Wrong message from Financial Conditions

Goldman Sachs has come up with its new revamped G-10 Financial Conditions Index (sorry, no link is possible). The biggest problem is that it is pro-cyclical. A bit like the Credit Rating Agencies whose ratings are the best when the asset valuations are the highest and hence the risk is at the highest.

If the strength in the bond market and in the stock market is very high  – lofty valuations and tight spreads (credit spreads) – then the risk of a correction and financial conditions becoming tighter is also the highest at these levels. Therefore, as asset prices keep rising, financial conditions going forward come under increasing risk.

Financial conditions are the most favourable for growth when they are extremely tight. In other words, they cannot get any worse. They can only keep getting better. That is why 2009 was the best to invest and was also heralding the recovery (no matter how tepid it was).

This is a bit counter-intuitive. But, it needs to be grasped. Mainstream institutions have not grasped them because their incentives are not aligned to grasping this truth. Their business models would collapse, if they did so, perhaps. I do not know.

Unjustified Indiaphoria

A friend sent me a message this morning on WhatsApp:

“Rupee at 63 handle, Sensex over 30k, nifty over 9300! Too good to last? Methinks not!”

My response:

“Sorry, Sir. I am afraid so. Economic fundamentals do not justify them. I will be happy to short them all, if I can.”

That response was given as someone who was the Chief Investment Officer of a Wealth Manager and one who is a natural contrarian (with all its attendant risks and pitfalls) when it comes to investing.

In other words and in the interests of brevity, I belong to the school that believes in buying when no one has a good word to say on the market and sell when everyone is a cheerleader for a market. India, more or less, belongs to the second category.

In fact, its fundamentals are not great too. Its economic growth rate is exaggerated. The current real GDP growth is close to 6% or slightly lower. I had a blog post on it yesterday. Corporate earnings are improving but gradually.  On that, this is what I heard from a stock broker six weeks ago:

Despite an improvement in the economy after the demonetisation shock, the earnings downgrade cycle has continued. In the past month, consensus Nifty EPS for FY18 has seen 6% downgrade and that for the wider BSE100 has seen 5% downgrade. Commodity sector companies have seen the highest upgrades whereas the large downgrades were concentrated in sectors such as banks, telecom, and consumer discretionary. Consensus estimates still imply doubling of profit growth (ex-PSU banks and metals) to 15% in FY18, which looks optimistic to us, given limited scope for margin expansion.

Leather industry hubs in Uttar Pradesh have recently come under a cloud. Someone should visit them and check out the fallout of the ‘Cow Protection’ movement. Hotels are beginning to feel the fallout of the alcohol ban. See this article in FT (could be behind a paywall).

Then, the government’s orders on stents are backfiring. Pharma companies are fighting back. The government order on price controls and its rediscovery of the price controls as a public policy tool is rather unfortunate. My column in MINT yesterday was largely built around this. The consequences are not so much unexpected as they are unintended.

Reliance Jio has placed the financial health of many of its competitors under a question mark and the Reserve Bank of India has warned the banks of the risk of exposure to the telecom sector. Usually, public warnings mean that the situation is no longer a risk but a reality. It has asked banks to set aside higher provisioning.

E-Commerce start-ups are seeing big erosion in valuation and investors are marking them down in their portfolios. Further, there are other stories that sap investor morale and sentiment. In general, Indian PE/VC investing is a bit like Hotel California. You can check out but cannot leave.

Overall, bank credit growth to industry is contracting. Non-performing loans are holding back credit growth and the revival of capital formation in the country. This story is not an exaggeration. The extradition order on Vijay Mallya is a sideshow. Non-banking sources of finance are picking up market share, surely. But, they cannot be accessed by smaller firms. Not surprisingly, this article mentions that the International Monetary Fund, in its latest World Economic Outlook (April 2017), does not expect a big jump in India’s investment share of GDP.

Government’s tax and black money collection drives – laudable though they are as to purpose but condemnable as to process – are unlikely to help investment sentiment.

Notwithstanding (or, because of?) the Bharatiya Janata Party (BJP)’s political successes in elections including in Delhi, there is actually economic malaise in the country.

Financial markets and asset prices are largely a sideshow, supported by an equally unjustifiable and myopic global market sentiment. That is a separate story, however.

On robotic conclusions on robots

Following up on my earlier post on Branko Milanvoic’ blog post, I ran into the blog site called ‘Bank Underground’. It is a platform for Bank of England economists, analysts, et al, to post their comments on various issues that may or may not agree with the official policy of the Bank of England. Very good platform, in that sense.

I came across this somewhat optimistic post on the impact of robotisation. This one line, of course, does not do justice to the post which is, on the whole, quite thoughtful. It has many links. I clicked on one of them that took me to a FT Alphaville post of November 2015. In turn, it led me to a speech (‘Labour’s share) by Andrew Haldane in November 2015 and a piece by Martin Wolf in Feb. 2014. I read the latter and have downloaded the former.

Martin Wolf’s piece itself has many links. His final point is well made:

Above all, technology itself does not dictate the outcomes. Economic and political institutions do. If the ones we have do not give the results we want, we must change them. [Link]

But, that is true of a lot of things. Most prominent is money, for example. Or, may be, fast cars. They cannot kill or destroy character or whatever, on their own. It is supposedly up to us. But, that also assumes a lot of things about human abilities at self-control that years of research have shown us to be incapable of.

In his famous TED talk, Dan Ariely reminds us that we do not know our own priorities and that, when the decision becomes a trifle complex, we fall back on the default options chosen for us. That is what gave rise to the whole concept of ‘nudge’. But, ‘nudge’ can be both argued for and against.

At some level, ‘nudge’ is playing God. It assumes that people do not know their preferences correctly. But, that includes the people and authorities nudging too! Second, it can be, with some justification, deemed elitist and even manipulative. All, supposedly, for good reasons and for good causes. What if they are used for wrong ends?

So, giving us tools that we do not know how to handle is not a good thing. Therefore, some tools are better off being inside the tool box.

That reminded me of the quote attributed to Christopher Hitchens:

Everyone has a book inside them, which is exactly where it should, I think, in most cases, remain. [Link – btw, the link explores how the deceased writer Christopher Hitchens came to be associated with this pithy and witty observation]

I had left a comment on the Bank Underground website under the post on Robot Macroeconomics:

Just three minor points worth thinking about or keeping in mind:

(1) We do not know if jobs would really be lost or not. Assuming they are not, is it also possible to state that robotics would not come in the way of creation of jobs? In other words, what would be the counterfactual in terms of job creation but for robotics? Can we even guess that?

(2) In the developed world, robots could firm up resistance against immigration from developing economies that have a surplus of labour.

(3) It could also lead, through trade, to job losses in the developing world if robots enable developed world to regain competitiveness and they use their muscle to prise open markets for various goods and services in the developing world. The employment hit to developing economies will arise not only from lost export possibilities but also from having to compete with cheap imports, enabled by robots.

Lawrence Mishel’s blog post in the Economics Policy Institute (ht: Martin Wolf’s article) is hard hitting and rigorously argued. He rejects the argument that technology is responsible for inequality and instead points the finger at compensation in the financial sector and executive pay:

…. rising executive pay and the expansion of, and better pay in, the financial sector can account for two-thirds of increased incomes at the top. If superstar actors, musicians, and others were driving growth at the top we wouldn’t see the close association of top incomes with the trajectory of the stock market, especially during the last two crashes….

… On their specific claim about executive pay it is true that such pay grew as firm size grew in the last two decades but it is also true that firm size grew for many decades before that without any escalation of executive pay. [Link]

His detailed work that proves these points can be found here. Regular readers of this blog would know that I am very sympathetic to these points of view. Technological change may not be the direct cause of it but capitalists might find that a handy instrument to beat the workes with, heightening their sense of insecurity and curtailing their own wage demands. Agree that correlation is not causation but an adjunct role for technology may be appropriate.

I was quite tickled to read this observation by Martin Wolf that Lawrence Mishel rejects that technology changes caused wage inequality. I had read this morning an interesting article (ht: TCA Srinivasa Raghavan) as to how Ronald Fisher, the father of modern statistics, could not accept that smoking caused cancer. Randomized experiment was not possible in establishing causation from smoking to cancer. Hence, he could not accept that smoking caused cancer. Establishing causality in many social phenomenon is very hard. Too many other factors at play and very few of them can be controlled or allowed to remain unchanged. More often than not, temporal sequence passes off for causality.

Suffice to say that this topic is going to become very important in the coming years and hence, we need to keep reading and keep expanding of our knowledge of past experiences, patterns, etc., on the introduction of technology and the impact – both general and relative – on workers, wages and on the society. Facile conclusions and generalisations need to be avoided. More than a touch of humility is needed – to avoid arriving at conclusions but to keep minds open for understanding the phenomenon and to abandon one’s policies when evidence and trends change.

If past were the (only) prologue for the future, there would be no need for intelligence.

An interesting postscript to conclude this long blog post:

The legend of how a union leader told an official in a car manufacturing company who was proudly showing off his new robots to the workers’ leader that robots would not buy the cars that his company was making, has been retold many times with everyone adding their own bells and whistles. This link gives the full monty behind that legend. It is authentic but happened more than six decades ago. There was one real character involved – Walter Reuther – leader of the automobile workers’ union. But, his counterpart in the conversation was not Henry Ford II.



On Milanovic on robotics

Branko Milanovic has a blog post on robotics. He calls these three fallacies:
  • Labour displacement by robots – more short-term
  • Ability to predict our needs – they may expand making labour and robots co-exist
  • Limits to raw materials.
But, are they unreasonable, based on the trends in the last three decades?:
(1) Jobs may not be lost; what about jobs creation foregone? Wage growth? Is the trend of last thirty years not a warning of the consequences of capital displacing labour. If there is more of it, then is it wrong to fear consequences of robots on employment and/or wages?
Don’t these have social consequences as well?
(2) Is he confusing between needs and wants? Have our needs changed much? May be, our wants have? May be, marketing has created many wants which did not exist before and made us believe that they are needs. Has Milanovic read the ‘Joyless society’? The boredom of prosperity and affluence could become even more oppressive.
(3) Limits to raw materials – if we expand raw materials to include clean air, water, lakes and rivers, flora and fauna – our growth model has shown that it pushes against these limits sooner than we expect or reckon with. We are already feeling the impact. What if robotics pushes growth limits even further? What kind of growth would it create? Will we allow creative destruction to play a role? Or, will there be a greater demand for protecting those businesses displaced and threatened by the arrival of robotics? Then, what happens to productivity, economic growth and innovation in the overall economy?
(4) Lastly, will robotics help alleviate or complicate human psychological and behavioural limitations? Groupthink, herd mentality, elitism, cognitive dissonance have been amply on display in the response to the crisis of 2008 and in the US Presidential elections. Robotics or any technology can do nothing to help these. If anything, they can complicate them further.
(5) Humans do not understand risks that well. We neither anticipate them nor are we capable of assessing their magnitude and impact where we correctly anticipate them. Faced with unknowns, it is good to be on guard rather than be complacent. If the fears turn out to be misplaced, then adjusting to a positive surprise is a lot easier than the other way around.

US labour market health

In July, the US economy appears to have created a healthy number of jobs (+255,000). If one looked at the Current Establishment Survey (CES), the labour market appears healthy. The ‘Current Population Survey’ (CPS) paints a picture of slight deterioration in labour market health, at the margin. As per household survey, new jobs in the month of July were found only by those who either had no high school degree or had not gone to college. The unemployment rate for those who had not even completed high school dropped from 7.5% in June to 6.3% in July (seasonally adjusted). Unemployment rate for those with some college degree went up by .1% from 4.2% to 4.3%.

Among ethnicities, the unemployment rate for Asians went up by 0.3% to 3.8% in July. In the last several months (or, years) table A-7 is fodder for candidates like Bernie Sanders or Donald Trump. All labour market indicators for ‘foreign born’ are far healthier than that of ‘native born’. Given that multiple jobholders went up meaningfully by 164,000 (Table A-9), the total job creation (even from the household survey) of 420,000 is brought down to 256,000 in terms of new jobs created. Both the median and mean duration of unemployment had climbed steeply in July (Table A-12).

Table A-14 shows that the unemployment rates for workers in ‘Mining, Quarrying, Oil and Gas Extraction’, in ‘Durable Goods Manufacturing’ and ‘Information Technology’ are higher in July 2016 than they were a year ago.

From Table A-15, we note that the U-6 measure of unemployment rate had gone up marginally by 0.1% to 9.7% in July. It is still well below the rate of 10.4% in July 2015, however.

The Establishment survey paints a far healthier picture. The index of aggregate hours worked, index of aggregate payroll have expanded meaningfully in July.

‘Charting the labour market’ is a monthly chart pack put out by the Bureau of Labour Statistics. It is updated every month. For July, the chart pack was updated on August 5 and you can find it here. Charts 3, 4, 8, 12, 18 and 20 show the fragility of the labour market recovery compared to previous two recoveries which were not the strongest of the post-WWII recoveries either. BLS also publishes ‘Highlights of Current Employment Statistics’ which is a detailed Industry Employment Analysis. For July, the publication is here. Undoubtedly, the US has done better than Japan or Europe from their (non) recoveries from financial crises. But, clearly, the United States’ economic health is fragile and not sturdy.

Case (or not) for policy loosening in India

As we have come to expect from R. Jagannathan (alias Jaggi), he does not hesitate to question the policy orthodoxy or extant economic wisdom.

I am referring to this piece in Swarajya. I have had an email exchange with him on this. A debate is needed on this.

Personally, and in a theoretical sense, I am somewhat neutral to even sympathetic on this issue because there is empirical evidence to argue that

(a) Sustained inflation above 10% are bad but sporadic inflation bouts exceeding 10% do not cause lasting growth damage. Further, as Srinivas Thiruvadanthai (Levy Forecasting Institute in NY) would say, an inflation target of 7% would do no harm. There is no evidence that 4% is sacrosanct and a theoretically carefully chosen number. There is no magic to that number or the 2% number that developed countries have adopted. Until now, that is.

(b) There is a famous speech by the RBNZ governor last year (October 2015) in which he questioned whether central bank monetary policies influence inflation expectations, based on an empirical paper. His speech was based on this research:

I just finished reading the paper by Kaushik Basu (former CEA to the Government of india) written in August 2011. Worth a read. He also mentions the tenuous link between inflation rates and economic growth briefly.

Also, there are times when a higher interest rates improves supply of credit and leads to more credit being made available to the economy than the other way around.

Here is the link to Dr. Basu’s paper.

In the mid-1990s, Michael Bruno and William Easterly have written about inflation and growth, examining both cross-sectional and temporal correlations. The evidence is weak.

In other words, there is a need to make the case not loosely (pun intended) but rigorously, citing academic literature with empirical evidence. Foregoing is a sample.

Notwithstanding the above evidence that weakly (or, otherwise) support a case for monetary loosening in India, we need to take the Indian situation and context into account:

(1) Inflation and the poor and electoral and political noise with inflation. Does higher inflation work in India or does it hurt India? Is it also politically unrewarding for the ruling dispensation?

(2) Inflation expectations are already high  – close to 10% – will another does of a higher inflation propel it higher further? Salary increase demands and competitiveness considerations will follow.

(3) Whether we like it or not – I do not like it – financial markets matter. So, if a policy loosening is interpreted by the market badly, with implications for bond yields (higher) and a weaker rupee, then the purpose will be largely lost.

(4) On credit growth in India, it is a fact that private sector banks’ credit growth is running at 25.7%. See here. This is data through March 2016.

So, where is the evidence that higher interest rates are a deterrent on credit growth? The real issue seems to be PSU banks’ balance sheet and corporate balance sheets.

So, is the answer in lower interest rates or in expediting PSU bank reforms including recapitalisation, consolidation and even privatisation?

If corporate balance sheets are stressed, capital formation will be a casualty. It is inevitable. The economy has to wait it out. Is lowering interest rates the answer or will it boost only more personal loans (growing at 20%), higher consumption, higher
aggregate demand and hence, higher inflation, without helping capex revival?

For credit growth to different sectors, refer to the table 177, Handbook of Statistics (real-time). Data through May 2016 are available. Personal loans are growing at 20% annual rate.

(5) Mudra loans: They may not be the answer to corporate balance sheet stress and lack of loans to medium and large industrial borrowers. They have grown exponentially in the nine months up to March 2016. But, do they work or are banks simply re-labelling existing loans into Mudra loans? How is its economic impact measured? Without having a framework for measuring their economic impact in place, pushing these loans might amount to wastage of resources.

So, to sum up, if we wish to influence policy for the better, I stress the following:

(1) State hypothesis
(2) Marshal literature and evidence
(3) Keep Indian context in mind
(4) Beware of the law of unintended consequences (e.g., financial markets).

Do institutions matter?

Well, as titles go, this one is both provocative and stupid. Of course, they do matter. The subtext is whether they matter as much as we think they do. If they do, in what manner and over what time -frame do they affect economic outcomes? We do not have clear answers to these questions. It was a coincidence that Narayan Ramachandran wrote about the National Green Tribunal in his fortnightly ‘Visible hand’ column the day I had finished going through the paper I am discussing here. As one would expect, Narayan had cited the work Daron Acemoglu in his column. Pl. note that I am using Narayan’s column as a peg for this post. This post itself is not a criticism of his column.

The paper I read was by Professor Nicholas Ziebarth of Northwestern University (Year 2011). The paper is titled, ‘Are China and India backward? Evidence from the 19th Century of U.S. Census of Manufactures’.

Professor Ziebarth compares the present-day India and China to that of the United States of the 19th century since, on many counts, they share similar characteristics, including that of per capita income, etc. If anything, the US had a far better set of institutions:

We started with a completely dierent country in terms of policy but a similar level of development in the form of the 19th century U.S. and ended up with very similar levels of misallocation. This casts doubt on the relation between policies and the measures of distortions suggested by HK. Whereas the policy distortions in India and China are well known in terms of state owned enterprises and previously the License Raj in India, it is not so easy to what those explicit barriers are in the U.S. at this time.

As I argued above, economic institutions were rather conducive to allocating resources. Government spending was limited. There were few market restrictions. Bank decisions were not overseen directly by the government. And where the government did play a larger role in developing communication and transportation networks, its eect appears to have been salutary.

Simply put, there is no good policy reason why the U.S. at this time should have such a distorted capital allocation. The only thing that it shares with China and India of the late 20th century is a similar level of economic development as re ected in real GDP per capita. I interpret these results as suggesting that part of the natural process of development is capital reallocation. Policy surely does play some role here, but I think it is much less than we think while development itself is the main driver. What this means is that HK succeeded along the accounting dimension while overstating the case for the institutional one. Rather than a cause of low income per capita, capital misallocation seems to be an effect.

In fact, we can simply accept the superiority of the U.S. in the 19th century in relation to the present-day India and China for this reason, even if for no other reason. Yet, it had a similar level of distorted capital allocation as these two have now.

Professor Ziebarth cites Atack and Bateman (1999) here:

No single early economic data source surpasses the nineteenth-century U.S. federal census manuscripts in quality, in consistency, or in comprehensiveness; from mid-century onward, the census enumerations oer a unique historical record detailing the transformation of the United States from an agricultural to an industrial economy.

How one wishes for better quality data from China and India?! Both lack that but for different reasons, of course.

His concluding paragraph suggests that we do not know remotely enough about the process of development:

This paper contributes to both the historical literature on the development of the American economy and more broadly to economists’ thinking on development. The results suggest that a sizable fraction of subsequent manufacturing TFP growth in the 20th century can be tied to a better allocation of resources. For the latter, this paper shows that the mapping from economic policy to the allocation of capital is not straightforward. It appears instead that part of the process of development itself is a natural reallocation of capital towards more ecient ends, totally independent of policy decisions. And this is endogenous process is now what needs a theory. Future work should attempt to ll in the gap between the results presented here for 19th century America and those in HK for the 20th century. In the end, much remains to be learned about development through the lens of economic history.

That must make economists very humble. So, we should add the question of ‘what makes economies tick?’ to the question of (as yet unresolved questions) of ‘what makes companies tick ?’ and well, ‘what makes movies tick?’

[P.S: What Ziebarth refers to as ‘HK’ is the paper by Hsieh and Klenow published in 2009: ‘Misallocation and manufacturing TFP in China and India. Quarterly Journal of Economics 124, 1403-1448.]