Japanese monetary trends remain lacklustre despite ultra-aggressive QE. Six-month expansion of narrow money M1 slipped again in September while growth of broad M3 was stable. Bank lending* has also failed to accelerate – see first chart.

As previously explained, the muted response of monetary growth to QE partly reflects stepped-up sales of government securities by banks, which have offset the Bank of Japan’s purchases. Holdings of such securities by domestically licensed banks** fell by ¥24.6 trillion between March and August (the latest available month) versus an increase of ¥38.0 trillion at the BoJ over the same period.

Monetary growth is weaker in real terms because of a pick-up in inflation, mainly due to the weaker yen. The six-month change in real M1 in September was the lowest since July 2012, while that of real bank lending turned negative in August – second chart. Real narrow money expansion is lower in Japan than in other major economies – third chart.

Real money growth, of course, will be squeezed further by the April 2014 sales tax rise from 5% to 8% – expected to boost consumer prices by about 2.5%.

The optimistic scenario is that banks will stop selling government securities, allowing ongoing QE to boost nominal money growth substantially before the tax hike hits. There is some support for this scenario in statistics showing a sharp slowdown in the rate of decline of their holdings in July and August*. The September monetary data, nonetheless, were disappointing. The assessment here of Japanese economic prospects will remain cautious until monetary trends improve significantly.

*The lending series in the chart adjusts for special factors, data for which are available only up to August.
**This category is a sub-set of the banking sector, which also includes Japan Post, among other institutions.
***Banks’ holdings fell by only ¥1.0 trillion July / August versus a ¥23.6 trillion reduction during the second quarter.

Leading indicators followed here suggest that the global economy will expand respectably through early 2014 (at least), a message supported by monetary trends.

The global leading indicators are derived from the OECD’s country leading indices, which combine information on economic and financial series that tend to move ahead of demand and output. The composition of the indices differs by country; the US index, for example, includes business and consumer survey responses, housing starts, new orders, weekly hours, share prices and the yield curve.

The OECD country-level data are aggregated and transformed to produce short- and longer-term leading indicators that have led turning points in global industrial output growth by an average of three and five months respectively in recent years.

The short-term leading indicator rose slightly for a second month in August, suggesting a modest pick-up in output expansion into late 2013. The longer-term measure, however, eased back after a recent gain, consistent with growth levelling off around year-end – see first chart.

The small decline in the longer-term indicator follows a reduction in global real narrow money expansion in June / July – monetary trends lead the economy by about six months, according to the “monetarist” rule. Real money growth, however, partially reversed this decline in August and remains historically solid – second chart. Monetary trends, in other words, are consistent with continued if moderate economic expansion through early 2014.

The emerging E7 component of the global longer-term leading indicator has risen recently, while the G7 component has declined – third chart. This coincides with a significant downgrade in the IMF’s growth forecast for emerging economies – the IMF always reflects the consensus view so is often a good contrarian indicator. These signals suggest adding to emerging market equities if forthcoming monetary data confirm that E7 real money growth has moved above the G7 level.

Peripheral Eurozone equity markets* rose by only 1.5% during 2012 versus a 13.2% gain for the MSCI World index. So far this year, they are up by 15.6% versus a 13.8% increase in the global index. This turnaround reflects better monetary trends and an associated improvement in economic prospects.

The first chart shows year-to-date price performance of various markets relative to MSCI World. Peripheral markets were weak during the first half but have surged since mid-July. They are now ahead of core markets and closing in on the US and Japan; the UK and emerging markets have underperformed, as have Canada and Australia.

The rebound follows a recovery in monetary trends. The six-month change in real narrow money in the periphery was still negative in December 2012 but had risen to the top of the global ranking by April 2013 – second chart. The turnaround reflects ECB support for peripheral economies since late 2011 in the form of rate cuts, long-term liquidity operations and the "outright monetary transactions" (OMT) programme – these measures have restored confidence, causing capital flight to reverse and increasing the demand to hold narrow money as spending intentions revive.

Will outperformance continue? The recent surge has probably been driven by underweight investors covering their positions. This process could extend further but six-month real money growth in the periphery has moderated since April and is now in the middle of the global pack – second chart.

The strong rebound in peripheral equities may offer a cautionary message to investors expecting further weakness in emerging markets: surveys indicate that these markets are significantly underowned, while emerging E7 monetary trends have improved since late 2012 – second chart.
 
*Spain, Italy, Ireland, Portugal and Greece, weighted by market capitalisation.

Two MPC members would be voting to raise Bank rate this week if the Committee were responding to incoming news in the same way as in the past, according to the “MPC-ometer” model followed here.

The MPC-ometer is designed to forecast the “average interest rate vote” of Committee members based on a small number of economic and financial inputs relevant for assessing the outlook for growth and inflation. Estimated in 2006, the model proved useful for predicting interest rate changes in the late 2000s; it also signalled the expansions of QE in 2011 and 2012*.

The model reading for October is +5 basis points (bp), suggesting two votes for a 25 bp Bank rate rise and seven for no change. The reading has risen from a recent low of -9 bp in May, when the MPC was contemplating further easing – three members voted to expand QE every month between February and June.

Some commentators suggest that a majority of the “old” MPC would now be in favour of raising Bank rate based on recent strong purchasing managers’ surveys – historically an important influence on the Committee’s thinking. The PMIs are included in the MPC-ometer but their influence has been countered by declines in several inflation indicators – in particular, consumer survey price expectations and CBI manufacturers’ price-raising plans. Financial market inputs, meanwhile, have yet to give a strong signal of imminent tightening.

Looking ahead, the MPC-ometer reading for November will depend importantly on the first estimate of third-quarter GDP growth released on 25 October. A moderately strong result, however, would not be sufficient, on its own, to push the model over the rate hike threshold**.

The new forward guidance framework, of course, suggests that the MPC will begin raising Bank rate later than in previous cycles, although Committee members have claimed that the framework simply makes explicit its existing “reaction function”. The guess here is that the MPC-ometer will cross the rate rise threshold around end-2013 but the Committee will delay acting until summer 2014, by which time unemployment will be at or close to 7.0% and inflation entering another upswing.

*The model was modified in 2009 to incorporate QE, with the relevant parameter implying that £75 billion of gilt-buying is the policy equivalent of a quarter-point rate cut.
**A quarterly GDP rise of 1.2% or more would be required, assuming no change in the other inputs.