Saturday, May 27, 2017

AI/Robotics/Automation - ITEmployees Cry Foul - Part1

Now that I have given my medium term outlook about Nifty, I will be spending the next few posts specifically addressing the sectors that are currently under pressure. There is so much being put out in different forms of media with some common keywords - automation, Artificial Intelligence (AI), Robotics, job losses.

For the common man, Robotics and Automation imply that tasks that can be performed by a machine with minimal human intervention. It is not as if robotics and automation came just yesterday. Automation started right from the time we had the industrial revolution! Two simple examples; imagine what happens when a Godrej storewell cupboard is being made (this is a common example I use in my lectures) We need sheet metal cut into different sizes according to where they will fit in the structure of the cupboard. Once upon a time, the entire cutting process was done by human beings. The challenge was that the productivity levels varied a lot and the quality levels varied a lot. Today we have a machine that cuts the sheet metal. Human intervention is needed to basically tell the machine - cut the sheet metal of this size. Another human intervention is to make sure that the supply of sheet metal is in place and once the cutting is done, take out the cut pieces. Advantage: What perhaps would take 100 human beings 100 hours to produce is now done by 1 machine in less than 10 hours. The variation in quality is extremely limited. Today there are millions of tasks in manufacturing and services where machines can easily replace human beings and perform a lot of tasks with minimal human intervention. Similarly in a financial services environment, a lot of accounts reconciliation can be automatically done by the system at the click of a button.

Anybody who says that automation is bad because we need to keep the jobs is akin to saying don't deploy machines to make the road or for that matter don't give the workers shovels. We need to keep the jobs so let them use spoons and ladles to build that road - lifetime employment. Is that what we really want? Yes I agree that there are some sensitive aspects where even if we have the option of an automation process, we perhaps should not allow automation because of security issues. A common example of this is drones for package deliveries. We know that it is in fact a fantastic innovation that looks great on television. My concern is that what do we do if some idiot misuses the technology and ships a bomb?? Air Traffic Controllers all over the world are already stressed with the tasks of managing airplanes and choppers. Imagine the scene when they have dots all over their screens because of these drones. Unless these security concerns are adequately addressed, perhaps we are better off keeping this innovation in the laboratory.

Artificial Intelligence is exactly what it says - the computer uses some sort of a program to perform analysis. We need to understand a key thing here - the system becomes intelligent only after some human being tells the computer what to look for and how to do the analysis. This is precisely the reason why the IT/ITeS industry in India is negatively impacted and job losses have begun. A lot of work that was earlier done by IT/ITeS employees by people are now being automatically done by the very computers and programs that have been created!

We need to understand one thing very clearly - India has grown significantly because of IT/ITeS companies but India has not contributed significantly to innovation in this very segment. All that India [and in turn the industry and people] has done is worked on the price advantage. Even for the basic computing tasks, if it costs USD 6000 / month in US [Monthly CTC for 1 employee] it costs USD 2000 / month in India. That is the reason why companies started laying off people in countries where salaries are high and gave those jobs to Indians. How many kinds of software innovations have been created by Indians in India? Google, Facebook, Microsoft etc etc etc were not India's contributions. Once that program was released, what Indians did well was fixing bugs and glitches. Now there are superior technologies that don't need so many people to fix these things.

We need to realize that automation is going to happen in many areas of manufacturing and services forever. What is important is that to get these automation techniques working, you need to be extremely smart and talented. The basic problem with a vast majority of Indians is that "smartness" and "talent" are rarely found. Yes thanks to an education legacy that the British left behind, we have a lot of English speaking graduates and post graduates but degrees do not imply "smartness"

So whether we like it or not, automation will continue and an employee is not smart enough to create and manage automation, s/he will eventually lose the job. Another question that then comes up is why do companies don't take initiatives to re-skill people. Human Psychology has shown that the greater the degree to which a person is used to think and do things a particular way, the greater is the challenge to make the person change. The old patterns and attitudes tend to be so deep-rooted that the unlearning process itself is a big challenge. And when you have so many young people joining the workforce - with fertile minds that can be easily trained, companies will take the easier option.

A lot of people complain that they are getting fired because they are not young enough, that is a wrong judgement. People are not getting fired because they are old but because they are not good enough for the new challenges. The people who stayed ahead of the curve and were talented enough are still being rewarded. And let us not forget that when the IT boom started, these very employees grew at the expense of employees elsewhere.

To be continued

Friday, May 26, 2017


Over the next few posts, I will be focusing on the gloom and doom part that has been hitting the newspapers, social media to the extent that there is some ITeS Union as well that has come in. I will talk about those aspects later. First, let us have the Nifty perspective in place.

I have seen a couple of bold comments like Nifty will double again from current levels over the next 5 to 7 years etc. foreign investors are still bullish on India etc etc etc. My take - we have been in a structural bull market since 2013 [some of my more experienced peers say from May 2009 - ok I buy that] Through Twitter, I had mentioned last week that a close below 9450 would be initial signs of weakness. My number is 9480 to be precise and I will be watching out for that this Friday as well. If the close happens to be below the 9450-9480 zone, it will be a confirmation of short-term weakness.

Another aspect we must not forget is that the dynamics of Nifty have changed significantly over the last 10 years. Earlier, Nifty was largely sensitive to Reliance, L&T, Tata Steel but that has changed. Some intelligent analysts renowned in social media have also been pointing out the same. Let us review the current Nifty 50 snapshot as of this week

One may review this file that can be opened with this link

I have taken the top 15 companies by relative weightage contribution to the index. With the telecom consolidation and volatile environment in IT/ITeS sector, some pain is on the cards. I have qualified the impact of recession on that particular stock. A careful look will tell you that the most severely affected firms are IT companies and banks. We need to understand one key thing; when we take the downward impact of IT/ITeS sector, the re-organization drives at telecom companies post Jio and the impending consolidation in the banks, the net impact in terms of affected persons would be at least 15 million [1.5 crore people!] What we many a time fail to realize is the multiplier effect. Every 100 direct jobs added in the IT or Telecom sector also added about 30 jobs indirectly.

That being said, I also disagree with the nay sayers who are predicting gloom and doom. There was so much panic and gloom in 2008 when the blood bath started and Nifty rapidly fell from 6357 to 2252 within a span of a year. From that point, Nifty scaled 6338 again, went all the way down to 4500 and hit 9k levels. From there it came all the way down to 6800 odd levels and now we are near record highs yet again. I do foresee a correction to about 6800-7200 levels once.

One key reason for my alarm bells on an impending correction is the market breadth. There is a divergence in index levels and overall market breadth. A few select stocks are taking the indices higher while a lot of stocks are actually going down. The distance from peak levels for a lot of stocks suggest a distribution pattern. The way mutual funds are advertising the SIP route with emotional appeal, my suspicion levels are inching upwards.And as I always maintain, corrections are good for the market and healthy too. I am particularly negative on the banking sector at the moment. The unsecured credit that is at risk at the market is near record highs. The aspirational young Indian is highly leveraged and by the time the restructuring exercise hits its peak in sensitive sectors, NPAs will be severely on the rise. For the first time perhaps in Indian banking sector, retail segment NPAs will outpace the business segment NPAs.

That being said, it will not be the end of the world. After dot-com, we had doom predictions; after Lehman Brothers, we had doom predictions. The same will happen again. The so called pundits and media spokespersons have selective amnesia and wrong anchor points. Even if we have a 25% correction from current levels, the markets will be way higher than they were in Jan'08 or Nov'10

In the next few posts, I will be discussing not markets but industry analysis and future prospects for people affected by organizational restructuring. Stay tuned and enjoy the ride