Category: Money Moves Markets

The US M2 money measure contracted over the summer, leading some commentators to expect the economic recovery to falter in late 2009, with attendant deflationary risks. Prior posts argued against this view on the grounds that the M2 decline was modest and followed strong growth, while the velocity of circulation was likely to be rising in response to miniscule interest rates and reviving risk appetite.

M2 now seems to be picking up again. It rose by 0.3% in both September and October and weekly numbers suggest a similar rise in November. Implied three-month annualised growth of about 4% is likely to be sufficient to support further solid economic expansion and could even be too strong if velocity rises by 4-5% over the next year, as an earlier post argued was possible.

Another approach to measuring monetary backing for economic growth is to add together broad money changes and mutual fund flows – the idea here is that fund flows are likely to be inversely related to the demand to hold money so provide an indirect measure of changes in velocity. As the chart shows, this indicator – the green line – has continued to expand robustly recently. (A post last week presented a similar chart for the UK.)

The concern for markets in early 2010 is not that monetary growth will be insufficient to support a sustained economic recovery but that stronger expansion and higher inflation will fully absorb the available supply, implying an absence of “excess” liquidity to push asset prices higher – in marked contrast to this year.

A recent post suggested that the surprise of the next six months would be the speed of the turnaround in labour markets. US payroll employment continued to decline in November but the fall of 11,000 was much smaller than expected while earlier months’ figures were revised up substantially.

The details of the payrolls survey were also encouraging, showing rises in weekly hours and temporary jobs – both tend to lead overall employment. The unemployment rate, based on a survey of households, retreated from 10.2% to 10.0% last month but this partly reflected a contraction of the labour force while an equivalent jobs measure from this survey was weaker than “official” payrolls.

The official series looks poised to  grow from December, judging from a jobs gauge based on the ISM manufacturing employment index, the monthly lay-off tally by Challenger, Gray and Christmas and the NFIB small firm hiring plans index – see chart.

Better labour market news will reduce worries about a “double dip” but will call time on current super-loose monetary policies, suggesting the removal of an important support for equity markets and other “risky” assets in early 2010.

GDP can be measured in three ways, by summing output, expenditure or income across the economy. The Office for National Statistics (ONS) relies on output information for early estimates of GDP growth, with expenditure and income data incorporated during the subsequent revision process.

The fall in GDP in the third quarter was last week revised from 0.4% to 0.3%. Many economists expect a further upgrade as the ONS improves its measurement of output and incorporates more information on expenditure and income. These hopes, however, are not supported by early data based on the alternative approaches.

Separate output, expenditure and income measures of GDP can be calculated from background information supplied by the ONS. The output measure fell by 0.3% last quarter – the basis for the published GDP drop – but the expenditure and income measures registered larger declines of 0.5% and 0.9% respectively. (These figures refer to “unaligned” estimates excluding adjustments to reduce discrepancies between the three approaches.)

The chart shows published GDP at constant market prices, indexed to calendar 2005, together with the three underlying measures. In level terms, the expenditure measure was in line with published GDP last quarter but the output and income measures were lower. A simple average of the three was 0.2% below published GDP. This indicates that the published measure already incorporates an assumption of future upward revisions to underlying data.

The chart also reveals a disagreement between the three measures about the start-date of the recession, with output – and published – GDP peaking in the first quarter of last year but the expenditure and income measures one quarter later.

Recent sluggish broad money growth in the US and UK is unlikely to signal economic weakness since investors are voluntarily shifting out of cash in response to low interest rates and perceived opportunities in markets. Reduced money demand releases additional liquidity to support economic expansion.

In further evidence of this shift, UK retail buying of mutual funds (unit trusts and OEICs) was again solid at £2.4 billion in October, according to Investment Management Association figures. Inflows are on course to exceed £25 billion for the year as a whole, well above the previous annual record of £17.7 billion in 2000.

The chart shows six-month growth in “M4 excluding other financial corporations” – i.e. money holdings of households and non-financial firms – together with a six-month running total of retail mutual fund flows, expressed as a percentage of the M4 measure. Fund buying has been on a similar scale to the rise in the money supply recently.

In the current context, the sum of money growth and mutual fund flows is probably a better guide to liquidity support for the economy than M4 itself. This indicator – the green line in the chart – bottomed in late 2008 and recovered significantly during the first half, with faster expansion maintained recently. This supports hopes of an imminent economic pick-up. (See previous post for a similar US analysis.)

The Bank of England this week revealed that it provided covert “emergency liquidity assistance” (ELA) to RBS and HBOS between 1 October 2008 and 16 January 2009, with the combined loan peaking at £61.6 billion on 17 October 2008. Why were analysts unable to uncover this operation from the Bank’s weekly balance sheet, the “Bank return”?

The RBS / HBOS loans, like the earlier Northern Rock facility, were probably recorded under “other assets” on the Bank return. “Other assets” were on a declining path from early 2008, reflecting the repayment and eventual transfer to the Treasury of the Rock loan. They surged, however, after Lehman’s failure, rising from a low of £21 billion on 10 September 2008 to a peak of £170 billion on 22 October. It now appears that about £60 billion of this increase was due to the RBS / HBOS loans.

The problem for analysts at the time was that “other assets” were simultaneously being boosted by US dollar repo operations, involving the Bank borrowing from the Federal Reserve under a swap arrangement and lending on to banks short of dollar funding. These operations began on 18 September 2008 and rose to a peak around the same time as the RBS / HBOS loans.

While it was possible to track the Bank’s dollar lending to the market, the swap facility could, in theory, have involved the Bank holding additional dollar cash in a reserve account at the Fed. This would also have been included within “other assets”. So although the dollar loans were smaller than the rise in “other assets”, it would have been a leap to conclude that the Bank was engaging in additional covert lending.

The hypothesis that the surge in “other assets” was due to the swap arrangement was seemingly supported by a similar change in “other liabilities” on the Bank return – these would have included the borrowing from the Fed and rose from £16 billion to £158 billion between 10 September and 22 October 2008. In fact, part of this increase was probably also connected with the RBS / HBOS support. In particular, “other liabilities” may have included a loan from the Treasury / Debt Management Office (DMO) to the Bank to finance the ELA operation, with the DMO funding the loan via debt sales.

“Other assets” stood at £192 billion last week (18 November) but now include a loan of £180 billion (19 November) to the asset purchase facility (APF). In other words, without the APF “other assets” would have fallen back to £12 billion – the level in early September 2007, just before Northern Rock imploded.

The recent recession and current revival in world industrial output bear a close resemblance to the deep 1974-75 contraction and subsequent recovery. This comparison suggests an ongoing economic upswing during 2010 but with growth momentum slowing as the year progresses.

A weighted average of industrial output in the Group of Seven (G7) major economies and seven large emerging economies (the “E7”) fell by 14% between February 2008 and February 2009 but had recovered by 7% by September – see earlier post for more details. The decline is similar to a 12% drop in G7 industrial output between May 1974 and May 1975 – the previous biggest post-war recession. It is reasonable to use G7 output as a proxy for world industrial activity in the 1970s, when emerging economies were much less significant.

Seven months after the May 1975 trough G7 industrial output had recovered by 5% versus the 7% rise in G7 plus E7 production between February and September this year. The first chart overlays the 1970s G7 path on current G7 plus E7 data, aligning troughs and rebasing the G7 series to equal G7 plus E7 output at the start of 2003. The comparison suggests that the recovery will be sustained during 2010 but  momentum will slow progressively from the spring into 2011. Output could regain its February 2008 peak level in October next year.

The downshift in momentum is clearer in rate of change data shown in the second chart. The implied slowdown in annual growth from a high to be reached in February 2010 is consistent with a recent peak and likely deceleration in G7 annual real narrow money supply expansion, which leads output by about six months – see previous post. Stock markets and other “risky” assets would probably anticipate slower economic growth later in 2010, suggesting a set-back early next year, although possibly after a significant further near-term gain. As the prior post explained, the liquidity backdrop may also turn less favourable for markets in early 2010.

Industrial output in the Group of Seven (G7) major economies recovered by 5% between March and September but was still down by 10% from a year before. G7 real money supply expansion, meanwhile, on narrow and broad measures is running at an annual 9% and 4% respectively – see first chart. The large gap between the rates of change of real money and output is a measure of the “excess” liquidity that has been pushing up stock markets, commodities and other “risky” assets.

On a six-month rather than annual basis, G7 output expansion has crossed above real money supply growth, implying that “excess” liquidity is no longer being created – second chart. The monetary backdrop for markets, therefore, is less favourable than in the spring but the huge money / output divergence of the prior six months should continue to buoy asset prices into early 2010. In other words, a “sell” signal awaits convergence of annual rather than six-month growth rates. Such a signal – annual output expansion rising above real money growth – occurred in 1987 and 1994, for example, and was associated with stock and bond market weakness respectively.

Real money growth – particularly narrow money M1 – leads output expansion by about six months. The annual rate of change of G7 real M1 reached a trough in August 2008 while the annual fall in industrial output bottomed in March 2009. Real M1 growth is likely to have peaked in September 2009, with a rise in headline inflation due to commodity price effects contributing to a significant slowdown into early 2010. This suggests that annual output expansion will reach a high next spring and fall back later in 2010.

G7 real broad money contracted in the six months to October. Previous posts have argued that this is unlikely to signal economic weakness because the demand for broad money has fallen in response to low interest rates and rising risk appetite. Put differently, broad money velocity is picking up after a plunge in late 2008 and early 2009. Real narrow money is a better guide to economic prospects – growth is slowing but remains solid by past standards, consistent with a moderation of output momentum within an ongoing economic recovery rather than a return to contraction later in 2010.

A big decline in pay growth since early 2008 is often cited as a reason for expecting lower inflation. This fall, however, largely reflects cuts in bonuses and working time rather than a slowdown in hourly wages.

The official average earnings index measures pay per worker and is separable into regular and bonus elements. The chart shows annual growth in total and regular earnings together with an adjusted regular pay measure taking into account changes in average weekly hours.

Headline earnings growth has fallen from an annual 4.0% in the first quarter of 2008 to 1.2% in the third quarter of this year with the bulk of the decline due to a slowdown in regular pay expansion from 3.8% to 1.8%. This latter reduction, in turn, mainly reflects a cut in average working time from 32.2 to 31.5 hours per week since last year’s first quarter.

Regular hourly pay growth, by contrast, has slowed only marginally, from an annual 3.5% in the first quarter of 2008 to 3.4% by this year’s third quarter.

Firms may base pricing decisions on trends in hourly pay rather than earnings per worker. This is because reductions in bonuses and working hours are usually associated with lower output so do not result in a fall in unit labour costs.

The limited response of hourly pay growth to labour market weakness may partly explain recent core inflation resilience, although this mainly reflects the impact of sterling depreciation and pass-through of global commodity price rises.

The relationship between the US monetary base – currency in circulation plus bank reserves at the Fed – and the performance of stock markets, commodities and other “risky” assets remains intact. The new high reached by the Dow Industrials last week followed a 9% surge in the base between the end of September and early November – see first chart.

The base contracted in the week to last Wednesday but this is unlikely to mark a change in trend. The Fed is scheduled to buy $500 billion of mortgage-backed and other agency securities by the end of the first quarter. This cash injection will be offset by a further decline in various forms of “emergency” lending, including term auction credit, discount window loans, commercial paper holdings and liquidity swaps with other central banks. These four items, however, currently sum to $172 billion – even in the unlikely event of the total falling to zero, the effect on reserves would be swamped by securities purchases.

The Fed could, in theory, sterilise the impact of its buying on the monetary base by conducting reverse repurchase agreements (repos), involving banks lending excess cash back to the central bank in return for securities. (The Bank of England has effectively sterilised the base effects of its asset purchases since August by cutting its short-term lending to the banking system.) Such an initiative, however, is unlikely before early 2010 and should be signalled in advance to reduce the risk of a large negative market reaction.

Expectations that US liquidity supply will remain ample at least until year-end have been reflected in further capital outflows from the US to Asia in particular. Currency board arrangements in Hong Kong result in a direct impact on the monetary base, which has climbed by 16% since the end of September, providing strong support for the local stock market and Hong-Kong-listed Chinese “H” shares – second chart.

Global industrial output has already recovered half of its loss during the recession. If the rebound were to continue at its recent pace, output would regain its prior peak level by May 2010 – far earlier than expected by the consensus but consistent with the “Zarnowitz rule” that “deep recessions are almost always followed by steep recoveries”.

The top line in the first chart shows an index of combined industrial output in the Group of Seven (G7) major economies and seven large emerging economies (henceforth the “E7”) – Brazil, Russia, India and China plus Mexico, South Korea and Taiwan. A smoothed path is also plotted, based on fitting log-linear trends to the separate G7 and E7 data since 2000. This path implies a current trend growth rate of global output of 3.5% per annum.

The global output measure peaked in February 2008 and fell by 14% to a trough in February 2009. It had recovered by 7% by September 2009, equivalent to an annualised increase of 13%. If this growth rate were to be sustained, output would regain its 2008 peak level in May 2010 and cross above its trend path in June.

The lower lines show the contributions of G7 and E7 output to the global total, together with associated trends. The outsized impact of emerging economies on global trend growth is immediately apparent. The curvatures of the trends imply annualised output increases of 8.1% in the E7 but only 0.4% in the G7. Put differently, the E7 account for 3.2 percentage points of current global trend growth of 3.5% per annum. If the recent growth differential were to persist, E7 output would surpass that of the G7 by 2014.

In addition to dominating the longer-term trend, emerging economies have been responsible for 5 percentage points of the 7% recovery in global output since the February 2009 trough. E7 output has already moved ahead of its prior peak and is only marginally below its fast-rising trend. If growth were to continue at its recent pace – plausible given current policy settings and functioning banking systems – emerging economies would overheat in the second half of 2010 and 2011.

The recovery in G7 activity, while less dramatic, has still been significantly stronger than forecast by most economists in early 2009. Output fell by 20% between February 2008 and March 2009 but had recovered by 5% by September. Short-term leading indicators signal a further solid gain in late 2009 – second chart.

G7 central bankers argue that their super-loose monetary policies are warranted by domestic “output gaps” that will exert strong downward pressure on prices. These gaps, however, may be smaller and closing more rapidly than they think. The deviation of G7 industrial output from trend has already narrowed from 14% to 10%.

Policy-makers, moreover, are ignoring inflationary risks from emerging-world buoyancy. Continued above-trend E7 output growth is likely to result in further gains in global commodity prices – third chart. If emerging economies overheat, labour cost and margin elements of prices of exports to the G7 will also increase. If E7 central banks tighten to prevent overheating while G7 policies remain lax, currencies should appreciate, with the same result of higher G7 import prices.

This week’s UK Inflation Report contained no discussion of such possibilities. The Bank’s view – reflected in Governor King’s five explanatory letters over 2007-09 – seems to be that any inflation overshoot due to external factors represents an exogenous “shock” that should not affect policy settings based on “output gapology”. Given the above-described global trends, this stance implies that actual inflation is much more likely to exceed than fall below the target – as it has in all but three quarters over the last four years.