Category: Money Moves Markets

Equities and other risk assets have rallied partly on hopes of easier monetary policies but major central banks have yet to inject additional liquidity into banking systems.

Previous posts discussed the leading relationship between the US monetary base (i.e. currency plus bank reserves) and US / global equities in operation since early 2009. Stock markets peaked in late April 8-9 weeks after a high in the monetary base in late February while the rally starting in early July similarly followed a trough in the base 8-9 weeks before in early May. The monetary base last week fell to its lowest level since the May bottom, suggesting increased near-term risks for markets – see first chart.

Eurozone trends may also be relevant and are similarly cautionary, with the monetary base in the week before last at its lowest since December (last week’s figures are released tomorrow). The recent large decline reflects a fall in demand for ECB credit from stronger banks, partly because new 12-month fixed-rate loans are no longer available. The reduction in excess cash in the system, however, has contributed to unwelcome upward pressure on market rates – second chart.

The 1.1% jump in GDP in the second quarter is partly pay-back for below-par performance in the first quarter, reflecting the VAT rise – which may have caused firms to restrain production temporarily – and weather disruption. The average gain of 0.6%, or 2.5% at an annualised rate, over the last three quarters is a better guide to the economy’s underlying path and accords with the message from business surveys and labour market data since late 2009. Non-oil growth has been slightly stronger, at 2.7% annualised.

A monthly GDP proxy derived from data on services and industrial output fell by 0.3% in April but surged 1.1% in May – see chart. The 1.1% quarterly growth estimate appears to incorporate an assumed 0.5% decline in June, for which the Office for National Statistics (ONS) currently has limited information. This indicates the possibility of an upward revision. The May level of GDP was 0.5% above the quarterly average.

Strong second-quarter growth should close off discussion of extending asset purchases at the August Monetary Policy Committee (MPC) meeting and supports Andrew Sentance’s argument that the current policy stance is increasingly misaligned with a strengthening economic recovery and ongoing inflation overshoot. The “MPC-ometer” model discussed in previous posts, indeed, suggests that interest rates would now be rising, if the Committee were responding to economic and financial data in the same way as in the past.

The US Conference Board index of leading indicators fell by a smaller-than-expected 0.2% in June and is still up by 2.6% over the last six months, suggesting a solid near-term economic outlook.

The bulk of the recent gain, however, is due to the component measuring the slope of the yield curve – the gap between the 10-year Treasury yield and the fed funds rate. This component has been contributing 0.3-0.4 percentage points to the monthly change in the index.

A positive yield curve slope is less likely to signal economic strength when short-term interest rates are unusually low. The US economy suffered a recession in 1937-38 when the yield curve was only slightly less steep than currently – see previous post.

A modified leading index excluding the yield curve rose by only 0.4% in the six months to June – see chart – and has fallen by 0.9% over the last three months.

The view here remains that a “double dip” will be avoided as long as real narrow money continues to expand – it rose solidly in June – and the 10-year yield remains above 2.5%. Economic data, however, may continue to disappoint over the next few months.

MPC inflation doves argue that slower core trends in other countries support the view that the “Phillips curve” is still working. According to the June minutes, “the recent downward trends in inflation, excluding energy and food, in the United States and euro area suggested that a substantial margin of spare capacity would cause inflation to fall back in the United Kingdom too, as the impact of temporary factors wore off.”

This argument, however, assumes that the observed falls in US and eurozone inflation have been caused by spare capacity. A closer examination suggests otherwise.

In the US, the annual increase in consumer prices excluding food and energy has fallen from a peak of 2.5% in August 2008 to 0.9% by June. This mainly reflects a slowdown in the “shelter” component, which has a 42% weighting in the core index. Annual inflation excluding food, energy and shelter has fallen only from 2.7% to 2.1% over this period, remaining above its average in recent years – see first chart.

The shelter component is dominated by “owners’ equivalent rent” – a theoretical sum paid by home-owners to themselves in return for accommodation. Its annual rate of change has fallen from 2.5% in August 2008 to -0.2% by June, a lagged reflection of housing market weakness. It is debatable whether imputed rent should be included in a cost of living index – it is omitted from UK and eurozone indices. In any event, the decline in US shelter inflation cannot be attributed to “spare capacity” in the sense of underutilised plant and machinery or labour.

Doves retort that housing weakness cannot explain the similar slowdown in eurozone core inflation (i.e. based on the consumer price index excluding unprocessed food and energy) from an annual 2.6% to 0.9% between August 2008 and June. This decline, however, owed much to a 17% rise in the euro’s effective exchange rate in the three years to July 2008 – second chart. The 12% fall in the euro since last October may be starting to lift core inflation, which bottomed at 0.75% in April.

The “big picture” is that global core inflation has proved much stickier than predicted by Keynesian models based on the large decline in output during the “great recession”.

Public sector borrowing was higher than expected in June, but recent figures remain consistent with an undershoot of the Budget forecast for 2010-11.

The Office for Budget Responsibility (OBR) projects a fall in net borrowing, excluding the temporary effect of financial interventions, from £154.6 billion in 2009-10 to £149 billion this year. Attempting to adjust for seasonal factors, borrowing averaged £12.25 billion per month between April and June, or £147 billion annualised – see chart. The OBR forecast, therefore, implies no further improvement – even a small worsening – over the remainder of 2010-11.

This is unlikely because, first, the benefits of economic recovery should grow as the year progresses and, secondly, the coalition has announced spending cuts and tax rises worth £8.1 billion in 2010-11 that have yet to be reflected in the monthly numbers. Put differently, even assuming no impact from the recovery, these measures together with the recent run-rate imply borrowing of £139 billion this year (i.e. £147 billion minus £8 billion), £10 billion less than the OBR’s forecast.

The forecasting approach in this journal places weight on the Friedmanite rule-of-thumb that changes in the real money supply lead demand and output by roughly six months. It is advisable to monitor the full range of money measures but narrow money, M1, has exhibited the most consistent relationship with the economic cycle historically.

The Friedmanite rule has worked well in recent years, albeit with a slightly longer lead than usual. The six-month rate of change of G7 real M1 turned negative in late 2007 ahead of the recession and bottomed in March 2008, 11 months before the six-month change in industrial output and nine months before that of the OECD G7 leading index, which is widely monitored by market participants. Real M1 expansion surged in late 2008 and early 2009, foreshadowing the strong recovery in G7 industrial activity over the last year – see first chart.

Monetary trends also signalled the current global industrial slowdown, with the six-month rise in G7 real M1 peaking as long ago as February 2009. The six-month increase in the OECD leading index reached a high eight months later in October 2009 with that of industrial output following in January 2010 – the same 11-month lag as at the cycle trough.

Importantly, however, the six-month change in real M1, while falling significantly, has remained positive in recent months, consistent with an economic slowdown rather than renewed contraction. It bottomed, moreover, in January 2010, picking up in May and, provisionally, June (the June estimate incorporates US and Japanese data only). Based on the lags at the prior two turning points, this suggests that the six-month change in the OECD leading index will bottom in September or October and that of industrial output in December.

Previous posts have explained that equities and other risk assets have, on average, performed better when real M1 has been growing more strongly than industrial output. The six-month change in real M1 crossed above that of the leading index in June, probably signalling a move above output expansion by late summer or early autumn. This could indicate an improving liquidity backdrop for markets later in 2010.

Monetary pessimists ignore M1 and emphasise recent contraction in G7 real broad money – second chart. The broad measure, however, failed to warn of the recession, continuing to grow strongly in late 2007 even as narrow money was weakening. The interpretation here is that the demand to hold broad money rose sharply as the financial crisis intensified but has been falling more recently, partly reflecting heavily-negative real deposit interest rates. Broad money, therefore, understated monetary tightness in late 2007 and is grossly overstating the degree of restriction now. Portfolio shifts are much less likely to have affected the demand for M1, which is more closely related to economic transactions.

Chinese consumer price inflation fell from an annual 3.1% to 2.9% between May and June but growth in the wider GDP deflator measure rose to 5.6% in the second quarter from 4.6% in the first – see chart. This is above the average of 4.4% since 2000.

Annual GDP deflator inflation has averaged 2.5 percentage points more than CPI inflation since 2000 but, unusually, the gap closed temporarily during the “great recession”, which, apparently, bypassed China. This could reflect price weakness in non-consumption GDP components; it is also consistent with a massaging down of the deflator data in order to boost published real GDP growth rates.

Deflator inflation peaks of 8.8% in 2004 and 11.0% in 2007 followed highs of 20% and 23% respectively in annual growth in narrow money, M1 – see chart. With M1 expansion having reached 39% last year, it would be surprising if second-quarter inflation of 5.6% marks the peak in the current upswing.

Chinese monetary policy is on hold but inflationary risks from the monetary overhang suggest no scope for growth-boosting actions.

US private job openings (vacancies) slipped back in May but the three-month moving average continued to climb, suggesting a further rise in employment – see first chart. (Private-sector numbers have been used to avoid distortions caused by temporary government employment to conduct the decennial census.)

UK vacancies are also on the up again after a temporary dip in early 2010, possibly related to election uncertainty, with an increase in private openings in the three months to June offsetting a fall in the public sector – second chart. Employee jobs in the quarterly workforce survey were still falling in the first quarter but employee numbers in the (more volatile) labour force survey increased in the three months to May.

The fall in annual consumer price inflation from 3.4% in May to 3.2% in June reflects a slowdown in fuel costs. “Core” trends remain stubborn, with a recent acceleration in services inflation providing further evidence that spare capacity is failing to exert the dampening impact expected by the Monetary Policy Committee.

Other points:

• CPI inflation averaged 3.44% in the second quarter versus the Bank of England’s 3.30% estimate in the May Inflation Report. This is the fifth quarter out of the last six that the Bank has underestimated current-period inflation.

• “Core” inflation excluding energy, food, alcohol and tobacco moved back up to its recent high of 3.1% from 2.9% in May.

• Core resilience reflects a pick-up in services inflation from 3.0% in January to 3.9% currently. Inflation optimists argued that services trends would slow in response to excess capacity, with a weak exchange rate having less offsetting impact than in manufacturing. The recent acceleration, however, had been suggested by business surveys – see chart.

• Goods inflation, excluding energy and food, has fallen from 3.3% to 1.9% since January, partly reflecting a stabilisation of the exchange rate. Imported pressures, however, remain strong, with manufactured import prices rising by 5% in the first five months of the year, according to May trade figures released last week.

• The Bank is likely to be forced to revise up its forecasts of 2.54% and 2.28% in the third and fourth quarters significantly in the August Inflation Report. On conservative assumptions, inflation may average 2.75% in the fourth quarter before returning to 3% in early 2011 as VAT is raised.

• The second-quarter outturn of 3.44% compares with a central projection of 0.73% a year earlier in the May 2009 Inflation Report. Bank officials have deflected criticism of this forecasting failure but an internal review may be under way, with the Bank’s recent Annual Report stating that the budget contains an allowance for expenditure of £2.4 million on a “new forecasting model”.

US recessions since the mid 1950s have been preceded by a flat or inverted Treasury yield curve. Some commentators take comfort from current curve steepness but the rule-of-thumb has been rendered obsolete by near-zero short-term interest rates.

A summary measure of the slope of the Treasury curve is the gap between the yield on 10-year bonds and the discount rate on three-month bills. This has averaged 1.4% since 1955 but fell below 0.3% before the last nine recessions, turning negative in six cases – see chart.

The three-month Treasury bill rate, however, was at least 3% across these nine episodes versus only 0.2% today. It is necessary to examine earlier recessions to gauge a “warning level” for the gap when short rates are very low.

The four downturns between 1935 and 1955 each occurred against the backdrop of a positively-sloped curve. The closest parallel with today is 1937-38, when the three-month bill rate fluctuated in a 0.2-0.6% range. The curve flattened before the recession but the 10-year / three-month gap never fell below 2.3%.

This 2.3% minimum may be a reasonable guide to a recessionary level of the gap today. Assuming that the three-month bill rate remains at 0.2%, this implies that the 10-year yield, currently 3.0%, would have to fall to 2.5% or below to generate a warning signal.

Interestingly, the suggested critical level of 2.3% for the yield gap is close to the 2.5% used by John Hussman in his original list of recession-spotting criteria – see previous post. The newer version of his list, however, employs a higher value of 3.1%, contributing to his recent recession call (since the current reading is 2.8%).

As explained in the earlier post, 10 out of 11 post-war US recessions have been preceded by a contraction of real narrow money – currently still expanding. The central scenario here will remain a temporary slowdown rather than a “double dip” unless monetary trends deteriorate and / or the 10-year yield falls beneath 2.5%.