Thursday, June 16, 2011

Responding to the ECB...again

So, the frank exchange of views with the ECB over our recent research into the ECB’s exposure to peripheral Eurozone countries rumbles on (albeit indirectly for the most part).

(See here for our previous response to the ECB)

Yesterday, the WSJ Real Time Economics blog covered comments by ECB Governing Board member Lorenzo Bini Smaghi suggesting that estimates such as ours were based on a "misunderstanding of the Eurosystem and its risk provisions" (we recommend reading the blog in conjunction with this post to make sense of it all).

Bini Smaghi makes three key points, which we try to address here: that the ECB’s revaluation account can actually absorb any losses it faces, that its stream of seigniorage from printing money adds to its financial strength and that only a default would really threaten the ECB’s balance sheet.

This seems to be a classic example of an EU institution trying to hide behind its own complexity. Frankly, we've seen it all before, and we're not overly impressed.

So here we go (excuse some of the technical language, we’ll explain it all further down).

Default assumption:

Bini Smaghi notes that both the bonds the ECB holds and the collateral it has taken on would only face losses if there was a default. Well, as you will know from reading our research, this was the very premise of our paper. Pretty much everyone, outside of the ECB and some Eurozone leaders, believes that Greece will need to default, in some form, in the near future. The ECB can continue to preach this line but we, along with plenty of others, believe they are living in a dream world. (Also stay tuned for an upcoming piece of research from us on the Greek crisis which should support this argument)

Revaluation account:

Firstly, the revaluation account was set up to help protect against losses on foreign reserves (including gold and dollar denominated holdings) resulting from changes in prices or exchange rates. As the value of these holdings changes any gains or losses are shifted to the revaluation account while the assets themselves remain listed at original prices. The idea is that these changes could be temporary and so it is financially prudent to add in this extra buffer (we agree on this point). Accepted, since the revaluation account is now over €300bn it is unlikely that all of this would be needed to buffer against price or exchange rate risks. However, it is not clear why, after being used for this purpose for over a decade, the ECB would switch to using it to account for broader credit and collateral risks. This also explains why we did not include it in our paper, since it is not defined as part of the capital base and was not created for the purpose now being suggested. Hence we stick by our comment that it would take only a 4.25% decrease in the value of assets to wipe out the ECB’s capital base (defined as capital and reserves, the standard definition).

More importantly though, this means that any money inside the revaluation account is unrealised (meaning it only exists on paper until the underlying assets are sold). Therefore, in order to cover any losses which appear in the ECB’s profit and loss account (realised losses) the ECB would need to sell some of these assets. So, the ECB would essentially be deleveraging to help cover its losses, a process which we’d expect might startle financial markets. This also means that Bini Smaghi is including the potential sale of assets as part of the capital base calculation, which seems far from normal. Including this money in capital and reserve buffers is very confusing since the revaluation account is listed as a liability while the actual holdings are listed as assets. These points seem to make it difficult for the revaluation account to be judged as a real backstop against the potential losses from a Greek default.

Additionally, since these gains are unrealised and the assets would have to be sold off for the revaluation funds to be tapped, the liquidity of the assets must also be considered. Given that demand for gold and dollars remains strong this may not be such a problem. However, these holdings will always be slightly constrained by their liquidity and will involve some transaction costs.

Furthermore, if Greece was to default we’d expect there would be significant turmoil in the financial markets, with money rushing out of the Eurozone and towards the 'safe haven' of the US. This would ultimately cause a massive drop in the value of the euro, likely wiping out a part of the revaluation account.


This is a highly technical issue, so bear with us. Seigniorage is essentially the income which the ECB and Eurosystem devise from being able to print money. Since they control the size of the monetary base now and into the future this income represents a significant financial strength, this is undeniable. However, it does have its limitations, particularly under a Greek default scenario. Firstly, it is dependent on the demand for currency, the interest rate and the inflation rate. These are all currently low, mostly due to the sluggish recovery in parts of the Eurozone, meaning that the level of seigniorage is limited. A paper published through the ECB last year notes that this income dropped to around €787m in 2009, and even at its peak was only around €2bn, not massive amounts considering our loss estimates of between €44bn - €66bn.

The ability to control future money production is important, but our point is more that the immediate hit which the ECB and Eurosystem will take from a Greek default would be passed through to taxpayers. Even with a large future potential income from printing money this cannot really be overcome without printing money immediately, which would be inflationary, as we suggest. Furthermore, if the ECB did draw on future incomes it could undermine the future financial strength of the ECB. Ultimately, our point was that the immediate hit which the ECB would take under a Greek default could cause it to need to be recapitalise or ramp up its printing of money (above the point where it is non-inflationary) and these costs would be passed onto taxpayers.

If you’ve stuck with us until now, thanks for hearing us out. Our research into the ECB was an attempt to scratch the surface on what is an incredibly opaque and complex issue. Once again, given its response, the ECB appears to remain in a state of denial over the problems it faces.

Update 16/06/2011 8pm:
As a keen observer has pointed out, the weakening of the euro would in fact increase the revaluation account. This is because the assets, such as gold or dollars, would be able to be sold or exchanged for a larger amount of euros. There might be an inflationary aspect to this, as with any devaluation, although we doubt that it would be large enough to have much impact. An admitted mistake on our part, though certainly not a vital part of our argument in any case.


Alan said...

Thanks for this. I only discovered your blog a week or so ago, and it has made fascinating reading.

I have a couple of points re the revaluation account (although I'm not an expert in central banking)

1) Would it not be possible to transfer some of the FX/Gold value from the revaluation account to the capital account by selling it (at the higher price) and immediately buying it back.

If this is so, what is to stop the ECB doing this as a "virtual" sell/buy back with no transaction costs. I know this would normally be a no-no, but they seem to re-writing their own rules.

They would be effectively re-setting the purchase cost of the foregin reserves to market value (or something below to be prudent). That would release funds?

2) If the Euro collapsed, the FX and Gold in reserves would actually be worth more (when priced in the devalued Euros), so the revaluation account is likely to increase not decrease if the currency tanks.

That may explain why the revaluation account is so high - the Euro has weakened since its 2008 high of $1.58, generating a large unrealised profit. If the ECB Mandarins expect it to regain its dizzy heights, that may be why they haven't released the funds.

Open Europe blog team said...


Thanks for your comments. Your second point is, of course correct, a misinterpretation on our part, possibly after spending too many hours wading through ECB docs. We will correct that point in our post, although we don't believe its vital to our overall argument.

On the first point, if we understand you correctly, we don't believe that such a process would release the funds as the ECB would have to buy back the asset at the higher price. It would realise the gains in the sense that it would shift the value from the revaluation account to the asset side of the balance sheet, as the asset would now be listed with its correct market value, but the funds would now be fully tied up in the asset. As such they would not provide a backstop against losses. Keep in mind that these assets aren't actually part of the capital base and so would not be scheduled to be used to absorb losses (which is essentially our first point). As far as we can see the only way to really access these unrealised gains is to sell the assets and then use the cash from the profits to absorb losses, but again this is far from what the fund seems to have been designed for (not that this has stopped the ECB before).

A very interesting point to raise and please let us know if this wasn't exactly what you were getting at.

Alan said...

Yup - you've convinced me. I hadn't really thought it through. Fundamentally, the only way to get extra capital is to sell assets, create money or cash-call.

Your point is absolutely right - do the populations of the Eurozone really expect their gold and foreign reserves to be sold off (I'm not sure the Germans I know would be too happy)?

Thinking more about the event of a default, a flight to quality may actually help the ECB with liquidity of foreign reserves. Euro money would likely go to USD and gold, and so there would probably be significant demand for the ECB's foreign reserves.

Not that a Central Bank should really be engaging in opportunistic trading. And it certainly should not be relying on that to cope with an almost certain default!