Gurdgiev's big mistake - the losses on the banks' loans from the ECB and Central Bank

Brendan Burgess

Brendan Burgess

Founder
Messages
54,757
This is a very interesting issue which has been discussed at length on Seamus Coffey's blog. It is a very long thread, but I am extracting this bit, because it is similar to Brian Lucey's infamous "Anglo Irish Bank should sell their deposits". What is so annoying is that these two economics professors can get these issues so wrong. It was also surprising that Joe Durcan didn't correct Gurdgiev, but to be fair to Joe Durkan, he would have probably been wasting his time as he had already corrected him so much, to no avail.

"Constantin Gurdgiev:[broken link removed] But let’s not forget that there roughly €110 billion which the Irish banks owe to the ECB which is clearly also in some line a liability of the taxpayer. And there is a very direct liability of the taxpayer to the €50 billion that the Irish banks owe to the Irish Central Bank.

But Morgan mentions but doesn’t do any allowance for the debts of the banks which we can actually say there is an assumption made that there about 17% average losses on the banking book which is going forward for the six institutions.
We can assume apply the same 16% percent as a possible risk weighting for the €50 billion that they owe to the banks. Bang! Another €8 billion is gone. You can apply the same amount to the European Central Bank as well. Bang! €17.6 billion gone as well on top of that. "

​
This is complete nonsense. We are not on the hook for any of the money the banks owe to the ECB or the Central Bank of Ireland and creating €26 billion of debt here is a little more than disingenuous. This money the banks owe to the ECB and CBoI is a liability. The problems we have with the banks is that their assets, primarily customer loans, are not worth what was originally lent out.



There has never been a problem on the liability side, and there never will be. The problem with banks is that they are suffering huge losses on their assets. It is an accounting impossibility to make a loss an a liability.



If I owe you money how I can make a loss. You can only make a loss on a asset. You don’t stress test liabilities, you stress test assets. The only way we will be liable for any of the money the banks owe to the ECB and Central Bank of Ireland is if they make additional losses above what has already been accounted for. Looking at the liability side of the banks’ balance sheet and what we might be liable for it is nonsense.



You have to look at the assets side and we know that banks are going to make huge losses. BlackRock Consultants and the stress tests told us that. We are putting money in to cover these losses. This money to cover loan defaults, and the repayments from performing loans, will allow the banks to repay in time, or replace somewhat more quickly, the liquidity the banks have taken from the ECB.
Click to expand...
 
Let's simplify this

Let's say that I guarantee a bank's liabilities to the Central Bank

loans to customers| €100
funded by: |
loan from the Central Bank|€90
Share capital|€10
Now, let's say that many of the loans to the customers turn bad and it is estimated that they will repay only €80 of the €100.

So I recapitalise the bank by personally investing €20 in it.

The balance sheet is now

loans to customers| €80
Cash|€20
Total assets|€100
funded by: |
loan from the Central Bank|€90
Share capital|€10
Total funding |€100
So Gurdgiev making a provision of 17% against the loan from the Central Bank is incorrect. He may well challenge the €80 valuation of the loans, but that is not the issue here.
 
How serious an error is this?
It is very serious for a professor of economics. It's a failure to think through the issues.

How common is it?
I think it has been very common. I have certainly heard people talking about the fact that we have guaranteed the liabilities of the banks to the Central Bank and ECB. I never heard anyone else point out what Seamus Coffey has pointed out, and it did not strike me at the time.

I described an article by Laura Noonan as excellent in which she pointed out "taxpayers here on the hook for about €50bn of emergency cash that's been channelled through the Central Bank of Ireland (CBI) into our banks." She addresses, but doesn't resolve the issue.

The descent into ELA essentially saw the Irish taxpayer assume a massive amount of risk for the non-repayment of that €51bn given to our stricken banks.On the face of it, it seems like an outrageous development, particularly given the lack of consultation or debate around the measures -- Belgium, for example, passed an explicit guarantee for its ELA in parliament.
But sources point out that the circumstances surrounding the surge in ELA mean that the taxpayer didn't actually take on more risk, the taxpayer simply renamed a risk it was already on the hook for.
Click to expand...

Why did Gurdgiev, a Professor of Economics, make this mistake?
It's very hard to know. It arises from doing our economic commentary on the airwaves.

We all make mistakes in our analysis. We double count. We omit something. We confuse assets and liabilities. In normal times, you get a sense that you have made a mistake because you know that the result seems wrong. But people are so used to huge numbers, that the reality check didn't work on this occasion.

Why was it not corrected earlier?
If Joe Durcan had twigged this, I doubt if he could have explained it on air.

When I heard Lucey on the radion asking Alan Dukes if Anglo had considered selling the deposits, I was astonished that Dukes didn't tell him he was talking nonsense. Either did any of the other panellists, which I think included Garret Fitzgerald.

The likes of Lucey and Gurdgiev spout out this stuff with such confidence that no one challenges it.
 
While I can sort of understand Gurdgiev's error, it is important to remember that his entire analyis in the Vincent Browne show was wrong. Or as Seamus Coffey puts it:

  1. Start with a phantom debt prediction of €225 billion from the IMF
  2. Add in €31 billion for NAMA and say the losses could be more and completely disregard the assets
  3. Add €16 billion for loan losses that will appear after 2015 with no provision for the earnings of the banks
  4. Add €4 billion of bank losses that will be covered by junior bondholders
  5. Add €25 billion for some “risk weighting” on the central bank liabilities of the banks.
This gives a running total of €301 billion and this is supposed to be the analysis we need of the debt legacy from the crisis we face. There is no one that believes that the 2015 public debt will be €300 billion yet we are expected to believe commentary like this which is provided on a regular basis. Every single element of this commentary is wrong. Not some of it, all of it. It is all wrong.
Click to expand...
 
Previously I (naively) thought that academics like Gurdgiev were just academics and therefore completely objective (at least initially, i.e. before they'd a view or position to defend or justify).

But then I heard that he's only a part time academic and that his "real job" is in treasury with IBM.

Surely these guys should be made disclose their interests?

Someone who's "just an academic" talking about (say) our 12.5% tax rate is one thing, but someone who works for a multinational but doesn't overtly claim to do so gives me a sense of unease.

Introducing him as "IBM's Constantin Gurdgiev" rather than "Trinity College's Constantin Gurdgiev" would make things more transparent.
 
Hi Gekko

I agree in principle, but I think that Gurdgiev is regarded as being very independent and something of a loose cannon. I doubt very much if he would simply promote what is in the interests of IBM if he didn't believe it. Of course, his judgment may be influenced by the fact that he is employed by IBM, but, to be honest, I don't think so.

Brendan
 
The way I interpreted Gurdiev's comment is that he was looking at the money injected by the ECB and CBI from the taxpayers perspective. The taxpayer, indirectly loans money to the banks through the central banks and is liable for those funds, in your example the €90. So, from the taxpayers perspective that €90 is an asset, not a liability.

There has never been a problem on the liability side, and there never will be. The problem with banks is that they are suffering huge losses on their assets. It is an accounting impossibility to make a loss an a liability.
Click to expand...
Technically he is right but I think this comment is missing a very important point. Yes, the underlying problem for banks is that on the asset side of things the value of assets has plummeted. But there certainly is a huge problem on the liabilities side, which is the flight of deposits and the inability to borrow money on the open market in order to cover the losses on the asset side. It is very misleading to completely dismiss the liability side in this way.
Yes, from the banks perspective the money it has borrowed is a liability, but that liability is someones asset, and in this case it is ultimately the taxpayers asset, so potential losses on this are very real to the taxpayer.
 
This is complete nonsense. We are not on the hook for any of the money the banks owe to the ECB or the Central Bank of Ireland
Click to expand...

Wait, what?

Ignoring personal attacks on the motivations of Gurdgiev or Lucey, which serve no purpose in understanding the situation, this doesn't seem quite right to me.

We own the banks, so the Bank's liabilities (regardless of guarantees) fall to the exchequer. Similarly, CBI losses have to be covered by the state, i.e. the exchequer, bankruptcy is not an option for a central bank.

So where's the problem?
 
The loans aren't given to cover losses. That's not how it works. Only capital can cover losses. The central bank is providing liquidity so the banks can fund their balance sheets. The only way the taxpayer is on the hook for the money owed to the Central Bank and the ECB is if the banking system collapses and the banks assets aren't enough to cover the liabilities. (Even then technically the ECB can only come after the Guaranteed debt that was used as collateral). The banks are or at least will be capitalised to cope with the losses according to the stress tests.

It's all very taking the banks liabilities and throwing them on to the Sovereign Debt level as some sort of fait compli but why don't they be consistent and take the banks assets and put them on Sovereigns balance sheet as well. No international analysis I have seen yet has put debt at the €250-300 billion I have seen mentioned. That's not to say it isn't an ugly picture.
 
Sunny, neither the assets nor liabilities are on the sovereign balance sheet yet. What Constantin is doing is pointing out what the likely liabilities are for the state which are well known. The problem is with NAMA in the way, propping up the price of property (where's that going BTW? are we near the bottom?), we don't know what assets will be coming on to the state balance sheet.

Constantin has been right a lot more than wrong in this. He certainly didn't go an exhort people to buy bank shares or property right in the middle of the crash.
 
astarmain said:
Sunny, neither the assets nor liabilities are on the sovereign balance sheet yet. What Constantin is doing is pointing out what the likely liabilities are for the state which are well known. The problem is with NAMA in the way, propping up the price of property (where's that going BTW? are we near the bottom?), we don't know what assets will be coming on to the state balance sheet.

Constantin has been right a lot more than wrong in this. He certainly didn't go an exhort people to buy bank shares or property right in the middle of the crash.
Click to expand...

No what he is doing is projecting debt levels by picking and choosing what he wants to include and exclude. His analysis doesn't make sense. Morgan Kelly has done the same. They have not given proper explanation of where they are getting their figures from and yet are allowed to use our media outlets to put their figures out there.
 
I really dont understand this.

The IMF agree Constantin Gurdgiev figures for 2014 and as for Brian Lucey and selling Anglo Deposit books the Government have just sold the Anglo deposit book to AIB.

What's all the fuss about ?
 
Brendan Burgess said:
This is a complex subject. The best way for you to understand it is to go back and read the excellent analysis by Seamus Coffey.
Click to expand...

No, the best way to understand it is to read NAMA Wine Lake or True Economics or The Property Pin. Just because Coffey manages to come to the same conclusion that you want to believe does not make him right. Breadth of reading.

As for picking and choosing numbers to suit an argument, isn't that exactly what everyone else in the public gaze does. cf Brian Lenihan as the prime example.
 
ang said:
I really dont understand this.

The IMF agree Constantin Gurdgiev figures for 2014 and as for Brian Lucey and selling Anglo Deposit books the Government have just sold the Anglo deposit book to AIB.

What's all the fuss about ?
Click to expand...

The IMF dont agree with figures and Brian lucey said Anglo could make €20 billion by selling the deposit book and solve all their problems.
 
astarmain said:
No, the best way to understand it is to read NAMA Wine Lake or True Economics or The Property Pin. Just because Coffey manages to come to the same conclusion that you want to believe does not make him right. Breadth of reading.

As for picking and choosing numbers to suit an argument, isn't that exactly what everyone else in the public gaze does. cf Brian Lenihan as the prime example.
Click to expand...

Everybody is making assumptions but people have to stand over their figures. Morgan Kelly wrote his article and then vanished despite numerous economists and commentators pulling him up on his figures. People might agree with his general sentiment but he used figures that were were plainly wrong to come to his conclusion.
 
ang said:
I really dont understand this.

The IMF agree Constantin Gurdgiev figures for 2014 and as for Brian Lucey and selling Anglo Deposit books the Government have just sold the Anglo deposit book to AIB.

What's all the fuss about ?
Click to expand...

The IMF number for our 2014 debt is €209 billion. Gurdgiev's numbers are substantially higher. I think Anglo got about €100 million for its deposit book. Not exactly a solution to a €29 billion black hole.
 
Sunny said:
Everybody is making assumptions but people have to stand over their figures. Morgan Kelly wrote his article and then vanished despite numerous economists and commentators pulling him up on his figures. People might agree with his general sentiment but he used figures that were were plainly wrong to come to his conclusion.
Click to expand...

I'll see your Kelly and raise you an Eddie Hobbs. Not to mention Austin Hughes, Donie Cassidy and Frank Fahy.

Anyway we were discussing Constantin who, I'm pretty sure, will have no difficulty standing over his figures based on the number crunching he does on a daily basis.
 
astarmain said:
I'll see your Kelly and raise you an Eddie Hobbs. Not to mention Austin Hughes, Donie Cassidy and Frank Fahy.

Anyway we were discussing Constantin who, I'm pretty sure, will have no difficulty standing over his figures based on the number crunching he does on a daily basis.
Click to expand...

Fully agree there are plenty of worse people out there. Constantin used Morgan Kelly's article and raised it so I would love to hear him stand over it. There are plenty of economists and commentators on sites like Irish economy and nama wine lake that provide quality analysis but have no desire to see their names and faces in the media. It's these guys I listen to. Not people who write opinion pieces or appear on programmes like VB giving lovely soundbites. Strikes me that there are some academic economists who want to be the next Morgan Kelly.
 
I got these corrections by email. They don't change the argument, but for the record, I am informed:


  • Gurdgiev doesn't work in IBM Treasury
  • Gurdgiev is not a professor
  • Neither he nor Lucey work in the economics department of TCD
  • Lucey has no 'interests' to disclose
  • Not only did Dukes correct Lucey, he agreed with him and stated they had looked into it
 
You must log in or register to reply here.
Back
Top