Since the publication of our recent thematic report on the declining representativeness and relevancy of M2 statistics, we have got many client questions on the historical performance of M2 proxy as a leading indicator. Furthermore, broader questions have been asked concerning indicators of monetary aggregates other than M2 and their respective performance as leading indicators. In this note, we demonstrate in more details in regards to why Adjusted TSF and M2 proxy growth are superior leading indicators for growth and inflation.

Both “M2 Proxy” growth and Adjusted TSF growth are proven to be superior leading indicators for growth and inflation compared with M2 in the past decade. An effective indicator of monetary aggregates is expected to lead aggregate demand growth, which in turn precedes changes in inflation (in a relatively stable supply environment). As we have discussed in our previous research on M2 growth, in retrospect, M2 growth have “missed” a few important turning points in the economic cycles since 2005—including the high-inflation episode in 2007-08 and the economic reflation since 2Q2016. On the other hand, both M2 proxy and Adjusted TSF growth was more successful in forecasting both.

“M2 Proxy” and Adjusted TSF growth leads economic activity growth by around 2 quarters. Monetary expansion and credit cycle leads aggregate demand growth –output volume, represented by electricity production growth in this case, responds to changes in both M2 Proxy and Adjusted TSF growth with a lag of around 2 quarters. It is worth noting that we have chosen electricity production as the gauge for underlying activity growth here. On the other hand, the correlation between monetary aggregates and industrial production volume growth has weakened since 2016, potentially attributable to some data quality issues.

“M2 Proxy” and Adjusted TSF growth leads changes in overall inflation by around 4 quarters. Understandably, acceleration of aggregate demand growth leads to inflationary impulse with a few months’ lag. Therefore, monetary expansion leads inflation with a longer time lag compared with growth. It is worth noting that while changes in aggregate demand leads overall inflation, the inflationary impulse among various product groups are determined by the supply-demand balance of the each sector, i.e. their respective pricing power. Since industrial raw material demand tends to pick up earlier in the cycle, the lead-time of monetary aggregates for industrial raw material prices is usually a few months shorter than that for consumer products (baring significant supply side changes). Adjusted TSF and M2 proxy growth leads changes in raw material prices, such as rebar and cement, by around 3 quarters. However, the coefficients between monetary expansion and sectoral inflation are not stable – since monetary expansion leads demand growth, and supply side changes specific to a certain time period may mitigate or exacerbate the inflationary pressure. Take the recent run up of rebar and aluminum prices as example – supply side constraints and expected capacity cuts have send the prices notably higher than demand fundamentals dictate.

Consequently, the two monetary expansion indicators lead nominal output and earnings growth by around 3-4 quarters. While M2 growth failed to foretell the run-up of inflation in 2007-08 and the ongoing reflation started in 2H2016, both M2 Proxy and Adjusted TSF growth demonstrated superior performance in predicting these important “turning points” in the cycle. Furthermore, M2 proxy and Adjusted TSF growth can also be seen as effective leading indicators for corporate earnings growth, as the corporate profit cycles tend to coincide with that of nominal output growth.

We observe both M2 proxy and Adjusted TSF growth as leading indicators since their relative performance varies over time – while changes in FX assets do matter for aggregate demand growth, the coefficient between nominal growth and domestic credit appears higher than the one with changes in FX assets. Since M2 Proxy= Adjusted TSF + commercial bank FX positions – fiscal deposits, the main driver for the occasional divergence between M2 Proxy and Adjusted TSF growth is change in FX assets. Rapid FX asset accumulation contributed to around one-third of the M2 Proxy growth in 2003 – 08, while Adjusted TSF growth consistently underperformed M2 Proxy growth by a wide margin. Focusing solely on Adjusted TSF growth in period of rapid FX assets changes may lead to misjudgment of the growth and inflation impulse – e.g. Adjusted TSF growth in 2003-08 pointed to lower-than-realized nominal GDP growth. However, nominal output growth appears more responsive to changes in Adjusted TSF growth then FX assets, as Adjusted TSF exhibits a stronger correlation with (lagged) growth and inflation when changes in FX asset growth are less pronounced, e.g. in 2015-17.

Growth of both M2 proxy and Adjusted TSF has moderated since the start of monetary tapering in 4Q2016, but neither has slowed in an abrupt manner, or entered the “deflationary” territory. Furthermore, we expect growth of both M2 proxy and Adjusted TSF to rebound in July 2017, as domestic monetary conditions loosened on the margin and FX outflow slowed. The current trajectory of both M2 Proxy and Adjusted TSF growth points to robust nominal growth in 2H2017 -1H2018 and continued economic reflation. We will continue to rely on both the Adjusted TSF and M2 Proxy growth to gauge the changes in the “quantity” of broad money supply, as well as the effective leading indicators of aggregate demand and inflation.