Want to know a big difference between Chinese and Indian GDP stats?

Well, we don’t think many Indian officials would say things like this, via China Daily:

Several local officials in China’s Northeast region sought to explain dramatic economic drops in their areas by admitting they had faked economic data in the past few years to show highgrowth when the real numbers were much lower, Xinhua News Agency reported on Friday.

“If the past data had not been inflated, the current growth figures would not show such a precipitous fall,” one official was quoted as saying.

The report cited several officials in the region who acknowledged they had significantlyoverstated data ranging from fiscal revenue and household income to GDP.

Irony thy name is fudged Chinese data.

And it’s not, we hasten to add, that Indian GDP stats don’t have their own problems — they do — it’s just that the lack of really strong centre out (and back in) incentives make us think the problem isn’t systemic so much as statistical. We pointed at the deflator before and SocGen said that “India’s Central Statistical Organisation (CSO) clearly has a lot more questions to answer.” Which seems fair.

But, while we wait for that, we thought this effort by Ambit — a Mumbai based brokerage — to construct an Indian version of the now famous Li Keqiang index was worth your time (h/t to Bloomberg for the nudge).

With particular emphasis on their estimates of GDP in the bottom par of the quote:

Against the backdrop of India’s own GDP data seeming unreliable, we update India’s own Keqiang Index (IKI) comprising of ‘real’ economy indicators which we first launched on 29 September 2015 [we'll put some notes on methodology in the usual place but suffice here to say it's comprised of ‘real’ economy indicators e.g. auto sales, cargo volumes handled, capital goods’ imports and power demand — we'll come back to that below too].

This Index (hence referred to as IKI) shows a strong positive correlation (+81%) with the old GDP series and a weaker positive correlation (+35%) with the new GDP series (see exhibits below)

More importantly the IKI points to the Indian economy losing momentum consistently from 2QFY15 onwards till 2QFY16 (see exhibits above and below).

In fact it is worth noting that if the new GDP series did not exist and hence we use our IKI to estimate the pace of growth in the real economy, our GDP estimate for 1QFY16 would be 6.3% YoY (vs the 7.0% printed by the new GDP series) and for 2QFY16 would be 6% YoY (vs the 7.4% printed by the new GDP series) (see exhibit below).

Which is pretty strong stuff, even if we would note that this Li Keqiang index also suffers from the flaws of the original — namely that it doesn’t pay attention to services.

What it made sense to track in the Liaoning province of 2007 — Li’s region back when he disparaged the official GDP stats and suggested the eponymous measure, which was then wikileaked into the world — as the economy was indeed largely driven by heavy industry, didn’t end up making sense for China as a whole. Here’s what Gavekal said previously about the Li Keqiang index as a gauge of China’s GDP:

The index is actually correlated only with the performance of heavy industry, and its relationship with the rest of economy and particularly consumer spending is quite weak…. As we have shown in previous research, electricity growth is basically a proxy for the volatile metals industry, rather than for the economy as a whole. And rail freight is actually closely related, since the main cargo of railways in China is coal, which is the main fuel burned by power plants to generate electricity.

We should remember that Li Keqiang used these indicators under very specific circumstances. In the Liaoning province of 2007, the economy was indeed largely driven by heavy industry, so it was very reasonable for him to look at these data. However, Liaoning was and is more reliant on heavy industry than the rest of the country.

It is also less so in the India of 2015 where, hell, manufacturing is desperately needed and services, IT for example, are the real bright spots in the economy even if they don’t employ the numbers that India needs to employ. We asked Ambit about this and here’s their reply, with our emphasis:

Firstly, when we model GDP growth in India – we find that historically Services sector growth is a function of Industrial sector growth. This makes sense since only when the industrial sector is buzzing with activity will there be demand for bank credit, transport services, hotels and other services. Whilst we have not directly taken into account services into our Keqiang, we know we are doing a good job of picking-up the pulse of the economy because IKI shows a significant and strong correlation with the old GDP series i.e. a series whose directional findings made more intuitive sense than that of the new GDP series.

Secondly, looking at the GDP from the expenditure side, private consumption expenditure and investments expenditure account for ~90% of the GDP in India even in the new GDP series. The four variables used in our IKI captures these two components of the GDP quite well. Specifically motor vehicle sales reflects both consumption and investment demand (as we use both passenger vehicles and commercial vehicles), power demand reflects both consumption and investment demand, cargo handled reflects investment demand and capital goods imports also reflects investment demand. Therefore from a demand side perspective too the index is a fair representation of the economic activities in the country.

Make of that what you will. Their basic point is that they are picking up the speed and direction of GDP via their measure, even if the absolute size of the growth rate remains unclear.

Which is fair even it kind of leaves us stuck on the fact that there is still very little certainty to be had in all of this — sorry — and that it is to be returned to. Not least because the enormous size of India’s informal economy is surely going to mask a bunch of services expenditure. But, hey, everyone knows GDP is a limited measure anyway, right?

It does seem though, that a bank/brokerage consensus on Indian GDP stats is being formed: India is growing quickly, and Modi’s BJP is making progress, but probably not as quickly as the headline numbers suggest. Basically, real time data, and the state of NPAs in the Indian banking system, should give pause for thought and no room for complacency.

As said, more in the usual place.

Copyright The Financial Times Limited 2024. All rights reserved.
Reuse this content (opens in new window) CommentsJump to comments section

Follow the topics in this article

Comments