accept delivery of the bank acount, instead of the shares EVEN IF THE SHARES ARE TRADING $65 ON THE DELIVERY DATE?

Somebody might take the other side of: I sell the shares with the specification that I will deliver the MORE valuable of the shares or the bank account. BUT THAT'S NOT AN ARBITRAGE!

Your arbitrage relies on someone being dumb enough to write you an option but not charge the premium. That will not move prices in a rational market.

Adam P.

Pay the counterparties 1 penny and they'll take it. We're dealing with options with a strike of zero and no expiration, so there's no option premium to lose.

PS: Mike, the value of an option is always greater than or equal to its intrinsic value. Thus, in you example where the share price is $61 the value of the option to deliver the cheaper of the shares or the account is worth more than $1 per share.

You have no arbitrage here, you also, by the way, have no theory of the price level. I suppose that will turn into a long conversation though, so forget I said it.

Mike, sorry we were posting at the same time.

How is it not obvious the option is worth at least $1 per share? There are other kinds of options than vanilla calls and puts. This is basically a cheapest-to-deliver option (usually abbreviated ctd).

In your story you give the short seller of the stock the right to deliver the bank account even if it is worth less than the shares. Why does the person who loans you the shares accept that?

Also: "We're dealing with options with ... no expiration..."

No, you can't short sell indefinitely. A lender who loans you shares, charges no interest on the loan and doesn't specify a date by which you have to return the shares has basically given you the shares. This not what happens in practice.

Stock lending requires the borrower to pay interest and the loans have a maturity date.

Adam:

Since everyone agrees that you can force a profit on an underpriced share by going long and taking delivery of the firm, I'm surprised that you'd argue against the exact opposite strategy of going short and making delivery of the firm in order to force a profit from on an overpriced share. It's something that most finance professors could confirm, but it's pretty clear that you won't take my word for it.

LOL

I was also planning on going long dollars, taking delivery of the central bank, and "unlocking the value" of the assets backing those dollars, earning a risk-free profit.

Alas, other arbitrageurs beat me to the punch, and the relative price of the dollar in relation to.. the dollar rose as a result, bringing the dollar back to its fair dollar value.

On the other hand, when the assets backing the dollar fell in value, then I knew I could short the dollar and make delivery of the central bank -- it is like a closed ended fund, right? -- and in the process, I would earn an arbitrage-free profit. Alas -- even here, I was defeated by swifter arbitrageurs that myself. As soon as the dollar price of the dollar-paying bonds fell, the purchasing power of the dollar in terms of dollar paying bonds grew, undoing my short-selling attempts. It was a vicious battle, trying to determine the "fair value" dollar price of the dollar, as a function of its backing by dollar paying bonds.

But Nick assures me that a dollar can indeed be worth more than a dollar, because of the refuge value in excess of the "exchange value", while Mike argues that taking delivery of the central bank can allow one to earn an arbitrage free profit by shorting the dollar when the dollar-paying bonds become cheaper in terms of dollars -- or is it more expensive?

:P

Ok, Mike we'll agree to disagree.

BTW, I have a bank account with 10,000 euros in it. Would you loan me 12,622 USD (EUR/USD is 1.2622 on bloomberg as I type) if I agree to pay back either the 12,622USD or the bank account. It's my choice which one to pay back. If you accept I'm ready to actually do this deal (we'll adjust to a particular eurusd fixing).

Please let me know.

PS: I'm happy to pay interest on the loan. That is, if I decide to deliver the US dollars I'll pay them back with interest (as long is the rate is no greater than what I can earn from a USD denominated deposit account).

Similarly, if I decide to deliver the euros I'll include any interest the account has earned.

Adam P:

You're thinking of calls with a positive strike, which of course sell for a premium. I've been talking about calls with a zero strike, which have no premium. For example, a call that allowed its bearer to buy $1 for 0 euros will always sell for $1, and if it ever sold for $.99 or $1.01 there would be an arbitrage opportunity on either side.

Adam P

Now that I think about it, this business of options with zero strikes or positive strikes is probably not a good way to make my point. Just think of that firm with the $60,000 bank account. Say the stock is selling for $59/share, which is underpriced. The arbitrage strategy is to buy the stock for $59,000 and take delivery of the $60,000 bank account. You might ask "Who would be dumb enough to give me the choice between 59,000 of stock and a 60,000 bank account?" Easy answer: The guy who is, by assumption, dumb enough to be on the wrong side of an arbitrage. It works the same way if the stock is selling for $61,000. Short the stock and deliver the bank account. Who would be dumb enough to go along? Answer: the same kind of guy as before.

It's one thing for you to be arguing about options and whether they have premiums. That's fine. But I wonder: Do you actually think that it's possible to arbitrage when securities are underpriced, but impossible when they are overpriced?

Mike, you're clearly not thinking at all.

In the case where the I buy the entire company for $59,000 I can take the contents of the bank account because I'm now the sole owner of the company. This sort of arbitrage does happen in real life, private equity firms make their living doing it.

In the case where the market value of the company is $61,000 I have to borrow the shares from their owner and sell them short. I now owe the owner of the company his shares. You suggest I "will deliver either the shares or the contents of the bank account" but what I actually owe is the shares. Furthermore, I can't literally deliver the bank account since I don't own the company.

Now, what you apparently mean is that I deliver to the owner of the shares $60,000 of the proceeds from the stock sale instead of his shares, I don't literally deliver the bank account that the company owns (this is an important distinction if you wish to make an analogy to company with other assets besides bank accounts, things like franchise value etc.).

The question then is, if the market value of the comany is $61,000 why does the guy who loaned me the stock accept only $60,000? Why doesn't he insist that I actually return the shares I borrowed, which are worth $61,000?

And Mike, the answer to this question: "Do you actually think that it's possible to arbitrage when securities are underpriced, but impossible when they are overpriced?", is that it depends on the security and the arbitrage strategy. The reason arbitrage pricing pins down the price of, say, a call option is exactly that the REPLICATING STRATEGTY doesn't require any one agent to do something stupid.

Your arbitrage requires somebody who owns a stock worth $61 per share to accept $60 per share for his holding. Why then is it limited only to stocks that have a bank account as the sole asset. Why does the owner even need to know what the assets are? Why not go to the owner of a share of GE, currently trading at $18.21, and offer to borrow the shares with the specification that you will return either the shares or $17.21 per share, where you get to choose which one to return? How is that different?

"Do you actually think that it's possible to arbitrage when securities are underpriced, but impossible when they are overpriced?"

To say that you are performing an arbitrage operation between an asset and currency is to say (in a convoluted way) that the asset is over-priced or under-priced. Your arbitrage operation will only succeed if you manage to predict the future price movements accurately.

The unit of measurement is currency. The unit of measurement of future value -- say value 10 years from now -- is the 10 year risk-free bond. The present value of future cash-flows is therefore the price of the 10 year bond in terms of currency.

It is non-sensical to talk about "arbitrage" between the 10 year bond and currency, anymore than it is sensible to talk about "arbitrage" between the inch and the foot.

Central banks operate by creating currency, and using that to buy risk-free bonds. They can also sell some of their risk-free bonds for currency (and then destroy the currency). In this way, they can adjust the relative quantities of different risk-free assets held by the private sector.

The asset side of the CB balance sheet is the residual -- it measures the running total of currency created and destroyed in the past. It is like the markings on the wall your parents made as you were growing up. Looking at the the asset side gives you information about the quantity of currency in circulation, but it does not set this quantity, anymore than the wall-markings of your childhood set your height. Getting the causality right is important.

Neither does the asset side of the CB balance sheet set the general price level or long term interest rates. You need to make a lot of assumptions about supply and demand for goods, the ability of the private sector to create credit (e.g. velocity), propensities to consume, expectations, government fiscal policy and tax policy, etc. The asset side of CB balance sheets doesn't shed a lot of insight about these things -- it sheds no insight that is not already embedded in these other factors.

This is how the Japanese took care of their poorly performing loans: http://reversewealtheffect.blogspot.com/

Saturday, December 27, 2008
Quote:
Bubble is caused by peoples’ expectation that the price of certain asset(real estate) will rise in future, with pouring high-powered money to the asset side of economic entities’ balance-sheet. So, to solve this problem, such asset bubble on economic entities’ balance-sheet must be gotten rid of, by the new system as below.

Adam P:

You agree that if the stock is underpriced, then the owner will sell it to you for $59,000 even though he is giving you the choice to either sell the stock for $59,000 or take the $60,000 bank account. Yet if the stock is selling for $61,000, you won't admit that the owner will let you borrow it in return for your promise to return the stock or the bank account. Obviously the owner is making a mistake in both cases, but you'll allow the one mistake and not the other.

It's a nice clean process: profit from an underpriced security by going long in the security and taking delivery of the underlying assets. Profit from an overpriced security by going short in the security and making delivery of the underlying assets.

RSJ:
So people never arbitrage against central banks? It's especially easy to see for a convertible currency. Some central bank stands ready to pay $1 for 1 peso. But the Bank has only $99 worth of assets backing 100 pesos. So customers arbitrage by bringing all 100 pesos back to the bank for $1 each, until the last peso-holder is left with nothing. If that same bank suspended (physical) convertibility, then either the peso must trade for $.99, or rival banks will issue their own pesos to get that $.01 free lunch, or people will return their overvalued pesos to the Bank. If the Bank doesn't use its assets to buy those pesos back, then the pesos will lose value.

Mike, you've completely missed the point.

The comments to this entry are closed.

Search this site

  • Google

    WWW
    worthwhile.typepad.com
Blog powered by Typepad