US public debt as a share of GDP is now higher than at any other time in history except after World War 2 - and rising: our US colleagues expect public debt to GDP to increase to 87% by 2020 (see US Budget Forecast Update: The Song Remains the Same, January 29). How policymakers will deal with this fact will likely be one of the main drivers across markets going forward. So what are the implications of high public sector debt for fiscal sustainability and inflation? To answer this question, we look at how the US economy escaped high debt following World War 2. We then quantify the inflation risks inherent in today's US fiscal position by asking what would happen if policymakers were to deal with the current debt overhang in the same way.
Stabilisation of public debt to GDP at current levels would require average inflation rates between 4-6% over the coming decade - even under much lower budget deficits than currently in place. On our numbers, even with budget deficits that are much lower than the current (and projected) levels, average inflation over the next ten years would have to be substantially above 2% to keep debt in check. Even a balanced budget would require 3% average inflation over the next decade. With an average deficit as low as 3% of GDP, debt stabilisation would require average inflation above 6%. Note that in the current fiscal year (FY) we expect a deficit of 9% of GDP, projected to decline to 5.2% of GDP by 2020. Suppose the government were to reduce the deficit to 5.2% from 2011 onwards - rather than by 2020. Stabilising the debt at current levels would then require an inflation rate of 9% on average over the next 10 years. What level of deficit would be consistent with achieving a 2% inflation target, on average, over the next 10 years? A 1% of GDP budget surplus.
It is clear that inflation risks of this magnitude are not in the price: currently, markets are anticipating inflation to be below 2.5% over the next 10 years, on average. Should we be worried about ‘debtflation' - the Fed engineering inflation to keep the debt in check? A forward-looking central bank may prefer to create a little controlled inflation now to the pressure of inflating a lot later on. And the idea of controlled inflation has influential advocates in policy circles. Former IMF Chief Economist Kenneth Rogoff has suggested the Fed announce a 4-6% inflation target for a limited period. Coincidence?
1. The Fiscal Consequences of the Crisis
The financial crisis and the Great Recession have increased US public indebtedness substantially. The debt to GDP ratio has shot up from 37% pre-crisis (fiscal year 2007) to around 60% in FY 2010, on our forecasts. With the exception of the World War 2 peak, this is higher than at any other time in the entire history of the US - including World War 1 or the Great Depression. From a fiscal perspective, it's as if the economy has just gone through World War 3.
And it's likely to get worse, implying fiscal and inflation risks. Our US colleagues expect public debt, as a share of GDP, to climb further to 87% by 2020 (see US Budget Forecast Update: The Song Remains the Same). Given this trajectory, fiscal sustainability remains a concern with investors and the public. Further, given the historical link between high public debt and inflation both in the US and internationally, such a precarious fiscal position may also pose a danger for price stability.
Quantifying these inflation risks with the help of history - and a simple accounting framework. Yet how large, exactly, are the inflation risks inherent in the current US debt position? Could inflation substitute for budgetary tightening in the pursuit of fiscal sustainability? Conversely, what is the size of the budget deficit or surplus consistent with low inflation? In short, what are the options policymakers have to keep debt in check? To answer these questions, we look to history for guidance. We ask through what mechanisms - the budget balance, economic growth, or inflation - did the US economy escape the record World War 2 debt levels? In other words, what mix of fiscal and monetary policies ensured fiscal sustainability after World War 2? Assuming the same mix is applied to the current situation, we can then put a number on long-term inflation risks.
2. Looking Back: A History Lesson
War debt burden was reduced not through budget surpluses... World War 2 left the US with a large debt overhang. In 1946, US public debt was 108.6% of GDP. Nearly 60 years later, in 2003, public debt to GDP was just 36%. Within two generations, debt had been reduced by over 70pp of GDP. This corresponds to an average decrease of debt/GDP (‘the debt ratio') by 1.2% every year. How was this achieved? Remarkably, between 1946 and 2003 the federal budget was, on average, in deficit, to the tune of 1.6% of GDP as the surplus in the primary balance (0.3% of GDP on average) was not enough to cover interest payments on the debt (1.9% of GDP on average).
...but through (nominal) economic growth... So how was the debt ratio reduced despite the US government having, on average, run budget deficits? The answer is, of course, through growth in nominal GDP. The denominator in debt/GDP grew faster than the numerator, bringing down the ratio over time. By how much, exactly? Nominal GDP growth reduced the debt/GDP ratio by 2.8%, on average, between 1948 and 2003.
... with the inflation effect larger than the real GDP growth effect! But this begs a more important question: how much of the erosion of the debt was due to growth in the real economy and how much of it was due to inflation? We split the Nominal Growth Effect (NGE) on the debt ratio into a Real Growth Effect (RGE) and an Inflation Effect (IE; we explain the accounting framework in the Appendix in the full stand-alone note).
Our numbers show that while real GDP growth reduced debt/GDP by 1.3% on average, the effect of inflation on the debt ratio was larger: 1.6%, on average, between 1946 and 2003. (In relative terms, 56% of the total Nominal Growth Effect on the debt ratio is due to inflation, with the remainder being due to real GDP growth.)
Note that the largest contribution of inflation to debt reduction came in the decade immediately after World War 2 (1946-1955). Despite a primary surplus of 1.2% of GDP, overall the budget was in deficit by 0.3% of GDP on average. Yet, the debt was reduced by 4.9% of GDP a year, through a nominal growth effect of 5.2% annually, as nominal GDP growth averaged 6.5% over the period. This very large nominal growth effect is mainly due to a substantial inflation effect - inflation averaged 4.2% over the period - which reduced debt to GDP by 3.7% every year, and to a much lesser extent to real GDP growth, which on average contributed 1.5% of GDP to debt reduction. The 1970s - the time of the Great Inflation - also exhibited a sizeable inflation effect. How come inflation was so successful in eroding the debt? We see three main factors. First, outlays were not closely linked to inflation. Second, bondholders were surprised by inflation both after the war and in the 1970s. Third, in the first post-War decade, the average maturity of the debt was - at more than 100 months - exceptionally high.
3. What If? Looking Ahead
Suppose policymakers deal with the debt now in the same way they did after WW2. Assuming the same relative roles for inflation and real economic growth as in the post-War period, how much inflation is needed, for a given budget deficit, to keep the debt ratio from increasing? Conversely, assuming a given inflation target - say 2% - what is the size of the budget deficit or surplus required to keep debt from increasing?
Our assumption for the policy objective is stabilising the debt ratio at our current estimate for FY 2010 - 60% of GDP - rather than it increasing to 87% by 2020, our long-term projection (see US Budget Forecast Update: The Song Remains the Same). On the fiscal policy side, policymakers control the primary deficit - the deficit excluding interest payments on the debt - rather than the total deficit (at least in the long run). Hence, the choice between inflation and the budget deficit is really a choice between inflation and the primary deficit, given the size of interest payments (as a share of GDP).
The Deficit-Inflation Frontier. Given the choice of (primary) deficit and the historical sizes of inflation and real GDP growth effects, we can calculate the inflation rate required to achieve the debt target: the Deficit-Inflation Frontier (DIF). On the horizontal axis we have the primary deficit, on the vertical axis the inflation rate. The DIF with the solid line assumes the long-term average (1946-2003) IE and RGE. The line above that assumes the 1946-1955 IE and RGE - it is above the long-term average DIF because in the first post-war decade the erosion of the debt was heavily skewed towards inflation. The least inflationary debt erosion took place in 1996-2003. Based on the IE and RGE of that period, we obtain the lower DIF.
Primary surpluses of at least 2.4% of GDP required to achieve a 2% inflation target. According to our numbers, with inflation at 2% on average, a primary surplus of 2.4% of GDP is required in the benchmark case of debt stabilisation at current levels. Given 1.4% interest to GDP, this implies that a 1% budget surplus is required. After World War 2, primary surpluses of the required level have been achieved during one period only: 1996-2003.
A balanced primary budget would imply inflation of 4.7%. If government expenditure other than interest equals revenue, the primary balance would be zero. (The budget deficit would then be equal to interest expenditure.) In such a case, the inflation rate required to keep debt stable is 4.7%. What inflation rate would be consistent with the primary surpluses we have seen historically? The average primary surplus as a share of GDP over 1946-2003 was 0.3%. Stabilising the debt ratio at 60% with this primary surplus would require inflation of 4.3% on average. A primary deficit of 1.2% - the 1915-2003 average - would imply an inflation rate of 6.1%.
Caveats. Our framework does not take into account the following factors: First, a given level of inflation may not have the same effect on the debt because the average maturity is shorter - though rising quickly towards, and above, the historical average on our forecasts. Second, by now almost half of federal outlays are de facto indexed to inflation. What does this mean for the inflation risks we outline? Essentially, that possibly even more inflation is needed to erode a given level of debt - at least mechanically (see also our discussion in the Box of our stand-alone note). And finally, we don't take into account the potential effect of higher inflation on real GDP growth in the medium term - or of the potential repercussions of an inflation spiral. Nevertheless, these caveats do not substantially affect our main message. The level and trajectory of the debt imply tough choices between fiscal rectitude and price stability.
4. Debtflation Nation?
This leaves one question open. Why would the Fed - in principle an independent institution - want to generate inflation? Independence means the Fed cannot be forced to inflate - at least not directly. Recent threats to its independence aside, for inflation to take hold it must be because the Fed allows it to happen. Surely, this is inconceivable?
Maybe not. Consider a Fed that faces the prospect of an 87% debt ratio in ten years' time, with population ageing and all its negative budgetary consequences imminent. In that case, a rational central bank may prefer to create a little inflation now rather than having to create a lot of inflation later on (see "Debtflation", The Global Monetary Analyst, October 21, 2009). The forthcoming increase in the average debt maturity will help.
Last but by no means least, note that the range of inflation rates we have calculated here - around 5% for the case of a roughly zero primary balance - are already being debated in policy circles. Former IMF chief economist Kenneth Rogoff has advocated a 4-6% inflation target for the Fed, and ex Bank of England MPC member David Blanchflower has made similar proposals. And Professors Aizenman and Marion calculate - in a different framework - that a "moderate" inflation rate of 6% could reduce the debt/GDP ratio by 10 percentage points within four years (see "Using Inflation to Erode the US Public Debt", NBER Working Paper 15562).
We think investors should take note - and buy TIPS, rather than CDS, if they are worried about ‘default': while hard default is inconceivable, soft default through inflation is a clear risk.
The Inflation Report was a bit more downbeat than we were expecting: There is a clear downward shift in the BoE's forecast for the "most likely" path for medium-term inflation (the mode) apparent in the fan charts. The inflation fan chart, based on ‘market' interest rate expectations (1% policy rate by end-2010 and 2.5% by end-2011), shows this central inflation forecast significantly below the 2% target two years from now. Taken at face value, this implies that the MPC would anticipate raising interest rates more slowly than the market, consensus (and we) expect. The Inflation Report projections, and the very cautious tone of the press conference, underline that risks to our interest rate forecast (a first rate rise in 4Q10) are more in the direction of later rather than earlier tightening.
The decision not to extend QE: Given the changes in the BoE's forecasts highlighted above, most observers will be wondering why it did not extend QE in February. We think the central forecasts don't tell the full story. The Inflation Report describes the "projected distribution" of both GDP growth and inflation as similar to its November forecasts. The BoE sees the balance of risks to its inflation forecasts now as to the upside and sees less downside risk to its GDP profile. Focusing on the whole distribution, this helps to explain why it sided with pausing rather than extending QE on February 4. Nevertheless, we think that the decision to "pause" QE at the February meeting was probably a relatively difficult one, and we expect the vote (which will be revealed in the MPC minutes released next Wednesday) to have been close.
Further Detail
Cautious tone: The overall tone of the November Inflation Report and press conference was cautious. Governor King underlined that the door was open to a resumption of QE. In his opening statement, Governor King referred to the decision last week to not extend QE as a "pause" and said "it is far too soon to conclude that no more purchases will be needed". He also said that "the Committee will keep its options open, and further purchases will be made if they prove necessary to keep inflation on track to meet the target in the medium term". Again, the high degree of uncertainty in the forecasts was mentioned and reflected in the fan chart forecasts: "the Committee judges that inflation is, on balance, more likely to be below the target than above it for much of the forecast period. Though by the end the risks are broadly balanced".
Projections more downbeat than we'd expected: As expected, the near-term forecasts for inflation were revised up and the forecast for GDP growth looks somewhat lower than in the November Inflation Report. The medium-term forecasts though were not quite what we'd expected, given the MPC's decision to pause QE at the February meeting. We had thought that its central forecast for inflation (the one using market interest rate expectations, which now incorporates later monetary policy tightening than in November), would show inflation a bit higher than in November in the medium term. In contrast, the most likely (modal) profile for inflation on its forecasts has been lowered and looks further below the target two years ahead than it did in November. In the press conference, Deputy Governor Bean said that "the movements in the central projection of inflation are a reflection to a large degree of what has happened to the growth projection". When questioned in the press conference about whether the looser monetary policy profile had made any difference to the BoE's forecasts, he said that "it has some effect but I mean we've also made obviously various other changes in the forecasts, slightly tweaking estimates of the margin of spare capacity, of how quickly QE is passing through - all those sorts of things". Hence, these other factors appear to have more than offset the looser monetary policy profile incorporated. The strength of these ‘offsets' is most clearly apparent when comparing the February and November projections incorporating unchanged monetary policy (0.5% interest rates and the stock of asset purchases steady at £200 billion).
On not extending QE: This raises the question of why QE wasn't extended at the last policy meeting. However, one of the key messages that the MPC seemed keen to stress is that the average outlook going forward is pretty much unchanged from the last report in November. It considers the whole of the distribution, not just the mode. We won't be able to see the actual difference between the mode and mean of its forecasts until the numerical projections underlying the fan charts are released next week. However, the Inflation Report describes the downside risks to the MPC's GDP growth forecast as smaller than in November and "the risks around the most likely path for inflation are judged to lie to the upside".
Likely a split vote in February: We think that the vote to pause QE will have been close. We think that MPC members will have been somewhat divided on the likely amount of spare capacity in the economy and how long that is likely to linger. Some may also be a touch concerned to see some measures of medium-term inflation expectations picking up. The tone of the press conference and downward revision to its modal forecast for inflation (despite a similar ‘average' outlook) also don't sit completely comfortably with the MPC's decision not to extend QE in February. We would expect the vote to have been split, but think that the decision to pause QE will have been difficult and that the vote could have been very close.
Differences with our forecasts: The MPC central forecast for GDP growth remains more optimistic than our own. Although the MPC doesn't provide quantitative detail on projections, the comments in the Inflation Report suggest that, in particular, we may be envisaging a weaker profile for investment and stock-building. We would also probably incorporate more fiscal tightening. Our medium-term central inflation forecast, however, looks higher than theirs does. This likely reflects that the Bank of England envisages spare capacity lingering longer than we do ("output is also judged likely to remain some way below capacity throughout the forecast period", where their forecast period runs until the start of 2013). It might also reflect a higher weight on our part to the role of anchored inflation expectations (and we note the Inflation Report chart suggesting that some survey-based estimates of medium-term inflation expectations have risen) in determining medium-term inflation.
We think that inflation may continue to surprise the MPC's central forecast on the upside this year. At this point, we still think that the MPC will likely raise rates late this year (November) and do not change our forecasts. However, the tone and forecasts in the Inflation Report continue to highlight that the balance of risks to our interest rate forecasts is in favour of later rather than earlier tightening.
Other Interesting Comments (Credit Rating, the Special Liquidity Scheme and Exit Strategy)
1) Governor King was asked several times about the fiscal outlook during the press conference. Interestingly, one of his comments was that a downgrade to the UK's credit rating would be "peculiar" and he couldn't think of any reason why the AAA rating would be reduced.
2) On the SLS, Governor King's verbal comments indicated that the BoE is adamant that the SLS would not be extended, despite references in the Inflation Report to, for example, the "significant amount of funding maturing over the next few years" for banks and "the cost and availability of funds to replace maturing liabilities will be an important influence on banks' lending capacity". We continue to think that conditions in the banking sector will weigh on the growth outlook over the next few years. Interestingly, the MPC sounded somewhat more sanguine on this: "overall, the Committee judges that access to the capital markets and recourse to internal finance should enable output to grow at healthy rates without requiring a significant pick-up in net bank lending to UK households and companies".
3) On exit strategy, in the press conference it was reiterated that the MPC intends to return the gilts purchased back to private hands. Governor King also ran through the instruments that the Bank has: "The instruments are very simple - it's bank rate, it's asset sales. If we need to mop up liquidity in the short run we can do it by issuing bank bills. These are all tried and tested instruments - we've used them - there is nothing new here. The US faces slightly different problems in trying to devise new instruments - we don't need to do that. So the instruments are very clear. The big judgment in question is timing". We continue to think that asset sales will most likely start at the point that interest rates start to rise, but that the emphasis of policy tightening will be on interest rate changes. We still expect a very gradual profile of asset sales.
January exports demonstrated further strength, although seasonality resulted in sequential moderation: After the strong rebound in December, China's exports showed further recovery in January, rising 21%Y to US$109.5 billion. This was very close to our forecast of 21.6% growth, although it missed consensus expectations of 28%. On a seasonally adjusted basis, however, shipments dropped by 5.5%M (versus +5% in December), suggesting that the headline YoY strength was mainly supported by the low base and Lunar New Year (LNY) effects (LNY fell at the end of January last year versus February this year). As in recent months, imports outpaced export growth, given China's relative demand strength, and this continued to put pressure on the trade surplus, which narrowed by 64%Y to US$14.2 billion in January.
Strong domestic demand lifts imports: While we had forecast an aggressive 75%Y jump in imports in January, the result turned out to be even stronger: shipments actually rose 85.5%, although, on a MoM basis, imports slipped by 0.9%. In the primary product space, the biggest contribution to the surge in imports was seen in crude oil (+143%Y in value, +82% in price) and primary plastics (+116% in value, +36% in price). Nevertheless, we believe that the intake of manufactured goods contributed much more to the pick-up in import growth, paving the way for sustained robust export growth ahead, given the high import content of China's exports.
Monetary data - loan creation was well behaved following prompt policy action: Contrary to earlier rumors that new loans surged above Rmb1.6 trillion in the first couple of weeks in the month, they totaled Rmb1.39 trillion in the full month. We believe this is attributable to early policy action, namely the RRR hike on January 12. This amount represents a 14%Y decline, consistent with the normalization trend we envisage for the year, and eases YoY outstanding loan growth to 29.3% (versus our forecast of 30%), from 31.7% in December.
Monetary data - broad money growth slowed in line, M1 growth possibly biased upwards by LNY: Deposits (+27.3%Y in January versus +28.2% in December) and M2 (+26% versus +27.7%) growth have trended downwards in line with the gradual normalization in loan growth, very much in line with our forecast (we forecast M2 growth at +26.2%). The pick-up in M1 growth to 39%Y (+32.4% in December) may appear alarming, but we attribute this to the reallocation of funds towards more liquid forms ahead of LNY. Indeed, the PBoC has been injecting liquidity to the interbank market in the last few weeks through open market operations (reduced bond issues relative to maturing bonds, and temporarily suspended repos) possibly in preparation for LNY demand. To recap, the withdrawal of cash at the end of January 2009 created a high base for M0 and a low base for M1, resulting in the drop in M0 (-0.8% in January) and higher M1 growth this year.
Inflation - consumer inflation milder than expectation: January CPI rose 1.5% Y, noticeably milder than our (+1.9%) and especially consensus (+2.1%) forecasts. Prices rose 0.6%M before seasonal adjustment, but slipped 0.4% if seasonally adjusted. Both food (+3.7%Y versus our forecast of +4%, +5.3% in December) and non-food (+0.5% vs. +0.8%, +0.2% in December) inflation came in below expectations.
Inflation - upstream inflation heads up further nevertheless: Although consumer inflation came in milder than expectations in January, upstream inflation continued to head upwards, meaning that CPI inflation still sees further upside in the months ahead (preliminarily we forecast around 2%Y in February). Primarily attributable to the rise in energy prices (+19.3%Y in January versus +10.5% in December), RMPPI inflation jumped to 8%Y in January (+3% in December) only two months since it returned to positive territory. On a broader scale, PPI inflation picked up to 4.3% in January (+1.7% in December), led by raw materials (+8.6% versus +3.6%) and energy (+31.4% versus +17.6%), while manufacturing processing producers priced up their goods YoY (+0.8% versus -0.7%) for the first time since November 2008. On a MoM seasonally adjusted basis, we estimate that PPI gained 1.3% (+1.8% in December), while RMPPI rose 2.6% (+2.4% in December).
Conclusion and policy implications: The January datapack, although not complete (as retail sales, FAI and industrial production data will not be released until March for January-February in aggregate), offers a general picture of further improvement in macro momentum, but the results can be considered moderate. Specifically, this set of data, contrasts with that for December, which featured a marked acceleration in key indicators and should hold policymakers back from shifting the policy stance too quickly and aggressively. The well-behaved monetary data, in particular, should reduce the risk of a policy intervention surprise in the very near term and should be well received by the market.
In light of the latest desired developments, we reiterate our policy outlook as detailed in our recent reports, Upgrade 2010 Forecasts on Improved External Outlook, February 3, 2010, and Déjà Vu: Dissecting Heightened Uncertainty, February 8, 2010. We do envisage policy normalization throughout this year, but believe the pace of this normalization will be measured, in line with the improvement in the global economy. We expect to see a first interest rate hike of 27bp in 2Q10, followed by two more hikes in 3Q and 4Q. On the other hand, we stand by our call that renminbi appreciation will not resume until 2H10.
Hampered by sovereign uncertainties in Europe and the downside risk to global growth, it comes as no surprise to us that the Bank of Korea (BOK) kept its policy rate unchanged at 2% in its monthly meeting on February 11. The previous day's surprisingly high unemployment rate would likely also have contributed to the rate freeze decision, although we believe there was a one-off factor that caused the jobless rate to surge last month (see Reasons Behind Today's High Unemployment Rate, February 10, 2010).
March will likely be a crucial month because it will be the last for the current monetary policy committee to meet before three out of the seven members (including Governor Lee) are replaced by new members in April as their terms expire. If the current committee is in favour of rate hikes, it will be their last opportunity to do so in March. If there is no hike in March, then the chance of a rate hike in 2Q will be very slim, in our view, as the new members will likely need a period of observation before they make a decision.
Judging from the official BoK statement and Governor Lee's speech on the same day, it appears that the chance of a rate hike in March is small. The committee seems less certain of the coming growth outlook, and the risk of debt crisis in Europe was cited. The major changes in the BoK statement from last month also include the addition of "stable" inflation, "high volatility" in exchange rates and stock prices versus "a sharp run-up" last month, and mortgage lending having "slowed" compared to "grown on steady scale" last month. Governor Lee's speech was also dovish. Instead of saying recovery, he said the economy has not fully normalised. He also said there is "no direct threat from prolonged low rate", which stands in sharp contrast to his previous speeches. Governor Lee commented that the central bank will raise interest rates when the real economy shows signs of recovery.
Korea's production and consumption levels have all gone back to pre-crisis levels, and exports are clearly outperforming. We would have thought these to be strong signs of recovery, and in fact of one of the strongest-recovering economies in the world. If these are not considered signs of recovery, then it could mean that Korea will have to wait longer to raise interest rates, again pushing the timing of a rate hike possibly to 2H10.
In our view, there is a risk to global growth, yet inflation risk is also not to be underestimated. As of now, we can see inflation picking up as a fact, but double-dip risk remains only a possibility. We agree that there is not much domestic inflationary pressure in Korea due to effective anti-speculative measures imposed on the property market and regulated utility tariffs. Capacity is getting tighter as production surprises on the upside, but the latest rebound in capex should help to offset this. Nevertheless, rising inflation in the rest of the region could spill over into Korea, and we believe that such a trend should be closely monitored.
Read Full Article »