OSI, PSI, the IMF and fantasy

After much intense negotiation, it seems there is a new deal for Greece. Or at least that is how it has been presented in the media. But what sort of deal is it, and does it solve the Greek problem?

The wording of the Eurogroup's statement does not suggest that there is a "deal", as such. All it says is that the official sector - the ECB and Eurogroup - may be prepared to accept poorer returns on their holdings of Greek debt. The specific measures that they "would consider" are the following:

• 1% reduction in interest rates on the loans made available under the "Greek Loan Facility" (the first bailout)
• 0.1% reduction in the fees paid by Greece for EFSF guarantees of its debt
• deferral of interest for ten years on EFSF loans (the second bailout)
• extension of maturities on EFSF and bilateral loans by fifteen years 
• repatriation of interest paid by Greece to the ECB and national central banks (the "Eurosystem")

These measures are conditional on Greece maintaining its commitment to the deflationary fiscal programme imposed by the Troika AND successfully buying back some of the debt currently held by the private sector. I will return to the buyback shortly. But first, let's consider what these measures would actually mean for the official sector creditors and for Greece if they went ahead.

None of these measures is "officially" a writedown of Greek debt. The face value of the debt will remain the same, which will enable politicians in creditor countries such as Germany to claim that Greece is not being "forgiven" its debt. But this is misleading. The value of debt is given not just by its face value, but by the return that it generates. Cutting the interest rate (and, in the case of the Eurosystem, refunding interest paid) reduces the return on debt. This is because over time, money loses value due to inflation, so the face value of a 10-year bond is less in "real terms" at maturity than it is at issue: the interest on the bond compensates for this loss in value, plus the opportunity cost to the lender of not having that money available for other things. Cutting the interest rate reduces the compensation to the lender, and if the interest rate falls below the inflation rate during the lifetime of the bond, the lender actually loses money as the value of the principal is eroded. Extending the maturity of debt also effectively reduces the return to the lender, since that money is "tied up" for longer, increasing the risk of principal loss. Therefore these measures constitute a writedown of the value of Greek debt held by the official sector, in  much the same way that Brady Bonds reduced the value of Latin American debt in the 1980s.

For Greece, the effect is a reduction in the anticipated debt burden by 2020. Now this is something of a moving target. Sovereign debt is usually quoted as a percentage of GDP, and this is how the Greek debt burden is presented. So it is not just the actual amount of debt in euros that matters, but the performance of the Greek economy.

Each time a bailout deal has been agreed for Greece, estimates of its expected GDP over the next 10 years have been produced. But every estimate has been wrong - and not just slightly wrong, but totally wrong in both the scale and the direction of GDP growth. As this graph from Zero Hedge shows, every estimate has assumed a return to strong positive growth and creation of a primary budget surplus within a year or two of the deal:


















But the reality is that the Greek economy has been falling down a recessionary black hole: with each year that passes, economic activity declines, tax revenues fall and the prospects for return to positive growth, let alone a primary budget surplus, recede futher into the distance. The swingeing cuts Greece has made to its budget to meet Troika demands have singularly failed to make any significant dent in its deficit: all they have done is trash its economy.  And the debt pile grows ever larger, both in reality as Greece is forced to borrow more money to meet its essential spending commitments and service its debt, and as a percentage of GDP, as GDP falls ever lower.

The estimates on which this deal depends are that Greece returns to strong growth by 2015 and runs a primary budget surplus (i.e. after debt service) of at least 4.5% from 2016 onwards. If they manage to achieve this AND both the debt buyback and the official sector NPV haircut go ahead, their debt will be at 124% of GDP in 2020, which is supposedly sustainable - although many people would doubt that this is sustainable in practice. Without those measures, the debt level would be 144% of GDP, which is almost certainly unsustainable, and if Greece fails to achieve the growth and budget targets then the level will be even higher - some estimates have put it as high as 190% of GDP.

The significance of keeping debt at "sustainable" levels is that the IMF, which is a contributor to the bailout funds agreed in March 2012, cannot lend to countries with unsustainable debt. Therefore the bailout ALREADY AGREED depends on reduction of the debt burden to a "sustainable" level. You may ask how on earth the bailout was agreed in the first place if the debt wasn't sustainable. But remember what I said about GDP estimates far exceeding reality? The debt itself hasn't risen much more than expected. But GDP is far lower. Therefore a debt that in March was expected to be "sustainable", has turned out to be anything but.

So are the latest estimates of GDP any more reliable than the previous ones? I seriously doubt it. The Greek economy is now in its sixth year of deep recession and shows no signs of recovery. I suspect that Greece's GDP is falling back to where it was when Greece joined the Euro, since the whole of its economic growth during its Euro membership has been generated by debt-fuelled consumption rather than increased production of real goods and services. In fact since the Greek economy has actually lost competitiveness compared to its neighbours, its GDP may have to fall even further. In which case this chart shows that it still has an awfully long way to go:


Note: the flattening of the GDP line in 2010-11 does not indicate that GDP is static, it means that the World Bank does not have up-to-date figures. The ZeroHedge chart above shows that GDP has been on a steeply downward trend since 2008.

If I am right, there is little chance of recovery any time soon. And this also makes the target of a 4.5% primary surplus from 2016 highly unlikely. But even if Greece by some miracle managed to achieve this, the terms of the bailout are that all of that surplus plus 30% of any excess must go towards debt reduction, rather than being invested in the Greek economy. It is very, very hard to see how the Greek economy can possibly return to strong growth if all its deficit reduction efforts simply enrich its creditors. Sadly, I think it is far more likely that Greek GDP will continue to shrink, achieving primary surplus will drift further and further out of reach and the debt pile will grow ever larger. And both official and private sector creditors will eventually have to accept that Greek debt simply is not repayable. Further writedowns will inevitably follow. Not that there will be much scope for further reduction of private sector holdings; the private sector took losses of around 75% of NPV on its bond holdings earlier this year, when the Greek government coercively swapped  "new  bonds for old". And the debt buyback in the current deal is redemption of those new bonds at 28% of face value. That really doesn't leave much Greek debt in the private sector. The media headlines have focused on the official sector's contribution this time: but the private sector is being asked to take a much larger hit.

Not surprisingly, the private sector is fighting back. The Greek banking lobby has protested that the buyback at 28% would bankrupt them (link h/t @Alea_), and called for additional EFSF funding to restore their capital levels if they sell their holdings - which they feel psychologically bound to do, even though it seems unlikely that the government could impose a second coercive debt restructuring on the private sector.

This sounds odd to me. The redemption price is the average market price of those bonds at close of business on 23rd November. The bonds should already have been marked to market if they are held for trading purposes, so the banks should already have taken the hit on their profit & loss accounts. If they haven't, then either the bonds were intended to be held to maturity or they have not been marked to market properly. If the former, then redemption at such a discount would indeed cause serious losses - although it could be argued that holding distressed government debt at par, even to maturity, is perhaps not the wisest investment strategy. But I suspect the latter. The market price for Greek bonds has indeed fallen by 70% since the new issue, but the market has been thin for quite some time, which tends to make market prices volatile. Holders of Greek bonds are therefore likely to have marked them to some kind of model, no doubt using parameters that flatter the value of the holding. This is similar to the methods used for valuing CDOs in the run-up to the financial crisis, and the outcome might well be similar too.

So the debt buyback could cause widespread bankruptcy of Greek financial institutions. This seems utterly counter-productive, given that the bailout funds were partly intended to recapitalise the Greek banking system. I don't have a great deal of sympathy for financial institutions that take a grossly unrealistic view of the value of their bond holdings, but collapse of the Greek banking system as a consequence of the debt buyback would be ludicrous.

It is also by no means clear where Greece will find the money for this buyback. Ollie Rehn insists that Greece will have "all the money it needs" for the buyback, but there is no new money on the table, the 2nd bailout funds have not yet been disbursed and the "additional measures" - including the refund of Eurosystem interest payments, which seems the most obvious source of funds - are conditional on successful completion of the buyback. Maybe I'm missing something, but I can't see how this buyback can succeed with nothing but imaginary money.

In fact the whole deal looks like fantasy. A debt buyback funded from thin air and possibly resulting in bankruptcy of a newly-recapitalised banking system: yet more wishful thinking on GDP figures: official sector involvement that is dependent on meeting impossible conditions. And above all, no recognition of the reality of the Greek situation. Greece's debt is unpayable and and its economy is falling off a cliff. It needs comprehensive debt forgiveness and the equivalent of a Marshall plan to restore its economy to health, not fiscal "reforms" that drive it further into the ground with large amounts of Eurofudge to keep its official creditors sweet while stiffing the private sector.

The IMF is evidently uncomfortable with the hard line taken by the Eurogroup and the ECB, and this is the closest it has ever come to pulling the plug on the whole deal. I really wish it had done so. This Greek tragedy must be brought to an end before it becomes a global disaster.

About those UK CPI inflation figures....

The UK's CPI inflation figure for October 2012 rose by 0.5% to 2.7%. A small rise had been predicted, but this was much larger than expected. The ONS attributed the rise to increases in tuition fees, food and transport costs. I shall discuss the implications of these components shortly.

Not surprisingly, the inflation hawks were out in force. Market Oracle led with a "shock, horror" headline:

UK CPI Inflation Soars Despite Bank of England Deflation Propaganda, Shocks Academic Economists

 And they showed this chart as evidence of what they called an "inflation mega-trend":

(for larger version, click here)

And Andrew Sentance, in a news release from Price Waterhouse Coopers, made the following pessimistic prognosis:

"UK inflation remains stubbornly above the 2% target. And with further energy and food price rises in the pipeline, it could rise further in the coming months. This would reinforce the squeeze on UK consumers and add to concerns about the Bank of England's ability to achieve its price stability objective. If above target inflation persists through next year, it will add to the pressure on the MPC to raise interest rates sooner rather than later."

Well, his first statement is correct. UK inflation has remained stubbornly above 2% since 2010. But this chart from Trading Economics shows that in fact it has been above 2% on average since 2005:

(for larger version, click here)

So given that, what on earth is the tearing hurry to bring it down to 2%?

Now, I will admit that this chart does show that there has been high inflation in the last two years - as indeed there was in the run-up to the financial crisis (really the BoE should have been paying more attention to the trend!). But since the peak of 5.2% in October 2011 the trend has been sharply downwards. In fact CPI inflation has been falling nearly as sharply as it did in 2009. Yes, there have been a couple of spikes, but there is no doubt that the underlying trend for the last year (at the time of writing) has been deflationary.

There are two possible ways of viewing the current spike. One is that it is just that - a spike - which will not affect the overall downward trend. If this is correct, then we should expect CPI inflation to fall again in November or December. In support of this argument, the ONS notes that of the 0.5% rise in CPI, 0.3% is due to a substantial increase in student tuition fees that has just come into force: this would unwind itself in due course. And as it also does not affect "ordinary" consumers, it is arguably a distortion. That leaves an increase of 0.2% due to externally-driven food and fuel rises partially offset by price falls in other sectors. If price-cutting continues in other sectors, CPI inflation may well return to trend even if world commodity prices continue to rise.  

The alternative view is that this increase in CPI is a trend reversal - that the September figure was the turning point and this is the start of a rising trend in CPI inflation similar to that in 2010. This is clearly Sentance's view and it may also have been the reason for the MPC's decision to end the QE programme. Clearly the 0.3% rise due to tuition fees cannot be regarded as a trend reversal, since it is a one-off change. However, Sentance is probably correct that food and fuel will continue to rise in price, because the world price of essential foodstuffs is rising due to drought affecting supply, and the oil price is also rising due to world economic conditions. Furthermore, utility bills are expected to rise substantially in the next few months. If this is a trend reversal, price-cutting in other sectors will not be enough to offset inflation in these key sectors.

So it is by no means clear what the direction of CPI inflation will be over the next few months. What is clear, however, is that this month's CPI inflation increase is not due to pressure from a growing domestic economy. Here's the UK's GDP growth rate since 2008:

(for larger version, click here)

In the last quarter, the UK economy managed to grow by a measly 1%. Inflationary growth? Hardly. No, the CPI inflation increase is due to Government policy (the tuition fees increase) and external factors (rising price of commodities). So despite Sentance's gloomy prognosis, raising interest rates would not help. In fact it may make things worse. Many households and businesses are interest rate sensitive at the moment: a 50bps rise in the base rate, at a time when real incomes are falling, would increase the level of serious financial distress, with consequent impact on business production and consumer spending. Really it would not be clever to try to choke off a possible inflation trend reversal at the price of UK economic growth. 

The Bank of England should hold its nerve and continue to support economic growth. Even if this is a trend reversal, inflation really is not the main issue at the moment.  

A sensible marriage (of politics & economics)

On Friday 9th November in a letter to the Governor of the Bank of England, the Chancellor of the Exchequer changed the rules of the game for the Bank of England's Quantitative Easing programme (QE). (Yes, I know that the day before the MPC decided not to continue with QE - I'll come to that in a minute.)

QE involves the Bank of England purchasing and holding gilts (UK government debt) on the open market through its Asset Purchase Facility (APF). Gilts are interest-bearing securities, so the Government pays interest on the Bank of England's holdings of gilts. Up till now, the Bank of England has kept that interest on deposit. But now, the Chancellor has decided that the Bank of England should return those interest payments to the Treasury.

This is an eminently sensible decision. As the UK has a fiscal deficit, the Government has to borrow money to pay the interest on existing government debt. It does this by issuing more gilts or treasury bills to make coupon payments on gilts in circulation. If you think this is a Ponzi scheme, you are right. Obviously there is no alternative regarding gilts held by the private sector (other than eliminating the fiscal deficit so that interest can be paid from tax revenues). But it is frankly ridiculous that the Government is borrowing money so that the Bank of England can hoard it. Other national central banks (the Fed, the Bank of Japan) already remit coupon payments back to their respective treasuries.

So, if remitting these coupon payments back to the Treasury is a) eminently sensible b) consistent with how other countries behave, what is the issue? Why has there been such a kerfuffle about it, with claims that this is an accounting "fudge" and will cost the country more in the long run?

Firstly, there is a great deal of confusion about the accounting. There is no net profit or loss to either the Exchequer or the BoE from coupon payments on QE. It is purely a matter of cash flows between two components of the public sector. The Bank of England is wholly owned by the Treasury and its accounts are consolidated in the Whole Government Accounts (WGA). Both the BoE's holdings of gilts and the interest it receives on them are eliminated against Government debt in the consolidated accounts. An internal transfer between the Bank of England and the Treasury will make no difference whatsoever to public sector finances.  It's a wash.

The pricing is a wash, too. Basically the Government has now redefined all gilts held by the Bank of England under the APF as zero-coupon, whereas the price the Bank paid for them would have included expectation of coupon payments. In theory, therefore, the Bank should suffer an immediate (unrealised) capital loss. But these gilts will still be marked to market as interest-bearing securities. Unless the price of gilts drops as a consequence of this decision - for which there would be no rational reason, since the Government will continue to pay coupons on private sector holdings - there can be no capital loss in reality. And even if there were, the expectation is that the Bank would eventually sell the gilts back to the private sector at the coupon-inclusive price. As I said, it's a wash.

Unrealised mark-to-market profits are excluded from the proposed transfers, which are to happen quarterly. But Osborne allows the Bank to retain cash from the coupon payments to cover unrealised mark-to-market losses. This is slightly odd, since the effect of QE is to raise the gilt price: I suppose he is concerned about a future speculative attack on gilts. Allowing the Bank to retain cash to cover unrealised losses would protect its balance sheet.

The argument against the Bank returning its coupon payments rests on the idea that the market price of gilts would have dropped by the time they are sold. This is likely, as unwinding QE would be done in a growing economy in which interest rates would be at a more normal level. So the Bank of England would indeed suffer real losses on its holdings of gilts and could reasonably expect the Government to indemnify them for these losses - which as both Osborne and King note, would effectively mean return of the coupon payments. The pessimistic OBR assumes that this would have to be covered by new gilt issuance. But as Britmouse points out, an economy in better shape would have better tax revenues. Any loss incurred by the BoE would be offset to an unknown extent by increased Government income. So the Treasury might not have to issue debt to compensate the BoE for trading losses on its gilts portfolio. It could be yet another wash.

In fact there's nothing to look at here, really. Except for one little snippet in King's reply to Osborne that makes me wonder if King is losing his marbles. Osborne says:

"net coupon income transferred from the APF to HM Treasury should be used solely to benefit the public finances and to reduce debt". 

So, no fiscal easing then. But in his reply, Mervyn reinterprets this:

"your intention is to use any funds transferred to the Exchequer to reduce the stock of outstanding government debt."

Er, no, this can't be right. The UK has a fiscal deficit. Total outstanding debt CANNOT be reduced. All that can be done with this money is either fiscal easing (which Osborne rules out) or deficit reduction. Now admittedly Osborne has not made his meaning entirely clear - "reduce debt" must mean issue less debt, not reduce existing debt - but really King should know better.

King then concludes on this basis that remitting this interest is equivalent to a small monetary easing, because the private sector will hold less government debt and more money. And in deciding to end the QE programme, the MPC seems to have bought King's view that gilt coupon remittance amounts to monetary easing. So are they right?

Well, if Osborne does use the money to reduce the fiscal deficit, yes they are right - there would be a small monetary easing. Not because any of the EXISTING stock of debt would be reduced, but because less would be issued in future, so the private sector would be forced to hold more cash or make riskier investments - which is the same effect as QE. Fiscal easing - increasing government spending and/or cutting taxes - which is Osborne's other option, would also release this money into the economy, but without reducing debt issuance.

And this brings me to my final point. The timing of this announcement is remarkable. Osborne is in danger of missing his fiscal reduction targets because of the poor state of the UK economy and collapsing tax revenues. This money is a political windfall to him, enabling him to claim that he has reduced the fiscal deficit despite difficult economic circumstances. And he can follow this up at a later stage with a fiscal boost just nicely in time for the 2015 General Election. This little piece of chicanery is from the same stable as his cutting of the top rate of tax earlier this year, which as I pointed out at the time was purely aimed at improving the Conservatives' chances of re-election in 2015.

However, it is still sensible economics. For once, Osborne's political instincts and the needs of the UK economy make a happy marriage. We should not criticise this move.

The illusion of safety

I have recently been reviewing various proposals for making the financial system "safer". Most of them involve some kind of full reserve banking, while one (Gary Gorton's) looks at improving the safety of the shadow banking system without subjecting it to the same rules as licensed banking. But all of them rest on the idea that there is such a thing as a "safe" asset.

The concept of "safe" (i.e. risk-free) assets has existed for a long time. Pricing models base themselves on the "risk-free rate". The instrument that is usually used as the proxy for "risk-free asset" is the United States Treasury (UST), but until recently all sovereign debt was regarded as risk-free, at least for bank capital adequacy requirements.

This led to banks loading up on the debt of Greece, Spain, Portugal etc. because those assets required no capital allocation but gave a better return than German bunds or UK gilts. They should have paid more attention to the yield. Yield is a much better indication of real risk than any regulatory view of an asset class. "If it pays better, it's riskier" is axiomatic in financial markets. It is unfortunate that in recent years investors and banks alike forgot this basic rule and believed that they could get high returns for zero risk. And even more unfortunate that retail bank depositors thought so, too.

Gary Gorton notes that one of the main drivers of the 2007/8 financial crisis was the failure of supposedly "safe" assets created by the private sector. This is a much neglected matter. The popular perception is that mortgage backed securities and their derivatives were known to be risky assets that banks and others chose to trade because they believed they would be bailed out if the assets went sour. The last statement here is correct - the financial sector did indeed believe it would be bailed out. But the first statement is not. The financial sector believed that these instruments were safe, because they were backed by property which (in the US) had been a stable appreciating market for a very long time. The UK does of course have a history of property market crashes, so it is slightly surprising that banks in the UK were fooled: perhaps they thought that the US property market was not subject to the same instability as the UK - or perhaps they just had short memories.

So Gorton argues - correctly - that it is not possible for the private sector to create safe assets. But he then states that only governments can create "safe" assets. This is in my view where he comes unstuck, for reasons that I shall explain shortly.

Gorton is not the only one who thinks government debt is "safe". John Kay also thinks so - or at least he used to. His "narrow banking" proposal would back all bank deposits with gilts. However, in his recent evidence to the Banking Standards Committee at the House of Commons, he was questioned about this by former Chancellor of the Exchequer  Lord Lawson, and was forced to concede that gilts could not be regarded as 100% safe.

A quick look at the history of sovereign debt defaults gives the lie to the idea that sovereign debt is in any way "risk free". However, there is no doubt that some sovereigns are "safer" than others: the UK and the US have never defaulted on their sovereign debt. So why shouldn't their debt be regarded as "risk free"?

There are two risks with sovereign debt. The first is sovereign debt default, which for the UK and US is a pretty low risk, though not zero. But the second - a much more serious risk as far as the US and UK are concerned - is variation in value. Gilts and USTs are actively traded and their price varies day-to-day. That creates a problem when using them to back customer cash deposits. As the whole point of holding these instruments would be to liquidate them quickly if necessary in order to meet deposit withdrawal requests, they would have to be marked to market daily. If  the bank was legally required to maintain 100% government securities backing, the bank would therefore have to trade gilts and/or USTs itself on a daily basis. A sharp drop in the value of government securities and/or freezing of the markets could bankrupt it in much the same way that writedowns of mortgage backed securities bankrupted banks worldwide in the financial crisis.

On the face of it, this weakens Kay's proposal and gives greater weight to cash-based 100% reserve banking proposals such as those from the IMF and Lawrence Kotlikoff. And although it doesn't completely wreck Gorton's ideas (since UK and US government debt is probably safer than private sector assets) it does mean that absolute safety of cash deposits in the shadow banking system cannot be guaranteed.

But a quick look at the inflation records of the US and the UK gives the lie to the idea that cash is "safe" either. Some people argue that "safe" means "won't suffer loss of principal" rather than "won't decline in value". But since the sole purpose of cash is to buy real goods and services, if £1 buys you a large loaf of bread today but only a small one tomorrow, the effect is that you have lost half your money. You still have £1, but you can only buy half as much with it. The value of cash depends on the willingness of government (via its central bank, usually) to control inflation. And that depends on a range of political and economic factors.

And it's not just cash whose value is affected by inflation. The value of government debt is, too. I've noted before the equivalence between yields on government debt and the inflation rate. Generally speaking, the higher the inflation rate, the higher the yield on government debt (and therefore the lower its value). This equivalence doesn't hold on daily basis, of course - government debt and currency values fluctuate independently of each other - and it doesn't hold in fixed currency systems such as the Eurozone, but over time it is a useful rule of thumb for floating exchange rate systems with no restrictions on movement of capital.  

So is anything safe? Well, proponents of gold think so. Those who support gold as a currency point to the stability of the "true" Gold Standard period of the late 19th century as a time when cash held its purchasing power. But in this article from Economic History (EH), the author questions why the gold standard during that time was stable, because on the face of it it shouldn't have been - and he concludes that there were two reasons: firstly, there were no major wars at that time, and secondly, both the private sector and the principal governments involved were fully committed to ensuring its stability. In other words, gold ITSELF is no more stable in value than any other sort of currency - as indeed the gold price chart shows. It is the backing of governments and trust from private agents that makes it so. EH notes that the "backing of governments" involves a commitment to settle trade deficits in gold: but the UK defaulted on its international trade obligations when it abandoned the gold standard in 1931, and the US did likewise when it ended the Bretton Woods system in 1971. The UK and US may not have defaulted on sovereign debt, but that doesn't mean they have never defaulted.

We also know from history that gold standard currencies don't survive major economic shocks, such as wars. The gold standard was suspended in the UK during the Napoleonic Wars, the First World War and the Depression: in the US during the Civil War, the First World War and the Depression: in France during the French Revolution and the First World War....I could go on, but do we need more evidence? Wars and major economic shocks need governments to create money in large quantities, which they can't do if they are locked into a gold standard - or any other kind of fixed currency regime, for that matter. And wars and major economic shocks disrupt international trade, which also undermines the gold standard. Returning to the gold standard at the pre-shock price is almost never achievable in reality. Abandoning the gold standard - which inevitably happens when there are major exogenous shocks - results in debasement of cash savings.

Those who support gold as an investment, but not as a currency, point to the rising value of gold over time. It is true that over the long term, gold does tend to appreciate in value - though so do property and land. But as a short-term investment it is volatile, which makes it completely unsuitable as collateral in a fast-moving financial marketplace or as backing for transaction accounts. And as a vehicle for personal savings it carries another risk - it is a target for thieves. In those countries where for cultural reasons people tend to put their savings into gold, theft (often violent) is a constant threat - which is why the World Bank is trying to encourage people to put their savings into banks rather than gold or cash. To the people of the West, scarred by recent events and fearful of bank failures, this looks like madness. But in developing countries, banks are definitely safer places to keep savings than mud huts. Gold stashed under the mat may not lose its value, but it can be lost.

So, over the long-term there may be assets that can be regarded as "safe".....but on a short-term basis, there is no such thing as a completely safe asset. Which brings me to my final concern about the search for assets that are "safe".

The belief that certain assets are "safe" encourages risk-averse investors to buy them. This pushes up the price and reduces the return on these assets. We saw in the run up to the financial crisis how the demand for assets that were considered "safe" drove production of these assets, leading their creators to take on more and more risk. The nature of the asset changed: it was no longer low-risk, but the investors did not know that, so continued to buy it until the rising tide of defaults in the underlying property portfolios caused a massive price correction. Once the true risk of these assets was fully exposed, their price collapsed and they became worthless.

The same is possible if either government debt or currency is produced in sufficient quantities to meet demand without domestic production rising to match. If this happens to government debt, eventually the price drops and the yield rises: if it happens to currency, the exchange value drops and there is high inflation.

However, I want to make it clear at this point that I do not believe we are anywhere near this state of affairs in Western economies generally at the moment, except for some members of the Eurozone. We have the opposite problem, namely that governments are unwilling to produce either debt or cash in sufficient quantities to satisfy the demand for "safe" assets, so we are seeing the price of government debt and the value of currency rising (and yields on debt and trend inflation falling) in those countries that are considered "safe havens".  Once yields are in negative territory, investors LOSE MONEY. And that is what is happening at the moment. Investors are so concerned not to lose their principal that they are loading up on supposedly "safe" assets which actually cost them money to hold. Remember I said that, for cash, loss of value due to inflation is just as real a loss as principal wipeout in bank default? In the same way, negative yields on government debt are just as real a loss as principal wipeout due to government debt default. No way are they "risk free".

And one final thought. We do not know what causes asset bubbles, but Austrian theory of business cycles suggests that the next bubble starts to form during the aftermath of the bursting of the previous one. In the aftermath of a major crash, investors are understandably risk-averse - as we are seeing at the moment, with their flight to traditional "safe havens" and their willingness to accept principal erosion rather than risk total wipeout. That suggests that, rather than greed and the hope of a high return, it may be fear and the search for safety that drives the creation of asset bubbles. In seeking the illusion of safety, investors may set up the very disaster they wish to avoid.

Supermarket banking

At the recent Occupy debate on "Socially Useful Banking", Andy Haldane argued strongly for culture change in banking, and particularly in retail banking - the sort of banking that ordinary people rely on for their payments, their short-term savings and their loans. And he claimed that structural separation of retail banking from what are viewed as "higher-risk" activities such as investment banking and trading would end cross-contamination of culture and enable retail banking to return to its roots in relationship management and judgement based on local knowledge. He said that the ring-fencing of retail banking as proposed in the Vickers report on reform of banking may not be enough and full separation would be considered if necessary.

The roots of his remarks lie in the prevalent belief that the changes in retail banking in the 1980s/90s were a consequence of the liberalisation of investment banking in the so-called "Big Bang" of 1986. He comments that:

".....financial and human resources were diverted away from retail banking services and non-bank activities towards investment banking.  At the same time, the culture and practices of investment banking infiltrated retail banking - a sales culture which culminated in harmful cross-selling and unlawful mis-selling."

I don't agree. I think the changes in retail banking arose from introduction of American marketing and retail sales techniques, and from technological change that enabled economies of scale in back-office processing. Investment banking had very little to do with it: by the time of Big Bang, retail was already a low-margin, cut-throat business in which the weaker players found it very hard to make significant profits. I would suggest, rather than investment banking draining retail, that the significant move towards investment banking throughout the 1980s was driven to a considerable extent by the poor returns in retail banking. Nor is it true to say that investment banking drained resources from retail: yes, the people working in retail were reduced in number and de-skilled, but there was enormous investment in IT at the same time. The change in retail banking was exactly the same as the change happening in retail shopping at the time - the crowding-out of small high street shops with huge out-of-town shopping malls: the replacement of personal service with glitzy product displays and sales staff paid on commission; investment in IT for highly efficient, low-cost manufacture and delivery of "products" using largely automated "production line" technology.

I have reason for saying this. In the late 1980s and 90s I worked at Midland Bank, which was the first bank to transform its retail operations into the sales-driven model with which we are now so familiar. Midland was dreadfully short of money: it had been (as we now know, fatally) weakened by its purchase and subsequent sale to Wells Fargo of the highly-indebted American bank Crocker, the last in a long line of aggressive domestic and overseas acquisitions, many of them merchant and wholesale banks. To those of you familiar with the causes of the downfall of RBS, this should sound horribly familiar.....but Midland was not bailed out by taxpayers. Its future was different.

In 1986, when I joined the bank, Midland was already going into a period of retrenchment. It consolidated its merchant and wholesale banking activities with its international businesses to create its Midland Montagu subsidiary, which was both functionally and legally separate from the rest of Midland Bank. In other words, it created structural separation between its core retail business and its investment banking activities. I worked for Overseas Systems Support at the time, and I remember the transfer of my department to Midland Montagu about a year after I joined: we all had to sign new employment contracts that differed significantly from our Midland Bank employment contracts.

Driven by the American Gene Lockhart, head of Domestic Banking, Midland embarked on a massive programme of automation and centralisation of its retail back-office operations. It created huge dedicated processing centres, highly automated and thinly staffed. Back-office functions such as cheque processing and payments were removed from branches and consolidated in these processing centres. Branches were redesigned to be like retail shops, staffed with sales staff whose job was not to provide a service but to sell "products": they were given challenging sales targets and their pay became partly commission-based. Some branch staff were made redundant - especially the older, more experienced (and more expensive ones): the rest were retrained as sales staff. Branch managers were redesignated as sales team leaders and the job became less senior: their authority was severely curtailed as Midland created specialist credit officers, not based in branches, whose job was to make lending decisions on the basis of credit scoring and customer history.

The changes made by Midland did improve the performance of its core retail business and probably enabled it to stay afloat longer than it otherwise would have done. But inevitably, other retail banks such as Nat West soon followed suit, and the competitive advantage that Midland had gained proved short-term. Midland also suffered losses in its investment, wholesale and overseas banking activities, including a completely ridiculous loss of £20m caused by "wrong positioning" on money market trading (traders had taken an open punt on the direction of interest rates and were caught out when the Bank of England raised interest rates), and was heavily exposed to Latin American bad debt. Provisioning against this virtually wiped out its profits, so it embarked on a programme of cost-cutting and redundancies, which it  euphemistically called the Profits Improvement Programme (PIP). No-one was fooled. A new word joined the lexicon: anyone made redundant had been PIPped.

But at the same time that it was cutting costs all over the place, Midland also embarked on a large (and expensive) project to improve its financial and management information-gathering and reporting so that it could make better use of its capital and manage funding and risk more effectively. I worked on that project: its budget was more than Midland's entire profits in 1991. Something had to give.

In 1992 Midland was taken over by HSBC, which was looking for an escape route from Hong Kong in anticipation of the return of Hong Kong to China in 1997. This takeover was  presented to the world as a friendly merger - and indeed it was the only realistic solution to Midland's problems: chronic shortage of money can only be solved by marrying money...... But internally it was nothing of the kind. The financial information project that I was working on was cancelled (after all, HSBC had pots of money, so it didn't need to use advanced capital, risk and liquidity management techniques to glean a few more basis points on RoE). All Midland's systems were migrated to HSBC's platform. Hundreds of staff at all levels, but especially in middle management, lost their jobs. But HSBC didn't reverse the changes in retail banking that Lockhart had driven. The sales model in retail banking remained, and indeed remains to this day at HSBC.

The final nail in the coffin of "traditional" retail banking was the conversion of building societies into banks, starting with Abbey National, which led to significant consolidation of retail banking and the creation of the "super-banks" we now have. The age of "supermarket banking" had arrived.

The purpose of this extended tale is to show that the cultural issues in retail banking are not because of "cross-contamination" from investment banking. They came from the retail sector itself. The model for the future of retail banking was not investment banking: it was shopping. Banks created "one-stop shops" where retail customers could go to buy all the products they needed, including pensions,  insurance and, for business customers, derivative products: banks knew they could make far more money from selling these products than they could from providing core banking services, so branch staff were given aggressive sales targets for selling non-banking products. The mis-selling that Haldane refers to stems from this cause, not investment banking. And despite all the scandals, NOTHING has changed. Branch staff still have sales targets, still exist to flog you "products" rather than provide banking services. My local branch recently started opening on Saturdays - but not to do boring banking things like depositing money. No, all you can do is discuss your financial product requirements with sales staff.  The cash desk is closed.

Haldane cited Svenska Handelsbanken as a current example of a retail bank that was "doing it right" - creating small local branches that served their local communities and rewarding its employees on a "pool" basis rather than individual sales targets. I absolutely endorse this: just as we are beginning to rediscover the benefits of small high-street shops providing a personal service, so we also need to rediscover the benefits of small bank branches providing a personal service.

But I would issue a word of caution here. Actually, for retail banking the shopping model is not a bad one: the problem with "supermarket banking" was the aggressive selling, the drive for short-term profit at the expense of long-term customer relationships, and the lack of effective regulation. Providing efficient, low-cost banking services does require centralised, automated operations, and with the advent of telephone and internet banking the need for local provision in ordinary banking activities is reducing fast. In my view there is still a place for supermarket banking, just as there is still a role for supermarkets: not everyone has the time to visit lots of little shops to get their weekly shopping, and not everyone can afford the higher prices that small shops tend to charge. Banking is the same: small scale and personal service comes at a price.

The abuses in retail banking must end. But to do that we need to understand what caused them, because otherwise we will not know what to change. Separating retail and investment banking made no difference at Midland Bank. I strongly suspect it will make no difference to retail bank culture this time round, either. The changes must come from within.