Friday, 31 October 2008

Global Imbalances - The Basic Story

One of the root causes of the financial crisis has been the unnatural sight of capital flowing uphill, that is from poor to rich countries. In a famous calculation, Robert Lucas (1990, AER) showed that if a rich and poor country have equivalent production technologies and differ only in income per head, then because the amount of capital employed in the poorer country will be less than in the richer country, the marginal efficiency of capital must be higher in the poorer country and so attract capital. For example, latest EIU data suggests that Chinese PPP per capita income is around $5,000 and the US is $45,000 which implies (calculation available on request) that the rate of return on capital in the US should be around 3-4% of that in China and should mean that China runs a current account deficit financed by a US surplus. The reality has, of course, been the obverse with US recycling China’s capital flows.

Let us examine the basic problem. Global savings equal investment at a single world interest rate (absenting risk). Figure 1 draws the equilibrium for the two country world of China and the USA, given their savings and investment schedules. In a closed economy, US interest rates would clear the domestic market for saving above R* and output would be determined accordingly. But when we open up to capital flows at the world interest rate, R*, the US expands its investment demand relative to savings, running a deficit, and at those interest rates China generates a current account surplus. The surplus (deficit) in each year adds (reduces) to net foreign assets in each year in the creditor (debtor) country.

The counterpart of savings excess in China is excessive investment in the US - recall that this comprises both public and private investment. Is a small reduction in US demand (investment) the answer? Not necessarily. Even if demand falls sufficiently to eliminate the US current account deficit at stable world rates, R*, then China would still have excess savings. This excess would drive rates down from R* and lead to the re-mergence of a current account deficit, albeit with lower world rates and a lower level of global imbalances. Obviously with large enough falls in US demand you could get zero current account balances in both countries at very low R and low US demand. Perhaps this is the solution, as we stare at a global recession, that we are heading towards?

We can also consider a number of alternative solutions. For example, a shift up in Chinese demand to clear the surplus at R*, at the original equilibrium, will mean excess demand in the US continues and thus world rates R* will go higher and there will still be a Chinese surplus and a US deficit. Obviously again if Chinese demand shifts up even further we can have no capital flows but at significantly higher world interest rates and high world demand. This may not be the solution we are heading towards!

The problem with this diagram as far as I can see is that any R can lead to an initial equilibrium providing China is willing to lend (or borrow) and US is willing to borrow (or lend). And what we learn is that if we want to adjust to some different level or direction of capital flows is that if only one country adjusts the overall change in rates and output will be greater than if they both adjust somewhat.

So what is the constraint or target? It must be something to do with the equilibrium level of net foreign assets to GDP and flows from one country to another to meet that target. The US is a debtor nation, (at around 6-7% of global GDP) implying that its current demand will be met by saving from higher future income. In this scheme (and I do not know the actual numbers but according to the IMF’s WEO Emerging Asia is in credit by around 5% of global GDP) then China will be the creditor. But for the reasons given earlier China should probably be the net debtor and borrow from its higher future income. And so if we are to get to a situation when eventually China becomes a debtor so capital flows downhill (from rich to poor), this will imply the need for a surplus in the US and a deficit in China, which implies lower US demand, greater US savings, higher Chinese demand and lower Chinese savings. But we probably knew that!

Thursday, 23 October 2008

Here’s That Rainy Day

Like the visitation of a medieval plague, Recession seems to have returned to our shores. These events are rare. So what exactly is a recession? I quite like Christopher Dow’s (Major Recessions, Britain the World, 1920-1995, OUP, 1998) definition of a fall in the level of GDP in one year compared to the previous year. According to Dow we only had five major recessions in the post-WW1 20th century (three after WWII) and it looks as though we are about have our first of the 21st century. In a book which will now be re-read, Dow finds that in these recessions, which typically have duration of around 1-3years, the level of output typically falls by up to 10% below the level it would have arrived at if trend growth was maintained. (He attributes around half this fall to a reduction in productivity.) During these episodes unemployment typically rises by up to 3-4% points.

A key additional finding is that recessions seem to have been unpredictable in real time, reflect changes in demand as much as any supply reductions and occur nearly simultaneously in major economies. Clearly the simultaneous occurrence of negative growth across economies might well be a candidate explanation for the severity, as trade multipliers will be running at full pelt. I do not know whether this recession will be a major one or not but the extent of the fall today in Sterling (closed on 22nd October 88.3 compared to previous day at 90.4 with Sterling falling over 6c against the US$ alone) and in the equity index (FTSE-100 and FTSE All-Share both fell by over 4%) suggests that the financial markets expect both a strong negative domestic interest rate response, which drags down sterling, and a large fall in corporate profitability, which drags down the equity index (recall that equity prices should be present value of expected corporate profits, which are in return a function of output).

In a speech on Tuesday, the Governor of the Bank of England gave two reasons for the recession. The credit market shock, which has deprived household and firms of liquidity against their collateral in some degree and so reduced demand, and a global price shock to fuel and commodities that has reduced household disposable income, rather like a large unanticipated tax rise levied by a foreign government. I suspect the financial shock has another consequence. The impecunity of the UK household balance sheet has been exposed - as perceptions of net wealth have been eroded with sustained falls in asset prices. And this means that the UK private sector balance sheet, currently well in deficit to the tune of around 5% of GDP, is certainly even more in need of a sustained bout of savings, which will surely reduce demand further.

Each of these arguments concern demand and so can, to some extent, be offset by appropriate interest rate policy. The Table above dates the trough and peak in the business cycles around the time of Dow’s major recessions. I also note level of policy rates at around the same time with their nearest peaks and troughs.

The peaks in interest rates occurred sometime after the business cycle peak but there is no clear pattern in the interest rate troughs (as in the most recent recession, policy rates were constrained by ERM membership). When inflation was more of a problem, in the earlier recessions, interest rates did not move especially far in proportional terms – falling by around 25-30% of their peak value in both cases. But when inflation was reasonably under control, as in the most recent recession, interest rates fell by a much larger fraction, by around 65%. If we mechanically apply these boundaries to the current scenario, we arrive at a corridor for base rate at 4.00% to 2.00% as the floor this time round. Clearly the current consensus is for inflation to reduce radically as commodity prices fall in response to lower world demand and the increasingly large negative UK output gap drives down pricing power of firms.

And so we might conclude for the moment that the markets have priced more of the latter scenario in than the former and hence the large exchange rate and equity price responses we have observed. But as ever we shall have to wait and see what that rainy day actually brings.

Tuesday, 14 October 2008

Part-Nationalisation of the Banks and the Tide of History

At school I learnt about Atlee’s post-war Labour government nationalising large swathes of the British economy. At about the same time in the real world, there was a prolonged period of de-nationalisation of many of those self-same swathes under the Tory government of Mrs T. The tide of time had first pushed up the size of the state, partly in response to the fatigue of war, and then cranked it down again, as the wave of free market ideas reached an apogee and I thought that was that. The part nationalisation of three major UK banks yesterday, following the nationalisation of Northern Rock and Bradford and Bingley earlier this year, seems, by general consensus, to have been as unavoidable as it was surprising given the recent tide of ideas.

In fact, as recently as this spring, I argued at a Public Policy Round Table that the state had already become too large and that during a sustained period of full employment growth a succession of government budget deficits, since the fiscal year 2002/3, did not make a great deal of macroeconomic sense. I argued that we needed a commitment to a smaller state and with it would come the room for manoeuvre the Bank of England would need to cut interest rates and offset the ongoing credit market shock. The sequence of deficits meant that the public sector net debt to GDP, excluding bank liabilities, was pushing pretty hard at the government’s ‘self-imposed’ target of around 40% to GDP. But the bail out of the banks now seems to have made my wish for a smaller state pretty unlikely as public debt looks likely to swell to postwar levels, which will in the longer term increase the incentive for a bout of sustained inflation and cause me to dust off my textbooks on the dangers of public ownerships and directed state action.

So how did we overturn the tide of history over a weekend and (partially) enact an infamous promise in the 1983 election manifesto of the Labour Party to nationalise the banking system? The short answer is that the banking sector found itself desperately short of capital and could not find sufficient funds from the private sector to raise that capital. The irony is that banks normally stand on the other side of this problem, as they are themselves reluctant to lend to individuals who need capital and ask for significant collateral or evidence of good standing before parting with any money. Or as many a businessman has put it, banks do not give out umbrellas when it is raining. But in this case, there has been no rain for a long time and a drought has resulted as liquidity has dried up and many of our banks have had no option but to buy some water from the largest reservoir in town: The State. For exercising that right to buy, they have given up a controlling stake in their future.

To recap, the UK government, in a plan that seems now to have been mimicked throughout the world, has taken preference shares (hybrids of debt and equity, where an fixed dividend rather than a profit share is paid to holders and where holders have senior claim on assets in the event of any liquidation) in three major UK commercial (or high street) banks. These preference shares mean that the government in return for an injection of £37bn owns over 55% of RBS and HBOS and around 43% of Lloyds TSB. The UK government thus has a (temporarily until the share are sold back) controlling stake in these banks. It has also offered to insure interbank lending for UK banks providing they re-capitalise privately or through the state. Recall that interbank lending has all but ceased and the interbank rate spread over LIBOR has shot up to reflect both liquidity and credit risk. The insurance, if appropriately priced, offered by the government should help eliminate the liquidity risk I would expect as a spillover also help to alleviate credit risk. Various extensions to the Special Liquidity Scheme have also been announced.

In sum these measures aid liquidity of banks and by shoring up capital allow them to have a cushion against losses and thus promote some semblance of confidence in the financial sector. All this has been required because having extended loans on the basis of relatively little equity capital, with the blessing of the regulators, the banking sector faced extinction of their equity capital from a number of possible avenues. That the default rates on loans would be higher than expected, that collateral held against those loans would be insufficient and that funding for those loans from short term deposits from households or the wholesale money markets would dry up. Given the interconnected nature of these events, it turned out that a shock to collateral triggered defaults which triggered mistrust in wholesale money markets, not only between banks but also between the net supplier of funds (pension and insurance companies) and banks.

So what we discovered was the correlated nature of these risks, that they were not independent events and that the scale of liquid reserve or capital that banks held against these shocks was insufficient. And so banks have to ratchet up their levels of capital to cover both previous losses and to have a sufficient liquid store of wealth against future shocks. The trick is to try and engineer this capital injection quickly so that lending can now resume in the face of a rainstorm...the banks need to giving out galoshes a well as umbrellas right now.

As far as temporary part-nationalisation of the banks, to coin a phrase: There Is No Alternative (for the moment).

Tuesday, 7 October 2008

“Countries Don’t Go Out of Business...”

I was this morning reminded of the Chairman of Citicorp’s 1982 famous injunction that a country cannot go bust. What Walter Wriston said at the time of Mexico’s default on its debt to commercial banks was: "Countries don't go out of business....The infrastructure doesn't go away, the productivity of the people doesn't go away, the natural resources don’t go away. And so their assets always exceed their liabilities, which is the technical reason for bankruptcy. And that's very different from a company." But when it was reported today that the country of Iceland may indeed go bankrupt I wondered, not for the first time, where the bottom of this financial crisis may lay. There is a nice irony here in that this time it is banks that are threatened with bankruptcy and they require the help of the nation-states to bail themselves out. To some extent, the early 80’s debt crisis was the other way around.

Developments across European banks, starting with Ireland last week, really are quite disturbing. The unilateral guaranteeing of all deposits and Tier 2 debt is basically akin to the guarantee given by the UK government for Northern Rock last year after its bank run. But as we know this was insufficient to prevent its eventual nationalisation - as no buyer could be found. The guarantee stopped the run on the Rock as far as retail depositors were concerned but did not allow the bank to continue with its wholesale funding model. These recent decisions seem to have been made at the national level with no real consultation with the EC (and in this case the ECB) which as a side-issue will raise a pretty big question mark over co-ordination within the EU and the Eurozone.

But these or any guarantees do not address the fundamental problems of banks that have been far too reliant on wholesale funding and whose asset quality has deteriorated markedly. That problem has been replicated again and again in the UK and elsewhere in Europe. In the case of Ireland and Iceland it is also magnified as it is not one or two banks among many but the whole banking system which seems to have adopted the same model. A blanket guarantee does not help the quality of assets and hence funding liquidity - all it does is re-assure depositors (retail rather wholesale) that their money is as safe as the finances of the Irish state or Icelandic state.

So how much can individual European states help their banks? I plot here the ratio of the sum of major bank assets to GDP for the many of the main European nations (Source: The ratio tells us neither about the riskiness of each Euro of assets on the bank’s balance sheet nor about the capability of the state to capture its tax base, GDP, per se but does perhaps allow us to put the scale of the problems into some national context. And immediately we can see the scale of the Icelandic problem with bank assets at nearly 11 times GDP, where even a 10% default (if it was backed by the state) would increase public debt to GDP by nearly 110%. Under these circumstances the blanket guarantee would simply not be credible, as an increase of debt-GDP permanently of this size will require a permanent increase in the primary fiscal surplus by some 2-3%!

We also learn that Irish bank assets to GDP are around 250%, which is only around the Eurozone average (obviously not the UK, Iceland or Switzerland) and tells us that in asset terms the Irish banking system is perhaps no larger (or vulnerable) than that of the EZ average. But the extent of the fragility, which despite not having an excessive level of asset creation by international standards, was such that a government guarantee was required. And the need for this commitment device tells us something important about perceptions about the quality of bank assets and the likelihood of continuing funding.

The extent of heterogeneity of individual countries bank asset size is interesting and of particular note is the relative strength of the German state with respect to German banks with major bank assets at only around 140% of GDP. And that even though Italy looks superficially well off the raw number of 170% does not probably deal with the relatively poor credibility of the Italian fiscal authorities, where the provision of a guarantee may be proportionately more difficult. (That said the lending there may not have been quite so risky.) So even if the Irish and others have dealt with their banking stress with a government guarantee - if a scheme of this sort was to be extended to the whole of the Eurozone it is likely to need the guarantee of the German state.

This situation is not unlike trying to devalue a currency within a fixed exchange rate zone. The analogy is that other countries will not be happy with the unilateral increase in relative competitiveness. And there will be some pressure for a further round of devaluations (guarantees), which we have now seen. Most obviously those countries who are most vulnerable and whose assets are closest substitutes will be under most pressure to give a similar guarantee. And again that is what we are now seeing.

Given the juxtaposition of dwindling deposits and deteriorating assets, I would expect governments to will want banks to increase their Tier 1 capital, if at all possible, and that may mean some deals to be struck with Sovereign Wealth Funds and also perhaps the need for governments to take a direct capital stake in banks. Bank mergers will also be a regular event as the market compresses. But given that the root cause is poor quality assets, this may only be the first step on the road to a proliferation in TARPing or some form of insurance support for assets as well as liabilities in the Eurozone and elsewhere. The onus of the government or the public sector as a solution to the immediate crisis will mean that nation-states look likely to become inextricably linked to banks once again but this time as creditors.