Emerging market equities have underperformed developed markets by a further 5.5% so far in 2015, according to MSCI indices. The price to book ratio of the MSCI Emerging Markets index stood at 1.55 at end-June versus 2.24 for the MSCI World index of developed markets – a 31% discount, the largest since 2003. Is it time to increase EM exposure?

Emerging equities have lagged since 2010, when six-month real narrow money growth in the “E7” large economies fell sharply to below its 1996-2009 average of 5.5% (not annualised) – see first chart. This slowdown signalled that economic prospects were deteriorating at a time of exuberant sentiment towards EM – the MSCI EM index reached a peak price to book premium to MSCI World of 18% in 2010 while 98% of global fund managers polled by Merrill Lynch claimed to be overweight.

E7 real money growth recovered modestly in 2012 but remained below the historical norm and fell away again in 2013-14. Fund managers, meanwhile, have filed for divorce – a net 20% are now underweight, representing a 1.9 standard deviation shortfall from the long-run average, according to ML.

Previous bouts of underperformance in the late 1990s and late 2000s were also associated with a sharp decline in real money growth to below average – first chart.

A simplistic strategy that would have worked well would have been to invest in emerging markets when E7 six-month real money growth was above 4% and had not fallen by more than 2 percentage points over the prior six months but revert to developed markets if either of these conditions was no longer met. Such a strategy would have outperformed a passive investment in EM by a cumulative 176% between end-1995 and end-2014, or 5.5% per annum – first chart.

E7 six-month real money growth has recovered recently but remains weak, at an estimated 2% in June. It may be advisable to await further improvement before increasing EM exposure.

How likely is a further pick-up? The second chart shows real money trends in the individual economies. The E7 aggregate has been suppressed by significant real money contraction in Russia and Brazil; a recovery in these two economies, where interest rates have respectively fallen and stabilised, could be sufficient to lift the E7 measure to the 4% “buy” level.  

The PBOC has released additional monetary data for June, allowing calculation of the “true” M1 measure monitored here. To recap, M1 is conventionally defined as currency in circulation plus demand deposits. Chinese M1, however, includes only demand deposits of corporations – a major deficiency at a time when policy is attempting to promote consumption- rather than investment-led growth. True M1 corrects this omission by adding a separate series for household demand deposits to the official M1 measure.

The six-month change in real true M1 (i.e. deflated by consumer prices) slumped in late 2014, warning of economic weakness in 2015 – see first chart. Growth has recovered but is no higher than a year ago and below that of real official M1 (and M2). The suggestion is that recent policy easing has been only partially effective and economic expansion, while reviving, will remain below par.

Corporate demand deposits, including deposits of state bodies, are rising solidly in real terms but household deposits continue to stagnate – second chart. A near-term economic pick-up based on higher corporate / state spending is unlikely to be sustained without consumer follow-through.

The stock of banknotes put into circulation by the Bank of Greece rose by a further €5.3 billion in June to a record €50.5 billion – see chart. The central bank’s liability to the rest of the Eurosystem related to the supply of notes is now €22.8 billion, on top of a €107.7 billion TARGET2 deficit.

The stock of notes is equivalent to 30% of forecast GDP in 2015 and 37% of bank deposits of Eurozone residents in Greek banks at end-May. For comparison, the Eurozone-wide stock equals 10% of GDP and 9% of bank deposits.

The ECB has accommodated a huge shift in Greek liquidity preference caused by the confidence-wrecking manoeuvres of the former finance minister and associated “Grexit” fears. His claim of deliberate “liquidity asphyxiation” is surreal.

Bank of England Governor Mark Carney last week suggested that the MPC will consider a rise in Bank rate “around the turn of this year” and that increases will “proceed slowly ... to a level in the medium term that is perhaps about half as high as historical averages". Mr Carney claimed that the pre-crisis average level of rates when inflation was at target was around 4.5%, so "half as high" implies a "new normal" of about 2.25%. Even if he is correct, however, the tendency of rates to cycle around their normal level suggests a future peak significantly above 2.25%. The latter level, moreover, seems implausibly low.

Mr Carney’s remarks have been widely interpreted as signalling a peak level of rates no higher than 2.25%. This assumes that the MPC will be able to calibrate the timing and extent of policy tightening in order to keep actual rates in line with their recovering normal level, consistent with achieving the inflation target. Such a perfect glide path is highly unlikely to be achieved given uncertainty about the correct inflation “model”, imperfect information about the current state of the economy and “shocks”. Historically, rates have rarely, if ever, increased slowly to an extended plateau. They have, instead, tended to rise abruptly to a short-lived peak. The process of reaching the normal or equilibrium level, in other words, has involved an overshoot and correction.

There were five rate rise episodes between the start of inflation targeting in October 1992 and the onset of the financial crisis in late 2007. Bank rate peaked at 6.625%, 7.5%, 6.0%, 4.75% and 5.75%. The mean deviation from Mr Carney’s estimate of a pre-crisis equilibrium rate of 4.5% was 1.625 percentage points (pp). If the “new normal” is 2.25%, therefore, it would be reasonable to expect rates to overshoot temporarily to 3.75-4.0%.

The minimum deviation from Mr Carney’s “old normal” was 0.25 pp in 2004 but rates were subsequently reduced by only 0.25 pp before embarking on another sustained rise to 5.75%. Even in the one case when rate rises stopped around the supposed equilibrium level, therefore, a substantial further increase ensued.

Mr Carney’s estimates of old and new normals of 4.5% and 2.25% respectively, meanwhile, seem low. He claims that 4.5% was the average level of rates during the pre-crisis period when inflation was at target. In fact, inflation was exactly equal to the central objective between the start of targeting and the onset of the financial crisis, while Bank rate averaged 5.4%, not 4.5%, over this period*.

Why should the new normal be 2.25 pp below the old? Mr Carney mentions headwinds to growth and inflation from global economic weakness, sterling appreciation and fiscal tightening. Such restraints, however, are temporary – they warrant proceeding slowly with rises but do not imply that the long-run equilibrium level of rates will be lower than in the past.

The main reason for believing that the new normal is significantly lower, unmentioned by Mr Carney, is a changed relationship between bank interest rates and Bank rate since the crisis. For example, the average bank interest rate vis-à-vis households** is now 2.0 pp above Bank rate versus an average premium of only 0.3 pp in the nine years to end-2007 – see chart. This suggests that a given Bank rate now has the same impact on household economic behaviour as a level 1.7 pp higher before the crisis.

Assuming that the spread between bank interest rates and Bank rate remains at its current level, an old normal of 4.5% may, therefore, imply a new normal of about 2.75% rather than Mr Carney’s 2.25%. If the old normal was 5.4%, however, the suggested new equilibrium is about 3.75%. These estimates, of course, will increase if the spread between bank interest rates and Bank rate continues to trend lower.

*The target was initially expressed in terms of RPIX inflation, with a central objective of 2.5%, but was switched to CPI inflation at end-2003, with a 2% goal. RPIX inflation averaged 2.5% over 1994-2003 while CPI inflation averaged 2.0% over 2004-07.
**Average of lending and deposit rates.