Jen (former student of Krugman BTW) seems to be optimistic about the $ over the next year but much less so after that:
Nationalisation ? Currency Weakness
September 19, 2008
By Stephen Jen & Spyros Andreopoulos | London
Morgan Stanley acted as advisor to the United States Department of the Treasury in the announced restructuring of the Federal Home Loan Mortgage Corporation (‘Freddie Mac’) and the Federal National Mortgage Association (‘Fannie Mae’).
Please refer to notes at the end of the report.
Summary and Conclusions
Conventional wisdom has it that, as a government fiscalises the contingent liabilities of nationalised banks, the currency of the country in question should depreciate. More generally, banking crises are, very often, accompanied by balance of payments (or currency) crises. The US, being a country with still out-sized ‘twin deficits’ (fiscal and external deficits), will likely see the dollar weaken because of the Treasury and the Fed’s decision to effectively nationalise some of the large financial institutions, so the argument goes. An inconvenient fact, however, is that nationalisation of banks, historically, did not tend to lead to further currency weakness. In fact, very often the financial sector and the currency in question reach a trough just as the government takes steps to address the banking crisis. Thus, currency weakness tends to precede, not follow nationalisation.
Popular Thesis on Nationalisation and the Dollar
The notion that nationalisation of banks should lead to currency weakness is popular mainly because it is intuitive. Since nationalisation of banks is ‘not good news’, and runs counter to the principles of capitalism and the free market, some have the visceral reaction to sell the currency in question.
Further, as highlighted by Kaminsky and Reinhart (K&R) (see Graciela Kaminsky and Carmen Reinhart (1999), “The Twin Crises: The Causes of Banking and Balance-of-Payments Problems”, The American Economic Review 89: 3, June), there are many historical examples of ‘twin crises’, whereby banking crises and currency crises occurred simultaneously. The more memorable examples include Argentina in the early 1980s, Sweden and Norway in the early 1990s, Japan in the late 1990s and Thailand in the late 1990s. In fact, K&R found that, during 1980-1995, of the 23 banking crises, 18 were accompanied by balance-of-payments crises.
This link between banking crises and currency crises is genuine, and the usual dynamics are well-summarised by ex-Governor of the Riksbank (Sweden’s central bank) Mr. Bäckström (see What Lessons Can Be Learned from Recent Financial Crises? The Swedish Experience, Kansas City Fed Seminar in Jackson Hole, August 1997): “Credit market deregulation in 1985 … meant that the monetary conditions became more expansionary. This coincided, moreover, with rising activity, relatively high inflation expectations, … (T)he freer credit market led to a rapidly growing stock of debt… The credit boom coincided with rising share and real estate prices… The expansion of credit was also associated with increased real economic demand. Private financial savings dropped by as much as 7 percentage points of GDP and turned negative. The economy became overheated and inflation accelerated. Sizeable C/A deficits, accompanied by large outflows of … capital, led to a growing stock of private sector short-term debt in foreign currency”. This description applies quite well to the US right now.
Moreover, nationalisation of banks will increase the fiscal burden of the government. For a country that already has a large fiscal deficit, this is clearly negative for the interest rate outlook. For one that also has an external deficit, a large public borrowing need, ceteris paribus, should translate into a weaker currency, so the logic goes. At the same time, the central bank may be tempted to ‘monetise’ the debt, or run a monetary stance that is easier than otherwise – again currency-negative.
The Inconvenient Historical Fact
While the arguments above may sound logical and compelling to many, the inconvenient fact is that the historical pattern of how currencies perform before and after nationalisation or bail-outs tells a very different story. Averaged across five episodes of prominent banking crises, the nominal exchange rate tended to fall before nationalisation, but rise thereafter.
The historical pattern suggests that financial markets tend to be forward-looking and try to price in the deterioration in the state of the banking system by selling down the currency and financial sector stocks, but the government is usually not compelled to act until conditions deteriorate significantly. As a result, more often than not, government interventions have coincided with the lows in currency values. In other words, even though K&R’s observation that currency crises often occur simultaneously with banking crises is correct, there is no strong proof that nationalisation leads to further currency weakness.
Other more visible examples are consistent with this link between banking crises and currency crises. The S&L Crisis and its bail-out spanned a protracted period of time. The dollar index did continue to fall from 1986 – the beginning of the S&L Crisis – until 1989 or so. (In 1986, the FSLIC (Federal Savings and Loan Insurance Corporation) – the deposit insurance scheme funded by the thrift industry but guaranteed by the government – first reported being insolvent (incidentally, the main reason why 1986 is remembered as the beginning of the S&L Crisis). The RTC (Resolution Trust Corporation) was established in 1989, and by 2003, the RTC had ‘resolved’ US$394 billion worth of non-performing assets of US savings and loans. (The total cost of the clean-up of the US S&L Crisis reached US$153 billion, in ‘current’ terms equivalent to some 2.6% of US GDP in 1991. This translates to US$375 billion in 2008 dollar terms.) The dollar index essentially moved sideways in the early 1990s. The dollar did falter in 1994/95, but that was attributed more to the inflation scare than to the S&L Crisis. Similarly, Japan’s government did not explicitly address its banking crisis until 1998-99 and again in 2002-03. After each episode, USD/JPY actually collapsed toward 100, i.e., JPY strengthened in the ensuing quarters. Finally, in the case of Thailand, the banking crisis did indeed lead the currency crisis. But bank bail-outs did not take place until 1998, and USD/THB drifted in the 36-42 range between 1998 and 2000 – significantly below the peak of 56 reached in January 1998.
The case of the US at present is also illustrative. Between the onset of the credit crisis in August 2007 and the collapse of Bear Stearns on March 16, 2008, EUR/USD rose from 1.35 to 1.58, and lingered around the latter level as the Fed and Treasury assisted other financial institutions in the subsequent months. Since July, EUR/USD has collapsed from 1.60 to a low of 1.39 last week. Even with recent dramatic events, there is no evidence that ‘nationalisation = currency weakness’. If anything, the dollar has held up remarkably well this week, despite several dollar-negative factors, including: (i) a higher probability of the Fed cutting the FFR than the ECB reducing the refi rate; (ii) a diluted Fed balance sheet, from the substitution of US Treasuries for other lower-rated securities; and (iii) large increases in the future fiscal burden of the US, from the contingent liabilities that are fiscalised. In fact, the only dollar-positive factor this past week was lower oil prices. EUR/USD seems to be drifting back toward 1.45, but we see this move as rather innocuous, given the severity of the financial stress in the US.
In sum, banking crises are unambiguously bad for currencies, but nationalisation per se does not make the situation worse for currencies. In fact, it often marks the low in the currencies.
The US Fiscal Worries Over the Medium Term
Having said the above, the US does have quite a worrisome fiscal outlook in the years ahead, which may eventually have an impact on the dollar. Setting aside the issue of the fiscal burden associated with the assistance the official sector has provided the financial sector, US expenditures may be too high and revenue buoyancy may be undermined by the weak equity and property markets.
The Congressional Budget Office (CBO) released its budget update last week, and predicted that the US federal deficit will rise from US$161 billion (1.2% of GDP) in 2007 to US$407 billion (2.9%) in 2008. This sharp deterioration in the fiscal balance reflects a simultaneous increase in spending and a decline in tax revenues. (Total government spending will increase by US$226 billion, to close to US$3.0 trillion, reflecting both discretionary and mandatory spending.) The CBO forecasts that deficits will remain above US$400 billion in each of the next two years.
Investors will likely see it as key for the next Administration to control spending. However, it is also important for investors to appreciate how sensitive US revenue collection is to GDP. During 2001-02, for example, as the US economy fell into a brief recession, revenue collection plummeted from 21% of GDP to close to 16%. Thus, the strength of the US economic recovery in the coming years will have important implications for the overall budget position. These fundamental trends in revenue collection and ‘core’ spending are at least as important as the costs associated with nationalisation. The performance of the dollar in the coming years will, therefore, be a function of how the US government deals with spending and how rapidly the US economy recovers, in our view.
Bottom Line
Banking crises are bad for currencies, but nationalisation per se does not necessarily make it worse for currencies. In fact, it often marks the low in the currencies. We believe this is the case for the dollar in the current episode. What remains a lingering risk for the dollar over the medium term is the US fiscal position, unrelated to the costs of nationalisation.
Financial Assistance Plans and the Dollar
October 06, 2008
By Stephen Jen & Spyros Andreopoulos | London
Summary and Conclusions
In this note, we challenge several widely held presumptions by investors regarding financial assistance plans and currencies. Among other conclusions, we find that (1) large debt issuance may not have a lasting impact on interest rates in the US; (2) there will likely be no monetisation associated with financial assistance plans; and (3) the dollar should not be adversely affected by TARP.
Popular Presumptions Regarding TARP and USD
Last Monday (September 22), after the announcement of TARP by the US Treasury Secretary Paulson toward the end of the previous week, not one media report we could find had anything good to say about the dollar. Here are some quotes from news reports from that day:
“The dollar will get crushed”. Bloomberg, September 22, 2008.
“The US government probably is going to commit to spending about US$1 trillion bailing out the bumbled. That is roughly on par with the M1 money supply… Cheaper dollars may make it easier for the government to repay its debt, but overseas borrowers will demand higher interest rates”. WSJ, September 23, 2008.
‘The volume of fresh government borrowing and the fast expansion of the Fed’s balance sheet are both negatives for the dollar, carrying a potential risk of increased inflation… It is hard to believe that the dollar will continue to stand its ground as the crisis continues to deepen”. the Times, September 21, 2008.
After an initial sell-off, the dollar has rallied in recent days and looks set to gain more ground against virtually all currencies in the world, in contrast to the dire predictions about it.
The arguments against the dollar, with financial assistance plans, are manifold. First, federal debt issuance would explode in size, jeopardising the US Treasury’s AAA rating, and invariably raise the cost of borrowing in the US. Second, to pay for the toxic assets, the Fed will eventually be forced to ‘print money’. Monetisation would, in turn, erode the purchasing power of the dollar, i.e., lead to a depreciation of the dollar. Third, financial assistance plans, therefore, are bad for currencies. In the case of TARP, it could be devastating for the dollar.
Refuting These Popular Presumptions about Nationalisations and USD
We believe that the dollar’s strength is justified, and expect it to continue to strengthen against a wide range of currencies. While the main reasons behind our USD-positive view are related to risk-aversion and market positioning accentuating the dollar’s supremacy as the world’s dominant medium of exchange and unit of account, we also believe that TARP should not be negative for the dollar. We tackle the three popular presumptions in turn:
• Presumption 1. Large debt issuance associated with funding financial assistance plans will raise the cost of borrowing for the US. It is sensible to assume that changes in supply of and demand for securities should alter the asset prices in question. In the case of the US, if US$700 billion worth of additional US Treasuries, over and above what the US would have issued in the absence of TARP, will need to be supplied to the market, yields in the US should rise. However, past experiences with large net increases in government debt have not been accompanied by sharp rises in interest rates. We examined the cases of Japan, Sweden, Finland, the US and Thailand, and found no systematic relationship between changes in the net public debt outstanding and changes in the yields on government bonds.
Looking at Japan, there is no clear relationship whatsoever between these two variables, even though Japan ran fiscal deficits for several years in the range of 8-10% of GDP. Large new supplies of JGBs did not lead to spikes in the JGB yield. The scatter charts for the other cases we looked at showed similar ‘clouds’.
There are several possible explanations for this non-result, in contrast to the popular opinion that there should be a relationship. First, many large real money accounts are indexed, and the indices themselves are a function of the size of the various debt markets. Thus, in a way, supply generates demand. Second, the concept of ‘Ricardian Equivalence’ suggests that nations smooth out their consumption pattern over time to accommodate swings in public deficits. Large public spending today could trigger a rise in private savings and therefore an increase in the demand for bonds to match the additional supply. Indeed, under this line of argument, the effect of the TARP on interest rates will partly depend on private sector expectations of the Program’s net fiscal cost. Third, sharp increases in public debt usually occur in a weak economy, where consumption should be weak and savings high. The preference for bonds over equities should also be stronger in such an environment, ceteris paribus. In short, investors should not jump to conclusions about what might happen to US interest rates if TARP is passed. Financing for the additional US public debt may be less of a concern than many may have in mind.
• Presumption 2. The Fed will be forced to ‘print money’ to monetise the debt, which will lead to dollar weakness. This is also a popular notion that we contest. The Fed stopped monetising the issuance of Treasury debt with the famous Treasury-Fed Accord of 1951. This agreement, which was an intellectual cornerstone of the modern, independent central bank, is a foundation for current monetary policy. Today, monetary and fiscal policy are independent of one another, and financial assistance plans, ironically, involve fiscal policy. There is no money printed! To the extent that TARP succeeds in undoing the knots in the financial system and encourages banks to start lending again, the intermediation process should recover. In that sense, broad monetary aggregates could rise, but that would be precisely the desired outcome as growth recovers, and not something that would be negative for the dollar.
The Fed’s independence is protected by the Federal Reserve Act, signed into law on December 23, 1913, by President Woodrow Wilson. There is no reason to believe that the Fed will be coerced to ‘print money’ to inflate this debt away. However, it is possible that the Fed’s monetary stance may be less hawkish than it would otherwise be, because of the economic slowdown associated with the banking crisis. But this is very different from TARP causing the Fed to intentionally print money to drive down the dollar.
This notion of ‘money printing’ was also popular during the Asian Currency Crisis in 1997. Thailand, Malaysia, Indonesia and Korea, it was argued, would need to resort to monetisation to deal with the large debt associated with the bailout of their banks. None of these central banks ‘monetised’ the public debt associated with the bank bailouts. Why would investors think that the Fed would do so?
• Presumption 3. Financial assistance plans are bad for currencies. In Nationalisation ? Currency Weakness (September 19, 2008), we already stated our view that this notion, though intuitive, is not corroborated by historical experiences. In none of the cases of financial assistance plans we examined (Sweden, Norway, the US, Japan and Thailand) did the currency in question depreciate sharply after the nationalisation operation, even though, in all cases, the currency in question had depreciated with the banking crisis, up until around the time of the nationalisation. We found that, in the episodes we looked at, on average bank bailouts took place only after the currencies had already crashed, and nationalisation actually marked the lows in these currencies.
Bottom Line
We disagree with the presumptions that (i) interest rates in the US will rise because of the large supply of public debt associated with financial assistance plans; (ii) the Fed will be compelled to ‘print money’ to inflate this debt away; and (iii) financial assistance plans will be negative for the dollar. Neither history nor the facts support them, in our view.
De Facto Dollar Zone versus De Facto Euro Zone
October 20, 2008
By Stephen Jen & Spyros Andreopoulos | London
Summary and Conclusions
We do not share the view that the dollar’s recent rally is due only to ‘deleveraging’. While we ourselves highlighted early on that cross-border risk-reduction would be a powerful driver pushing the dollar higher in an adverse global environment, we have also argued that there are important fundamental reasons why the dollar should rally. In this note, we highlight one particular macro factor that is dollar-positive. The composite C/A balance in the de facto dollar zone has steadily improved in the past three years, as it is in fact approaching a fully financed position, while, during the same period, the balance in the de facto euro zone has steadily deteriorated.
Give the Dollar Some Credit
The recent dollar rally came against the prevalent view – back in June/July – that the dollar could collapse due to the financial turmoil in the US. The general opinion during the sell-off in EUR/USD from the mid-1.50s down to the high-1.40s was one of scepticism. This scepticism persisted even as EUR/USD traded below 1.40 for the first time in this cycle on September 10. However, when EUR/USD definitively traded down to the mid-1.30s in early October, commentators grudgingly agreed that ‘deleveraging’ was the main and only reason why the dollar was able to rally, implying that, as soon as risk-aversion abated, the USD would resume its weakening trend.
While we had pointed out, early on, that cross-border risk-reduction would push the dollar higher (i.e., risk-aversion would push the world up on the left side of the ‘Dollar Smile’), we also suggested that there were other fundamental reasons why the dollar should strengthen against the EUR, that the dollar’s rise is genuine and more deserving than many USD-sceptics may have in mind.
First, a proactive Fed had already front-loaded most of the rate cuts early on. When inflation was the dominant concern, the dollar was punished for the easy Fed, and the EUR rewarded for the ECB’s vigilance on inflation. However, now it is more evident that the global economy is in a genuine slowdown, and because it is well-known that monetary policy changes act with a long and variable lag, a passive ECB should imply a risk premium on the EUR.
Second, the EUR is over-owned, while the USD is under-owned by many real money investors both in the US and in Europe. Cross-border risk-reduction will probably not be ‘reversed’, i.e., cross-border risk will not be rebuilt after the dust settles. We will not repeat our thesis (our observation of the US real money accounts having heavily diversified out of the US since 2003 was made in The Biggest Dollar Diversifiers Are American, July 19, 2007), but only stress that part of the desire to move out of the US was related to the US C/A deficit, which in turn was a symptom of the US credit bubble. Thus, a prospective repatriation of some of these outflows would be a fundamentals-based story as the secular credit cycle turns, not a ‘technical’ matter associated with risk-induced deleveraging that might turn out to be temporary.
Third, at 1.55-1.60, the EUR had been extremely over-valued. The first signs of Euroland weakness triggered a normalisation process, in our view. 1.60 was never a level that was sustainable, particularly when the global economy weakens.
Fourth, the underlying trend of the US C/A deficit has been improving steadily since 4Q05. This particular development – a narrowing in US external financing needs – has thus far been dismissed by most as an unimportant driver of the dollar. We disagree. In this note, we elaborate our thoughts on the financing needs of the dollar area.
De Facto Dollar Zone
We first introduced the concept of the ‘de facto dollar zone’ in 2002 (see A De Facto Dollar Zone, April 25, 2002). The idea is that the dollar area should not be confined to the national borders of the US. Rather, it should include countries with less-than-fully-convertible currencies that are either hard or soft-pegged to the dollar. For example, the GCC (Gulf Cooperation Council) countries are hard-pegged to the dollar (Kuwait has a basket peg with a heavy dollar weighting) and actions need to be taken (e.g., interventions and monetary policy adjustments) to maintain these pegs. Similarly, much of Asia has been soft-pegged to the dollar (e.g., China). If we consider the C/A balance of the entire de facto dollar zone, we see that the dollar zone now has a significantly better external financing outlook than three years ago. We illustrate the evolution of the trade balances for the US and the de facto dollar zone. The C/A trends follow a similar pattern.
The key reason for this trend improvement in external balance is that the rates of improvement of the savings-investment surpluses in the GCC and China have been even higher than the pace of improvement of the US C/A deficit. (Linearly) extrapolating based on the latest data, we expect the aggregate C/A balance of the de facto dollar zone to approach zero by 2009/10. In other words, within the next year or so, the de facto dollar zone could be almost fully self-financed, i.e., there will be no need for additional external financing from outside the zone.
A similar calculation can be performed for the de facto euro zone. In addition to the euro area, there are many Eastern European currencies that are hard and soft-pegged to the EUR, whose currencies are not fully convertible on the capital account basis. Similar to the de facto dollar zone concept, we compute the overall savings-investment balance for this de facto euro zone and monitor its evolution.
We have the following thoughts:
• Thought 1. Divergent C/A paths. While the aggregate C/A deficit of the de facto dollar zone has been declining sharply, approaching zero in a year’s time, the path of the aggregate C/A balance of the de facto euro zone has headed in exactly the opposite direction. Specifically, since 2002, we have seen a structural deterioration in the external balance of both Euroland and the de facto euro zone. While the overall trade balance of the euro zone still shows a small surplus, the euro zone’s C/A balance has actually gone into a deficit position (-0.5% of GDP expected in 2008, according to the IMF WEO), and the trend is likely to persist in the foreseeable future. These divergent paths reflect the opposing trends in the household savings of the US and the euro zone: rising in the former but falling in the latter. In terms of the levels of external balance of the de facto dollar zone and the de facto euro zone, the former could be more fully financed than the latter in a year’s time.
• Thought 2. Core versus periphery. The ‘core-periphery’ relationships are exactly the opposite in the two de facto currency areas. The US has a C/A deficit, but the other members of the de facto dollar zone have large C/A surpluses. In Europe, this relationship is exactly the opposite: the peripheral members of the de facto euro zone have large C/A deficits while the euro zone has a small deficit. And, within the euro area, Germany has a large surplus (7.3% of GDP), while Spain, Portugal, Greece and Ireland have large C/A deficits (10.1%, 12.0%, 14.0% and 5.0% of GDP, respectively). These relationships between the core and the periphery reflect, to some extent, the different levels of development and growth strategies of the AXJ economies and the EE economies. In any case, as far as currencies are concerned, the more committed are the GCC and the AXJ countries to the dollar, the better supported it is. On the other hand, for the euro zone, the more committed the East European countries are to the euro, the less supported it is.
• Thought 3. Loyalty toward the dollar. While we first introduced the de facto dollar zone concept in 2002, and have recognised the trend improvement in the zone-wide C/A balance for some time, we only make this observation now because, prior to July, it was not clear whether the GCC and the AXJ countries would ‘remain loyal’ to the dollar. Pressures were intense at times for the RMB and the GCC currencies to revalue against the dollar. However, now that the dollar has reasserted itself, it is much less likely that the members of the de facto dollar zone will choose to leave the zone. China has even stopped letting the RMB appreciate against the dollar; there is no more talk about GCC revaluation. Thus, the stabilisation of the dollar has significantly emboldened the structural integrity of the de facto dollar zone, making the discussion of the trend improvement in the aggregate C/A balance much more relevant now than it was three months ago.
Bottom Line
Cross-border risk-reduction is one, but not the only, reason why the dollar is strong. Among the fundamental factors supporting the dollar, we believe that the trend improvement in the C/A balance of the de facto dollar zone is crucial. In a year’s time, it is possible that the de facto dollar zone will be better financed than the de facto euro zone. |
