'The Shadow Banking System' and 'Off-balance sheet finance'

These are two phrases that the conspiracy theorists like to bandy around - and our former Prime Minister popularised the expression shadow banking system in order to try and shift the blame - when I try to explain that the UK banking system is:
• nowhere near collapse. I calculated net total cash inflows from mortgage and other lending at £136 billion a year back in 2008, which might be down to £100 billion by now, as interest rates have fallen,
• would not go *pop* even if house prices fell by half, and
• is in no need of taxpayer-funded subsidies (unless your aim is merely to re-inflate the credit and house price bubbles, which would require subsidies that are so huge that the golden goose, the productive economy, would be slaughtered).

The conspiracy theorists mutter darkly that all my calculations don't include 'off balance sheet finance' or 'the shadow banking system', therefore they won't work, without offering any sort of explanation of what they are talking about - are the UK banks hiding assets or hiding liabilities? what sort of sums are involved?

The whole point is that UK banks are not omitting assets or liabilities from their balance sheets, they are wildly over-stating assets and liabilities - if anything there is too much on their balance sheets, making them look much bigger than they really are (see Note G below). Here is the consolidated balance sheet of UK banks. I'll explain all the double counting in the notes:
Notes:

A. The figure for loans to customers looks 'about right'. According to Credit Action, total residential mortgages and other household debts are £1,457 billion, plus add 50% for business lending. The Bank of England's Table C1.2 also shows £2,208 billion as at January 2010 (Excel).

B. I can't be bothered digging any deeper into this, but the 'securities for sale' (assets) are largely all the mortgage backed rubbish which they bought from other banks. Similarly each bank has repackaged its own mortgages and issued bonds secured thereon and/or the money they raised from bonds issued (liabilities) was lent on as mortgages, so the two B's are two halves of the same equation, and could in theory be netted off to very little indeed.

Some mortgages were genuinely 100% sold on and can be removed from the banks' balance sheets entirely, some were with recourse and so had to stay on. There is a theory that Northern Rock still included a lot of the former category on its balance sheet on the assumption that they were with recourse to the bank, whereas actually it should have excluded them.

C. The same logic applies to inter-bank lending, it's just two halves of the same equation. In this example, the amount borrowed from other banks is more than the amount lent to other banks, so I assume that the net figure of £121 billion is from non-UK banks.

D. The figures for derivatives looks big and scary, but I wouldn't worry too much about that either. This is because of accounting rules (see Note G).

To give a simple example of what this might be, let's imagine
i. A UK importer who intends to buy €120,000s worth of goods from the Euro-zone each year, who is worried that sterling will fall even further against the Euro, which will increase his sterling costs. So he enters into a forward agreement at today's exchange rate of say 1.2. He commits to pay the bank £100,000 next year and in return the bank will give him €120,000.

ii. A UK exporter who intends to sell €120,000s worth of goods to the Euro-zone each year, who is worried that sterling will bounce back, so he enters into a forward agreement at today's exchange rate to pay the bank €120,000 next year and in return the bank will give him £100,000.

The bank takes a small margin deposit from each business to cover counter-party risk, and even if the bank didn't hedge these payments elsewhere, you can see it runs no currency risk whatsoever - it will simply take the €120,000 from the exporter and give it to the importer; the £100,000 goes in the other direction. The bank couldn't really care less what happens to the exchange rate and still earns its commission.

But, because of accounting rules, the bank has to include £100,000 as an asset and as a liability; and it also has to include €120,000 as an asset and a liability (at whatever the relevant exchange rate is).

E. This is the aftermath of Quantitative Easing and the bank bail outs. The government gave the banks a couple of hundred billion in soft loans, included under 'bonds' (liabilities), but told them they had to invest in 'quality assets' (government bonds, natch). Then it told them it would overpay slightly for those bonds if the banks were prepared to leave the sale proceeds on deposit with the Bank of England (which is just a form of government borrowing, but as it was just one form of borrowing replacing another, is of little relevance to the outside world).

The Treasury could quite easily get a couple of hundred billion back off the banks by netting off the two figures.

F. Besides the bail out money, the banks also owed the government £5 billion in tax, but have a deferred tax asset (i.e. they can offset past losses against future profits) of £20 billion. This can be thrown in the pot with E, seeing as the Bank of England and HM Revenue & Customs are all part of HM Treasury, which is in turn part of the UK government.

G. Accounting rules discourage you from netting off closely related assets and liabilities (such as the £200,000 assets and £200,000 liabilities disclosed under D above, which don't really exist), and banks are quite happy to go along with it, because it makes them look 'too big to fail'.

The only UK banks that were sorted out properly via debt-for-equity swaps, being broken up etc are the smaller ones such as Northern Rock or the Bradford & Bingley. The point is that this would work just as well for the big banks. Of course it would have to be controlled and managed so that savers and depositors don't panic, separate topic.

Sorting out the UK banking system, part 3/3

1. OK. We assumed a catastrophic house price collapse of 50% and mass bankruptcies - or at least a debt jubilee - in Part 2 earlier today. The resulting consolidated balance sheet of the UK banking system is shown below, and as you can see, the consolidated total assets are still looking positive, albeit down from £873 billion to £36 billion.

2. The next Big Myth is that banks have to refinance £800 billion in bonds in the next few years (the actual figure of £800 billion is correct), and they are unlikely to get this so the government will have to throw another £800 billion of taxpayers' finest at them. Let's assume that the banks really can't afford to repay these loans (which does indeed seem impossible) but similarly that they can't borrow new money either.

So let's turn to our old friend: the debt-for-equity swap. As long as the underlying business has some value and will make more money by continuing under new ownership than it would from being broken up, a debt-for-equity swap is always the most viable option. As to all this Basel-style capital adequacy nonsense, see footnote D.

3. We've also got to show that neither shareholders nor bondholders are somehow being robbed, and that they are no worse off than before - as the balance sheet shows, there is no reason to assume that they would be. See footnotes below:

Footnotes:

A. The total value of all the bonds and shares was £723 billion before the hypothetical house price crash and balance sheet restructuring (see Part 2). For company law/insolvency law reasons, all the bonds would be converted into shares, and the former bondholders would acquire a majority of the issued shares. Twenty five 'new' shares would be issued to bondholders for every eleven 'old' shares (the shares would thereafter be identical, or 'rank pari passu' as they say in the trade).

B. Banks will be charging mortgage interest rates of 4.5% on average - a bit higher than now, but they don't need to worry about house prices crashing any more because they already have done. They'll still be paying a miserly 1.5% interest rate on deposits, and still have typical running costs of 1%, so their maintainable profits from the £1,800 billion average customer balances (deposits or loans) will be about £36 billion per annum. If you don't mind, I'll gloss over corporation tax liabilities and ignore any residual income from the £571 billion worth of 'securities for sale' and the income from their trading or investment banking divisions, which will net off to very little.

C. UK banks' price earnings ratios are currently between 9 and 30. Let's pencil in a price/earnings ratio of 20.1 (it might be higher; it might be lower), which we multiply by the £36 billion maintainable earnings to arrive at a total market capitalisation of £723 billion, which is exactly what is was before we started - this is hardly surprising as the markets have already factored in what will inevitably happen.

In other words, if you own £10,000's worth of UK bank bonds or 'old' shares today, once the dust has settled, you will end up owning about £10,000's worth of shares afterwards. What's not to like?

D. Of course, having shareholder's funds of £36 billion to support total assets of £2,442 billion is far too low. But we are looking at the bottom of the cycle. If we want UK banks to have a ratio of at least 6%, then they'll just have to stop paying dividends for three years, hey presto, job done. Under the circumstances, whether profits are paid out as dividends or retained in the business has relatively little impact, see also 'Berkshire Hathaway' or 'Microsoft'.

Or they can speed up this process by flogging off the 'securities for sale' for £571 billion (they may well get far more than that, of course), which reduces total assets to £1,871 billion - combined with three years' retained profits, they'd have a capital ratio of 10% which is more than adequate.

Sorting out the UK banking system, part 2/3

1. Having accepted that bonds are part-ownership and not a liability, and done all the netting off and contras from Part 1, we arrive at the more respectable balance sheet position shown below, with a healthy Basel ratio of 25% (i.e. £873 over £3,279). This deals with the first Big Myth, that UK banks are likely to go *pop* any second. I'll deal with the last Big Myth - that UK banks have to refinance or roll over £800 billion in bonds in the next few years - in part 3.

2. But let's now confront the second Big Myth - that we have to throw everything we can at propping up UK house prices, because if they fall by one single penny, the entire UK banking system will collapse. So let's do an extreme stress test and assume that UK house prices fall by half (which is unlikely to happen). I've pencilled in these write down percentages, and the resulting balance sheet will appear in Part 3 later today. Article continues below:
3. To see how undramatic the effect of a 50% house price crash would be, we have to remember that not all mortgages are 100% loan to value, i.e. if your loan-to-value ratio is 50% and prices fell by half, you would still not be in negative equity.

4. According to Table 2.17 of the Bank of England's latest Financial Stability Report, 59% of UK residential mortgages had a loan-to-value ratio of less than 50%, so subject to certain assumptions - that every borrower in negative equity declared him or herself bankrupt and handed back the keys (again, highly unlikely to happen), the total losses would be in the region of 15% to 20%. The same sort of figure applies to lending on commercial properties, and there is plenty of non-land related lending. In my 'write down' column, I have assumed a write down/loss of 20%, i.e. at the higher end.

5. Heck knows what's buried in 'Securities for sale'. It'll be a mixture of stuff that is worth what they say it is, second hand mortgages worth 80% of their face value (see 4.) and other US-origin sub-prime stuff that might be worthless. This averages out to 60% of current market value, so let's write this lot down by 40%.

5. For good measure, let's write down banks' own fixed assets - which include their commercial premises and 'good will' - by 10%.

6. You can't 'write down' customer deposits or trade debts as this would be open fraud and politically impossible.

7. And we have to keep track of the market value of all the bank shares and bank bonds in existence. Market value of the shares is explained in Part 1. I've assumed that bank bonds are trading, on average, at 80p in the £. Therefore the total 'enterprise value' of UK banks is £723 billion - the object of this exercise is to show that neither bond nor shareholders would particularly lose out if house prices crashed AND/OR if UK government bail outs were to be halted and reversed.

Sorting out the UK banking system, part 1/3

1. One of the Big Myths put about by bankers and politicians is that 'banks are too big to fail', because if you add up all their balance sheet totals, you end up with a figure of £6,000 or £7,000 billion, which is four or five times the UK's GDP. This is because every bank owes all the other banks vast amounts of money, and because of accounting rules that force you to show closely related assets and liabilities (i.e. derivatives) gross, rather than netting them off to a small, manageable figure.

2. Another Big Myth is that if house prices fall, banks will somehow disappear in a puff of smoke and savers won't get their money back (let alone bondholders or shareholders), which I will deal with in Parts 2 and 3 of today's mini-series.

3. So I have printed off the balance sheets of the five largest UK banks (see footnotes) and done all the netting off for you, which gives a more realistic balance sheet total for the UK banking system of £3,602 billion (still more than twice GDP, but that figure can be whittled down further). Article continues below:


4. I trust it's obvious where all the QE money went - straight back back into the Bank of England!

5. This balance sheet total is still overstated, so I have proposed a couple of further contra entries:

a) The UK government has lent the banks a couple of hundred billion to bail them out, which is included in 'bonds' but it also holds a couple of hundred of billion of the banks' money at the Bank of England (so it has lent money and borrowed it back again). Then there's the deferred tax asset which I can't be bothered to explain. Let's net all these off to nothing.

b) According to their individual balance sheets, total UK bank borrowing from other banks is £359 billion and total UK bank lending to other banks is £238 billion, which I have already netted down to £121 billion. We can only assume that the net figure is borrowed from non-UK banks, so let's net that off with 'Securities for sale', which includes all the mortgage-backed bonds and rubbish from other banks, primarily from the USA, i.e. repay them with their own rubbish. Also known as 'doing an Iceland'.

I'll show the effect of these contras and also look at the impact of a house price crash in Part 2 later today,
------------------------------
Footnotes:

A. Balance sheets downloaded from here:
Royal Bank of Scotland
Barclays, page 19, pdf
Lloyds Banking Group
HSBC, page 357, pdf
Nationwide, page 40, pdf

B. I didn't include Abbey, which is part of Santander, and of course Nationwide is a building society, not a bank. I only included the 53.7% of HSBC which relates to European operations but not smaller UK banks and building societies. By and large, the overs and unders will net off - Barclays has quite sizeable non-UK operations and what we get is a fair picture of the UK banking system as a whole.

C. There is of course no clear dividing line between 'customer deposits' and 'bonds', but a dividing line has to be drawn somewhere between true liabilities and ownership. For example, if you borrow £50,000 from your uncle to set up in business, and a year or two later you have also run up unpaid invoices of £50,000 it is up to the bankruptcy courts to decide that your uncle is part-owner of the business and that your suppliers are normal trade creditors.

D. I added up the market capitalisation of the four banks using Yahoo Finance's numbers to arrive at the market value of the shares. Nationwide doesn't have a market capitalisation, so I have assumed this to be £nil.