so what currency do those iraqis stealing oil for the black market demand ? if it were dollars, that would undermine your argument. if currencies are manipulated by the authorities – i think it is legitimate to enquire what goes on outside the control of the authorities.

– and those convenient 500 euro notes mentioned by ‘guest’ above. some people even allege that they were deliberately designed to corner the black market and attract the cash-only operators to support the euro. five helicopter loads of 100 dollar bills were sent from the u s to support the kurds. in 500 euro bills that would have been one helicopter. . . .

I disagree with you on the yen $. there are three big long $/ short yen positions out there — call it real Japanese private money (pension funds and insurance funds holding $/ buying more $), real Japanese public money (the MOF’s $900b of reserves) and leveraged private money (borrow yen to buy $ bonds or $ equities). the two private stocks — the real money one and the levered one — are big, no doubt. tho perhaps not as big as the public money stock. but the MoF shapes expectations — japanese real money is more comfortable being long $ (unhedged) if they expect the MoF won’t allow a huge yen appreciation.

and Macroman’s point also applies here — all data suggests that central banks haven’t been big buyers of yen (i.e. sellers of $), so there hasn’t been a big central bank inflow into japan (unlike europe). that has helped to shape prices.

May I try to develop the debate between you and Brad?….I think it may be related to the point I made above, about the academic argument that sterilised intervention does not work (and maybe also my doubts about the ability of central banks to control interest rates).

If I understand correctly, you are arguing that private sector investors are supporting the dollar as much as the central banks, because they are willing holders at the present exchange rate. Maybe you would also say that central banks cannot affect the exchange rate, because any appreciation they achieve raises the probability of capital loss and prompts selling to the restore the balance of risk and return to its previous state – which (assuming that the central bank intervention has not affected returns and other sources of risk) is at the initial exchange rate. If so, the only lasting effect of intervention is to switch dollar assets from the private to the public sector. Am I right, or do I misrepresent your view? I assume that exchange rate risk is what you have in mind when you mention the sustainability of the US CAD.

The flaw in such an analysis is that its characterisation of the cost-benefit analysis of dollar asset holdings is too simple. There are other reasons to hold dollar assets besides risk and return. You mention liquidity yourself. Also, private sector investors may also have a certain amount of wealth to invest somewhere. Both of these reasons mean that private sector investors also care about the size of their dollar holdings to some degree. They may reduce their holdings of dollar assets in response to the increase in exchange rate risk generated by the intervention-driven appreciation, but not enough to entirely undo the appreciation. I believe this is what international economics textbooks call the portfolio balance channel.

My conclusion is that you are both right. The private sector is supporting the dollar as much as the central banks, but only with a smaller stock of dollar assets than would be the case without intervention. If central bank holdings of assets are small in relation to the private sector, intervention does not work. How big a proportion of the stock central banks have to account for to affect the exchange rate depends on the utility of those assets beyond risk and return. I would imagine that the utility of dollar assets is relatively large, and also that central bank holdings of dollars are relatively large, so I can believe that intervention is holding the dollar up.

One might view the association of CB purchases of U.S. dollars with current account surpluses rather than gross capital account inflows (meaning accounting period additions of liabilities to the external balance sheet position, not flow ‘turnover’) as arbitrary as well, other things equal. It is a similar mode of associating certain assets with certain liabilities within a larger balance sheet operation – an association of convenience rather than logical necessity (an error of ‘mental accounting’).

E.g. suppose one views CB reserve increases as offsets to gross capital inflows of U.S. dollars of equal size. Then no net long position is created in U.S. dollars by the central bank – there is a chain of intermediation consisting of U.S. dollar reserves // monetary base/domestic debt // foreign U.S. dollar debt. The net currency position for this chain is flat – U.S. dollar assets offset U.S. dollar liabilities. The net long U.S. dollar position by default must then be associated with private sector investment of the current account surplus.

But the profile of the partition can be forced, at least partially, by constraints on the relative size of the various funding channels. E.g. if both the current account surplus and CB reserve accumulation are large relative to gross capital inflows, then one really must associate the investment of the net long position and influence on the currency with the central bank to some considerable degree. Perhaps such a constraint is operational/binding in the case of the major reserve accumulators.

This all assumes that the settlement of the profile of the U.S. external balance sheet is a primary driver of exchange rates. In addition to this is the velocity of turnover through nostro account settlement of foreign exchange transactions. However, such turnover doesn’t affect the aggregate external balance sheet profile per se – it merely shifts the ownership composition of the overall profile, albeit at very high velocity.

Of course, the liquidation of foreign private claims could be funded to a degree by the liquidation of U.S. foreign claims – i.e. foreign private claims could be liquidated partially without affecting the current account, and therefore without affecting the potential for official asset accumulation to fund the deficit. This could happen through the liquidation of interbank claims on both sides of the U.S. external balance sheet, thereby shrinking its overall size.

This, in turn, inspires dollar selling against more undervalued currencies- RMB, RUB, etc., which leads to ever-more dollar buying (against these pegged currencies), which in turn leads to more dollar selling (against euros, sterling, etc.)

The data you ask for doesn’t exist in official form; it has to be calculated. Basic caluclations suggest a rather impressive net flow:

http://macro-man.blogspot.com/2006/11/harry-potter-and-fx-reserve-manager-why.html

A 1% trade (fairly modest by international bond/currency manager standards) for SAFE is now $10 billion. That sort of size definitely has an impact!

1) we agreed that – by not selling $ net , even buying a little – the private sector supports the current US$ level.

2) it follows that major exrates are not determined by CBs intervention (except when CBs want to keep their currency low, as JP did in the past)

3) so if exrates are inconsistent with what they should be, it requires that private participants are irrational – but this is a strong hp, as they bet money and we don’t.

4) so there must be an alternative possibility. That is, exrates are not irrational, if one looks as he should to the forward structure.

I took our common agreement a bit further, maybe we can agree on bits of it?

Note, further, that you keep aggregating all “private”: this approach hides that many privates are piling up $s. And they do it willingly.

But if you want to limit yourself to considering the private sector in aggregate “not selling $”, I think this will enough to recognise the decisive role of mkts in setting major exrates, and avoid describing a world where CBs set ex-rates, and set them at arbitrary, irrational levels, causing globalimbalances for pure political (policy) reasons.

So this may finally open up the discussion on global adjustment and the adequacy of current exrates! Are mkts irrational? Or if they aren’t, why do they keep the US$ spot rate at levels where it is not clear that the US CuA deficit is sustainable?

See next!

Regards

the argument against this that i find most compelling is that the buildup of central bank claims can only finance a US deficit if private creditors outside the US are willing to hold on to their existing claims (in aggregate — they can sell them among themselves, but they cannot sell them back to the US — that doesn’t work from a bop standpoint). so one of the conditions that allows $500b plus in official asset accumulation to finance a $850b deficit is that all existing private creditors (and the stock of private claimso n the US is large) have to hold onto their claims. otherwise, the official inflows would finance private outflows (a la turkey in 2001, or Argentina in 01, or other cases). that obviously hasn’t been the case.
but even in stock terms, the offiical sector’s share of total extenral claims on the US economy is rising.

I think we agree on CBs liquidity preference: it declines as reserves grow; I just pointed out that this trend is negative for the US$.

On the grow of official Reserves, I know the data (thanks mostly to your blogs). To estimate their relative importance, we need to put these data in context. You offer for this purpose some data on the private sector which I’m not sure I understand:

“Net increase in private exposure: $260-$360b. It depends on the size of the estimated growth in official dollar assets.”

Are they the result of: Total US CuA deficit – Reserve growth? Then how do they add up?

Moreover, I’m afraid that these data won’t solve our difference, I’m afraid that you didn’t get my point. Which is, again, the following.

If you take the “total official net flows” and you contrast them with the “total private net flows”, of course you’ll get more or less that CBs cover between 60% and 100% of the UD CuA deficit (let’s assume 100%, for the sake of the argument): I AGREE on this! Plane macro accounting.

But my point is that this partition between CBs and Private is arbitrary, and you draw some conclusions that are in my opinion arbitrary and wrong. The truth is that ALL those (institutions and individuals, private and public) who have positive net US$ flows do “support the dollar”‘s overall value. All of them. Take on the right side those who are “$ buyers” and on the left side those who are “$ sellers”, you’ll find on the right side many many people, and central bankers will be a very small fraction. So if you want to say that CBs net finance US CuA, and private sector overall doesn’t, I say Ok. But if you want to say that CBs are decisive in supporting the US$ against the private sector, and that without CBs the $ would be 20% lower, I say: wrong (look at the yen); and “ad hoc” to boost the “irrationality” of current global trends which you dislike. No, private players determine the dollar’s overall value much more decisively than CBs. And they are not irrational. Their rationality is bounded, as is ours, but no worse than ours.

Regards