Some thoughts on Spain

Since Friday, Spanish bond yields are clearly over the 7% threshold making market access and debt sustainability a challenge. I ‘ll take the opportunity to look into more details on the Spanish balance sheet regarding other residents.

The first table includes main liabilities to RoW based on BdE balance of payments data which are only available till 2012Q1 (they will be updated on 31 July):

What is clear is that the main driver of capital flight is a drop in securities investments. Money market instruments also show a large drop, mainly in the general government sector, although the relevant positions were already quite small. In my view this is a clear case of an asset class not being considered ‘safe’ anymore (in the Gary Gorton terminology) with investors moving out of it. Both the Spanish government and economy are now viewed as a higher credit risk instrument, with funding dropping both in the bond (Held to Maturity) and money market areas.

The same is evident if one looks at monthly flows of direct investment (with portfolio investment being the one mainly hurt):

Things are much clearer by looking at more specific data from the BdE (balance sheet data for the BdE and credit institutions), as well as the Spanish Treasury:

The first observation is the significant drop in RoW deposits in credit institutions, deposits which should mainly be considered as part of money market funding. The drop in securities liabilities is much lower than the one observed in the balance of payments data, suggesting that resident banks found new domestic sources of funding (most probably with the help of BdE).

On the government debt front, Treasury data show a large drop in RoW (permanent) holdings, which is covered by an increase in holdings by domestic credit institutions. By subtracting the Net position from the outright holdings, one can calculate the outstanding repo funding from the RoW. Repo funding reached a high in September 2011 but has dropped significantly ever since, especially during 2012. This is another indication of elevated counterparty/collateral risk which leads to a large ‘haircut’ in funding.

Adding up the drop in outright holdings and repo transactions, a large part of the external position (and target2 liabilities increase) deterioration can be explained by foreign investors moving away from the Spanish bond and repo market. The repo/money market drawback is evident in the balance of payments other investment data as well:

Clearly a warning sign is the large decrease in turnover in the outright transactions (bond) and repo markets since February. This was the main pattern observed in the Greek case, with outright trading coming to a halt and price discovery moving to the CDS market with bond yields following CDS spreads (instead of the opposite) since the derivatives market became the most liquid. One thing to keep an eye for is the CDS net notional. If that starts dropping with opposite movement in the gross notional that will mean that market participants are not creating any new net contracts (which ultimately require a short position on the underlying instrument) but only leverage existing contracts by shorting CDS while covering the position with an older long position. This would indicate an unwillingness to take the original credit risk and a difficulty in covering a short position in the bond market. The intermediation trade is probably a nice arbitrage trade since collateral will be provided by the original seller or the new buyer of the contract (depending on how CDS spreads move).

This chart for the Spanish CDS net notional (from ftalphaville) is not good. Current gross/net notional is $171.6/13.7bn while they were $163.2/13.9bn a month ago.

A positive number is the total government deposits available which stand at a little over €90bn. Given the current government debt redemptions schedule and net funding needs, the government does have some leverage in the short-term until probably October (when redemptions are €27.4bn), even if it cancelled all new auctions (except for T-Bills) since most of the redemptions actually concern t-bills (which i think can be covered by domestic credit institutions in any case). With that in mind, i don’t see an immediate reason for a full bailout, at least not until the ESM is allowed to be activated by the German constitutional court in September (since financing through the EFSF for sovereign needs would mean removing the Spanish guarantees). One can probably safely acquire (outright or as collateral in a repo transaction) any debt maturing until October.

Safe assets in Europe

Gary Gorton, who has written some very informative papers on the 2008 credit crisis, recently released a new paper on the safe assets share of total assets in the US economy. The basic observation is that, although total assets have exploded from 400% of GDP in 1952 to more than 1000% of GDP in 2010, the percentage of assets considered safe has remained very stable around 1/3 of total amount, including government debt (Treasuries), central bank liabilities (bank reserves), deposits and private sector debt (mainly MBS/ABS and high quality corporate debt). It seems that, despite the financialization and debt buildup of the economy in recent decades, only a certain part of assets (debt) need to be considered ‘safe’ (information insensitive) in order to facilitate transactions efficiently. During 2008, a large part of securitized assets considered safe, became information sensitive, lowering the transactions ‘multiplier’ and increasing demand for the remaining risk-free assets (mainly government liabilities in the form of Treasuries and bank reserves). Only when the government expanded the share of safe assets (by exchanging MBS for bank reserves, providing collateralized lending facilities which acted as a temporary asset exchange, expanding government debt and guaranteeing various private sector liabilities like MMMFs) did the credit crisis stop.

The following graphs from the paper are quite informative:

What is interesting is to try and do a similar exercise in the case of the Eurozone. The latter clearly faced a similar situation with bank credit increasing in risk (due to falling house prices) and demand for safe haven assets increasing. Unfortunately, not having a common government debt instrument backed up by the central bank (like in other countries) resulted in ‘asset selection’ with periphery sovereign debt becoming information sensitive and core countries securities acting as safe haven.

Assuming that private debt securities considered safe have probably not increased since 2009, we can look at the other available safe assets and their net change since then. I am assuming that ECB bank reserves/banknotes and core countries government debt (Germany, France, Austria, Netherlands, Finland) are safe assets, while periphery debt (Greece, Ireland, Portugal, Spain, Italy) moves to the risky assets category. Net loss is calculated with haircuts of 20, 40 and 60%. I am using the latest (24 April 2012) financial statement of the ECB instead of 2011 data to include the second 3-year LTRO (numbers in billion €):

One should bear in mind that a periphery bank can get hold of safe assets basically in the following ways:

• Sell its own assets to a willing buyer in order to acquire bank reserves and/or core countries sovereign securities, which in the current market will only exacerbate the risky assets price fall.
• Borrow secured or unsecured in the money markets, something which seems to be very difficult given the periphery banks balance sheet quality.
• Borrow bank reserves from the ECB using risky assets as collateral. In this case, ECB performs an ‘asset upgrade’ without pressuring the underlying collateral market price.

As a result, the net change in ECB liabilities plays a crucial role, much more important than an increase in the total amount of low risk sovereign debt in the market. The net change due to the 3Y-LTROs is quite evident as well as the reason why they provided a near term stabilization in the money and bond markets. Nevertheless, the current market turmoil suggests that either a higher haircut (more than 40%) is applicable and/or that higher losses in bank private credit assets are expected (which is highly likely due to the deteriorating situation in most of the periphery countries, especially Spain).

A stronger supply of safe assets seems to be needed. A European Term Securities Lending Facility would probably be very effective, with the caveat that the ECB does not have a portfolio like the Fed’s SOMA, apart from the SMP assets. The problem seems to be that European sovereigns do not have a similar asset upgrade mechanism for their own debts which results in an endless vicious cycle of austerity.