LTROpprobrium | FT Alphaville


So the European Central Bank has just announced it will restart its Long Term Refinancing Operations (LTRO) as well as begin to snap up government bonds.

But how well did the LTRO work the first time around?

Gary Jenkins over at Evolution Securities has an intriguing theory that the ECB’s mass liquidity provision, coupled with that steep yield curve, might have contributed somewhat to our current predicament. It’s not far off from the idea that quantitative easing resulted in a rally in risky assets.

Here’s what he says:

Last year, in order to ensure that the European banking sector had access to sufficient liquidity, the ECB held three one year 1% LTRO’s (as above), lending a total of €614bn, of which €442bn was lent in the June operation and is due to mature on the 1st July. At the time this was regarded as “free money”, but with the benefit of hindsight it might be that the manner of this liquidity, primarily the term of one year, was a major contributing factor to the recent problems evidenced in the government bond market.

After all, what can a bank do with a deposit which is a fixed one year term with (at the outset at least) no chance of being extended? They cannot really use much of it to provide finance to companies, but they would want to try and make a profit on it. The obvious asset to buy then would have been a risk free, liquid bond that gave a greater return than the 1% fee. With Spanish and Portuguese 1 year notes at the time both offering a sub 1% yield, there was an incentive to take advantage of the steep yield curve environment and invest in longer dated government bonds. Interestingly from the date of that €442bn injection (24th June) to the end of August Spanish 10 year yields declined from 4.164% to 3.769% and its spread to Bunds declined from 74bps to 51bps. This trend is also evident for Greek and Portuguese bonds.

Let us collate all the banks into one and then say that of the total amount of liquidity injected €400bn found its way into government bonds. Let us then say that this hypothetical portfolio consisted of three equal amounts of Greek, Spanish and Portuguese 10 year bonds. And why not? This would have been seen as a risk free way of making money at that time. However, whilst this strategy worked well at first, it has now been a disaster, with a total loss of 13.5%. Thus our pretend bank is sitting on a loss of €54bn based upon Friday’s closing prices. Plus the money is due to be returned within 7 weeks and the liquidity in such assets is more limited than it has been for years. Thus the bank is in trouble, but so is the government bond market because forced selling pressure will lead to higher and higher yields.

That is one of the reasons why the EU has had to provide such a massive support package now. It is not just trying to save the banking sector, but it is saving the government bond sector, and therefore itself.

The above analysis oversimplifies the situation and makes huge assumptions. It could be totally wrong. But even if it has elements of the truth in it (and it is not difficult to believe that banks all had the same trade on) it is an interesting lesson that all attempts to organize bail outs have consequences that are difficult to predict further down the line.

Let’s hope the ECB’s ‘shock and awe’ bail-out really is different this time. This time.

Related links:
So farewell, 12-month LTRO – FT Alphaville
The ECB as liquidity monster – FT Alphaville
This bank-engineered equity rally – FT Alphaville
Pricing risk redux – FT Alphaville