Markets Are Losing Steam but Other Assets Are Volatile
The big GST price cut announcements have come and it would appear that the auto sector could be the proportionately largest beneficiary
On Episode 671 of The Core Report, financial journalist Govindraj Ethiraj talks to Nikhil Varma, Technical Lead, India at Algorand, Sheetal Sapale, Vice President–Commercial at Pharmarack as well as Ganesan Ramachandran, Managing Director and Lead - Industry Groups at Accenture.
SHOW NOTES
(00:00) Stories of the Day
(01:00) Markets are losing steam but other assets are volatile.
(06:31) What the GST rate cut on life saving drugs really means.
(11:10) India’s new digital personal data protection bill can trigger key changes in handling of data and need for security.
(19:38) Agentic AI is a big opportunity. What do employers want from the future workforce?
NOTE: This transcript contains the host's monologue and includes interview transcripts by a machine. Human eyes have gone through the script but there might still be errors in some of the text, so please refer to the audio in case you need to clarify any part. If you want to get in touch regarding any feedback, you can drop us a message on [email protected].
—
Good morning, it's Monday, the 8th of September, and this is Govindraj Ethiraj broadcasting and streaming weekdays from Mumbai, India's financial capital.
Our top stories and themes for today,
The stock markets are losing steam, but assets like gold and currency are volatile.
What the goods and services tax rate on life-saving drugs really means.
India's new Digital Personal Data Protection Bill, or DPDP, can trigger key changes in the handling of data and the need for security and systems.
Agentic AI is a big opportunity, but what do employers want from a future workforce?
Volatile Markets
The big goods and services tax price cut announcements have come and it would appear that the auto sector could be the proportionately largest beneficiary, and at least that's how institutional investors are looking at GST impact and, of course, auto stocks.
Now, this was, of course, overall good for the consumer, though there will be some hitches and glitches in getting all of this going by September 22nd, the deadline. Small businesses could be hit in the short term for sure as they try and balance the higher costs they have paid for inventory versus the lower prices they will receive from customers, who in any case have paused most bigger ticket buying since August 15th, like cars or air conditioners and refrigerators. But last week was positive as a whole, thanks to GST.
For the week as a whole, the Sensex was up 901 points to close at 80,710 and the Nifty 50 was up 314 points to close at 24,741. This is for the week. That's all of last week.
Friday itself was mostly flat with the Sensex slightly down while the Nifty was slightly up one of those rare days. But the coming week will be driven more by external factors that appear including what's gonna happen or will happen in the United States. And the United States markets do bear mention for a moment if it's only to repeat what's been said in the past.
So specifically, last week saw the resolution of Google's antitrust case and that led to sharp rallies for both Alphabet and Apple. Broadcom shareholders welcomed a new $10 billion customer and Tesla stock was buoyed by a freshly proposed pay package for CEO Elon Musk, all according to CNBC. Of course, you have to hear this package to believe it.
Tesla's board has proposed a $1 trillion compensation package for CEO Elon Musk in what could be the largest corporate pay package in history. That's give or take 1 4th of India's GDP right now. Musk promised that he will convert Tesla into an AI and robotics powerhouse.
Meanwhile, a legal battle over his 2018 pay package then valued at, oh, just $56 billion continues according to Reuters. So the new award is roughly 18 times the size of the contested plan and is close to Tesla's current market valuation. So the US tech industry's $8 trillion companies gained something like $420 billion in market capitalisation last week and that took their total value to $21 trillion despite a fall in NVIDIA shares according to CNBC.
A senior index analyst at S&P Dow Jones Indices told CNBC that these companies now account for roughly 36% of the S&P 500, a proportion so great by historical standards that there are no comparisons. So while the US markets are zooming away, much of the rise is driven, as we said, by a few stocks and of course, one sector, you could say, and one phrase, AI. On the other hand, the internet, through which all that AI flows, is connected by physical cables and a reminder of that played out over the weekend when it was reported that undersea cable cuts in the Red Sea disrupted internet access in parts of Asia and the Middle East, though it wasn't clear what caused the accident.
There has been concern about the cables being targeted in a Red Sea campaign by Yemen's Houthi rebels, but they've denied attacking these lines in the past. Back home, while the stock markets may be somewhat directionless, the currency and gold markets are moving more sharply and thus reflecting the macro concerns that are weighing down the market. The rupee hit a record low and closed at Rs 88.27 against the dollar after touching an intraday low of Rs 88.38 on Friday.
So Rs 88.27 is where we are. Elsewhere, India's finance minister has said the government is keeping a good watch on exchange rates, stressing that besides rupees, several other currencies have depreciated against the US dollar. This is not the case only with rupee versus dollar, it's the case with many other currencies versus the dollar, so we're keeping a good watch on this, she told the Press Trust of India.
Actually, the rupee is amongst the worst-performing Asian currencies and has been hit harder, down about 3% against the dollar this year, which in itself has been somewhat weak, particularly after the tariff trauma started. In contrast, as of last week, the Thai baht, Singapore dollar, and Korean won have rallied more than 6%, and the offshore Chinese won has risen 3%, according to Reuters. So gold is the asset to watch, as we've been discussing here.
Prices are continuing to rise for reasons we've also discussed earlier, which range from overall economic uncertainty, that's global, and continued central bank buying as they switch out or reduce dollar dependence. As a medium to longer-term trend, gold prices appear to be telling us more about global economic mood than stocks. So gold's prices on Friday were very close to $3,600 per ounce, or just cents away, thanks to a weak US jobs data print, which further raised expectations of Federal Reserve rate cuts.
Spot gold was up at $3,596 per ounce on Friday, having hit a record $3,599. So the record is $3,599 per ounce. Gold is now on track for its strongest weekly gain in nearly four months, according to Reuters.
In India, gold futures, that's October futures, have risen about 6,500 rupees in the last one month, and have now hit a record of 1,07,807 per 10 grams on the multi-commodity exchange on Friday.
Life-Saving Drugs
Prices of some 33 life-saving drugs will come down with goods and services tax rates reduced from 12 to 5%. Now, these are drugs used to treat critical conditions like multiple myeloma, lung cancer, lymphoma, blood cancer, haemophilia, among others.
While the relief is welcome, the price at the top end runs into crores of rupees for one drug, as we will discuss shortly. On the other hand, these drugs are patent-protected, so it's useful to know why and how rate cuts in this space will play out, and also perhaps what the industry wants or maybe does not want. I reached out to Sheetal Sapali, Vice President at Pharmaceutical Industry Tracker Pharma RAC, and I began by asking her how she was seeing the price reductions and impact, apart from also getting a sense of industry performance for August or last month.
INTERVIEW TRANSCRIPT
Sheetal Sapale: Yeah, the GST slashing that has happened for the life-saving drugs, these drugs are typically used for some rare diseases or critical diseases like cancer or atrophy, muscular atrophy for children below two years or certain lung cancers, myelomas. So, these are diseases where the cost of therapy itself is very high and if you see at the cost of the drug for muscular atrophy for children, the cost of the drug is close to 18 crore with GST, it comes to 16 crore without GST. Certain anti-cancer injections, the cost is 2.24 lakh with GST, it comes down to close to 2 lakh. So, the drug still is a very expensive drug.
It's not that it becomes very affordable to the patient but there is some relief that the patient gets. Also, as far as access of these drugs to the patient, many times besides certain patient support programmes and schemes, there is a good amount of support that the companies give. So, it's not that the GST has made these drugs very cheap but that the GST has definitely helped in reducing the cost of disease burden to some extent.
Govindraj Ethiraj: Okay, and the selection of these drugs, the 33 that I'm looking at, so is there a clear rationale to this or could it have included something else or how are you seeing it?
Sheetal Sapale: Typically, when I try to study all these 33 drugs, all of them or majority of them are biologics under patent protection. So, they will be very expensive drugs and biosimilars for these drugs are not available right now. So, it is just the government that has tried to ensure that it becomes less accessible, more accessible to the patients because these are innovations, blockbusters and very nice therapies or solutions that are available to a disease.
Just because they are under patent protection, they're not easily available for the normal patient. And since it is patent protected, we will have to go by the price suggested by the innovator company. So, the innovator company puts a higher price.
Govindraj Ethiraj: The only lead that the government can give is in terms of GST. Right, and if there were to be future cost reductions or GST reductions, where would you think they should be or where do you feel they could be?
Sheetal Sapale: For the GST reductions, it should be in areas where the cost of therapy is high. See, right now, if you talk about certain cardio-diabetic conditions, most of the molecules have lost their patent. A lot of branded generics are available at a much cheaper price.
So, no need to have a GST reduction over there. But then certain sections of society where the solution should be easily available, I think there the GST reduction should help. But then mostly it is your critical diseases only.
Otherwise, everywhere we have branded generics, we have biosimilars, and we have genocide and generic generics. So, right now, there is not much need for a GST reduction as such. I mean, this is my opinion.
Other people may have different opinions. Because we have cheaper options or affordable options in most of the drug categories today.
Govindraj Ethiraj: Got it. Last question. So, the month of August has seen higher sales across categories, but the split in terms of regions is somewhat different this time.
Sheetal Sapale: That is because of the natural calamities. Maharashtra, Mumbai, Maharashtra, the entire state of Maharashtra, including Mumbai and Gujarat, have seen a lot of floods and a lot of rain-related calamities throughout the month of, you know, starting from the end of July and throughout the month of August. That's why there is some amount of dip.
Whereas the chaos in the North started somewhere in mid-August to the end of August. So, right now, in the month of August, there is not much dip. But then if the chaos continues, definitely there will be some down in sales in the northern states.
Govindraj Ethiraj: Right. Sheetal, thank you so much for joining us.
Sheetal Sapale: Yeah. Thank you so much.
India's new Data Protection Act
India's Digital Personal Data Protection Act, or DPDP-2023, came into effect in August 2023. The act provides a comprehensive legal framework for processing digital personal data, balancing individual privacy rights with the need for lawful data processing by both businesses and government entities.
It establishes a data protection board to enforce the law, sets out data fiduciaries, obligations, grants, data, principles, rights over their data, and includes provisions for penalties and exceptions, and there are also now Digital Personal Data Protection Rules 2025. But as the DPDPA moves from paper to reality, blockchain could help India Inc. operationalise it.
For instance, boards now need provable consent, real-time transparency, rapid breach response, and auditable data rights across vendor ecosystems. Web 2.0 stacks that were built for scale and not verifiability struggle with consent integrity, tamper evident logs, and cross-organisation policy enforcement. Dr. Nikhil Varma, Associate Professor of Management at Ramapur College of New Jersey and a blockchain expert told me.
According to him, blockchain can help India Inc. operationalise India's Personal Data Protection Act. I reached out to Dr. Varma, also a technical lead with blockchain company, Algorand Foundation, and I began by asking him how he was seeing the way forward with the new data protection rules.
INTERVIEW TRANSCRIPT
Nikhil Varma: One of the most important things that we need to note with the Digital Protection Act is the consent quality, and this quality is a pretty broad word. There are a lot of things that have changed within that perspective of consent quality. Consent has to be informed.
I mean, we all knew that, but then how does it kind of impact the business? Why is the business seeking that consent? That purpose has to be very clear.
It cannot be, I just got consent to be able to access this person's data. What is the purpose of the consent has to be now recorded. Now this, I think, impacts the consent life cycle.
In the past, there was no life cycle as such. This is just a one-off transaction to be able to get a consent, but now I am giving you consent to be able to access my records, to be able to do this specific thing, and after that, I will be able to withdraw my consent. The other thing which has to be recorded now is there has to be some kind of an audit trace for that, which means I have recorded the consent.
This is when the person gave me consent. This was the purpose of the consent, and this is when I gave back that consent to the person. So essentially, I have no more access to the data.
This creates even more complex scenarios when there is a breach of consent and agreements. So these are some top-level changes in that consent structure which have changed with the DPDP Act.
Govindraj Ethiraj: And who or what are the sectors that are likely to be affected most by this or could be affected most by this? Again, asking you as someone who specialises in blockchain.
Nikhil Varma: Looking at businesses, most of the businesses are data-driven. I mean, in fact, data is something that gives us competitive advantage. So every single business that touches customers in any form or the other is actually going to be impacted by this.
In many ways, I would say 99.9% of the businesses are going to be impacted out of this, except for businesses which are probably still working out of paper are the ones who would not have an issue with this, but I see the impact to be pretty huge.
Govindraj Ethiraj: Right. And when we say businesses must retain consistent logs and proof of action across apps and backups and vendors and data lake, are the enterprises ready for this sort of transition? Or how ready are they?
Nikhil Varma: The way enterprises had set up this whole construct is by digitisation. Digitisation is the first step to everything. If you're not digitised, this is not even a discussion.
But digitisation creates a lot of headaches. Now, how is digitisation currently set up? And from a blockchain perspective, I'll just get to it a bit later.
But what is currently set up with the Web2 architecture is that these organisations own the database, they own the transacting system, they might be using it as a service. But in many ways, that instance is theirs. It's their responsibility to be able to maintain the integrity of the data that cannot be handed off to a third party.
What is happening right now is that if I do a transaction in an organisation site, that transaction is going to be recorded in their database. If I suppose I give them consent, the consent is going to be recorded in their database, everything is going to be centralised and owned by the organisation. And I feel that that is where the current systems might not be apt to be able to cover all these different things, mostly when it comes to a breach and a grievance.
Because when it is a breach and a grievance, the trust factor goes down. A centralized system is as good as it is until something goes bad. Because data ownership is with the person who is being alleged.
And then it becomes a pretty complex process to be able to prove that nothing had gone wrong or if something had gone wrong, what had gone wrong.
Govindraj Ethiraj: Right. And the way forward, I guess, is tokenised trust. Tell us about that.
Nikhil Varma: So one of the things as Algorand Foundation in India that we are looking to do and we have done with a few entities is to be able to completely decouple the data. You know, essentially, I need operational intelligence, right? For operational intelligence, I need data.
And this whole construct of tokenised trust or what we've done is that we are decoupling the data and creating tokens. Essentially, if I have to transact with somebody who is over the age of 18, I don't need to know their date of birth. I just need to get to see if they are over the age of 18, essentially.
And if there is a token in their wallet, which has a verified data, which says, you know, this person is over the age of 18 and now you are able to do certain transactions with them or give them certain accesses, that reduces the issue of us having to deal with some kind of a breach of data. That being said, what tokenised trust is also building is that there is a back pointer to the data of the person, the actual data. Now, suppose we are getting into a banking transaction and that token is not enough for me to be able to kind of give that person a loan, but that token is enough for me to make a decision whether this person can possibly get the loan, right?
So if I am able to kind of get to that point where I'll say, okay, these are the candidates that could possibly get a loan, by the way, their identity is also concealed. There is only a token with which this bank is going to send them, hey, I'm going to send you this token. Would you like to get a loan from me?
And the person accepts only then, I mean, that is the consent event. And that's when the person will be able to share the data with the bank. So this mechanism kind of rewires the way we do things in Web2.
Web2 is essentially, in many ways, data intelligence is by getting data, data is power. Companies are selling data left, right and centre. And that's where all these problems are coming in.
With Web3, we will be able to solve this problem by not seeing data at all. And the token structure, trusted token that we are building has to be kind of robust, that architecture has to be robust, the information has to be more granular. Because if I give you a very high level token saying you are a citizen of India, that might not be able to do an address proof.
But if I say you are a resident of Mumbai, that might be address proof. But if I go too granular and get to your home address as a token, then that becomes a problem with privacy, right? So that token architecture is where we have to kind of sit down and see how operational intelligence can be drawn from this token.
But that is the whole idea behind the trusted token framework.
Govindraj Ethiraj: Right. Nikhil, thank you so much for joining me. And we're going to continue our conversation around how we're going to see enterprises gearing up for this post new world of digital personal data protection.
And we'll come back and discuss some specific use cases as well. Thank you so much for joining.
Nikhil Varma: Thank you.
—
Next week, Dr. Verma will speak to us on data providence and traceability, breach notifications, and access control and data sharing, all of which become important as companies use blockchain to operationalise data protection.
Agentic AI
Meanwhile, this is somewhat and has been a next-generation technology-heavy show, but that only tells you where the big changes are happening.
Consulting major Accenture is running the Accenture B-School Challenge 2025, and this is the ninth season. The finale is happening on the 11th and 12th of September, that's later this week, and listeners can catch the highlights on the course. We are media partners.
The team is Agentic AI, and I will be running a panel discussion on that team in Delhi later this week as well. Agentic AI is a big opportunity for India's youth workforce in the context of AI adoption and adaptation. But what are employers looking for and how should young college graduates align their approaches and their thinking? I spoke with Ganesan Ramachandran, Managing Director with Accenture Global Capability Network, leading industry consulting capabilities, and someone who's been involved with the B-School Challenge right from the beginning.
And I began by asking him how he was seeing the role of Agentic AI.
INTERVIEW TRANSCRIPT
Ganesan Ramachandran: It's a futuristic technology for sure. Companies always start having a good ERP, which means processes are in control. On top of that, they have data in control.
On top of that, you have an AI in control, playing AI. On top of that, generative AI. And on top of that is agentic.
So if you look at this continuum, a lot of companies are aiming for agentic when some things are not in order internally first. Notwithstanding that, the business cases, when it works, are very compelling value propositions. And that's what the future is all about.
So the trends are, how are we going to automate tasks which are now requiring a lot of human effort, which also may result in some errors, to completely ensure it runs in a smooth way at a very fraction of a cost, which will enable the companies to operate at a competitive level. And that's the trend now. But I still believe we are still a distance away.
Proof of concepts are helping, but you need to get these layers right. So from the company side, this is the issue. And from people who understand technology, another side of the issue which we need to talk about.
Govindraj Ethiraj: Okay. So if you were to look at industries that let's say you're familiar with, for example, automotive, where would you see the best use cases as things stand today? And given our understanding of how agentic AI can roll out.
Ganesan Ramachandran: Main low-hanging areas are B2C. When you have a call centre or a use case, that's the biggest use case. Let's take an example of let's say BMW or let's say Maruti.
Maruti has six dealers in Chennai. I am going to Maruti on the website and asking for a Grand Vitara, this model, this colour, and semi-automatic or automatic. And typically this goes to a CRM system.
They allocate to a particular dealer manually, and then the dealer may take some time and then come and talk to you. So it takes time. But now if you have an agentic AI in place, what it will do is understand your zip code, understand your product requirement.
And then what it will do is, there are multiple agents working together, which will first look at what is the availability of stock closest to your zip code. And then it will look at what is attractive prices, how much discounts are being offered by different dealers. And post that, it will come to you with a specific idea of this dealer, stock availability is now, the discount you will get is this much, and you need to travel this much distance.
This is what agentic AI is. Autonomous AI will ensure that even the dealer closes the deal with you, and if you're online ready, then even the whole process gets closed automatically, but we are still far away. But agentic AI will at least ensure the time lag between a customer expressing the interest and what he gets back when the lead is hot.
And that's going to be done very quickly with zero manual intervention and happens at the accurate level. And that's the power of agentic AI we're talking about.
Govindraj Ethiraj: Right. And you said that the base layer or the start point is ERP. And obviously mature companies, including large enterprises, would have that part already sorted out.
Let's say the gaps, if you want to call it, that really you were talking about is people. Now tell us about where and how does that manifest itself?
Ganesan Ramachandran: So I'm going to talk only about the Indian context in this call. So in the Indian context, we have 16% of the global AI workforce, which is a sizable number. But then they are in a particular skill set.
The skill set for agentic AI is more advanced than a simple AI. A. B.
The expectation of the requirements, which will be in 2027, is going to be so much. And we are not going to keep pace with it. So we are going to see while we are complaining about job loss and all that.
But on the other side, there is a huge opportunity waiting for us to tap into, which is how do we ensure that availability of skilled manpower is managed to ensure that the demand, which is going to come in the next two to three years, is addressed. And we need to do activities at multiple levels. But I'm saying the most important thing is how do companies do it?
There is an ecosystem of startups needed. Academia has to get involved. And all of them have to come together for a common cause to address this huge skill gap we are going to experience.
Govindraj Ethiraj: And if I were to ask you to dive a little deeper into the example that you gave, you know, an automotive dealership, which is trying to provide you as much information and intel as possible, which obviously helps with your buying decision and prepares you for your buying decision. How would you say or rather where is the sort of talent input that is needed? And where do you think the gaps would be in terms of the skill levels?
Ganesan Ramachandran: First is you need a LLM, which is large language models, which is trained to handle this query and able to understand what is coming from a prompt engineering, which is another one, which means what kind of prompts are coming, natural language, how do we communicate that to a proper machine level understanding. On top of that, agent creation and orchestration of them and the architecture to support that in terms of where it is happening. And the next layer is your ethical AI.
Are we getting any biases coming in any of this what we are doing? On top of that, IT security, we cannot expose data, which is important for a dealer and the dealer will be very, very cagey about exposing his data, which will make him non-competitive and he will blame the principle squarely for it. So there are multiple levels of intervention needed to ensure the whole ecosystem works in an ethical way, in a safe way, and it does what is intended to do quickly.
So there are multiple skills, technology skills needed to orchestrate this correctly.
Govindraj Ethiraj: Right. And in your understanding of the students that you see or your encounter or young employees that are coming into the system, what's your sense in terms of readiness?
Ganesan Ramachandran: I believe only five percent of engineers from tier one, tier two are ready to deploy AI resources. Five percent. That's also pathetically low.
But I'm not even talking tier three and other engineering colleges available in every pin code diamond desert. That's the first problem. Even MBAs, they come with a business running mindset.
Okay. So the intersection of business and technology, process and technology is not fully taught at the institute level. Most of the MBA institutes have it as electives, not as a compulsory course.
So when you are not understanding, when you understand business very well, you understand technology at a high level, then you're not able to marry both and bring the best of both the world for a specific purpose. And that's where the gap is, which means education institutions have to do a lot of things to do that. That is one side of the story.
Second, corporations or organisations have to invest a lot of money in re-skilling people who are likely to be affected and who have the aptitude to learn new things. So how are we going to do that upskilling to ensure that they are relevant to the growing demand coming in? So these are the two sides of the equation which need to be looked at.
Govindraj Ethiraj: Right. And you know, you're working on the Accenture B-School Challenge. So how do you see programmes like that feeding into this entire demand side challenge that you've just talked about?
Ganesan Ramachandran: So Accenture B-School Challenge, I'm the proud father. This is my ninth time running this. It was created by me nine years back.
So the idea originally we created was that Accenture is equal to strategy and consulting. But then we realised that every year we look at what is evolving as the latest things. The next two, three years it will be there.
And we weave in that topic so that we make students contextualise and aware the shape of things to come. And that is, so for this time, for example, is agentic AI. And the idea for them to realise, okay, you're learning business very well in your MBA.
You learned a bit of technology, but can you fathom the kind of impact agentic will have in the way we do business? I mean, just to give you an example, 10 years back, if I went to an MBA program in Ahmedabad, Bangalore, Kolkata, and told I'm a tech company, nobody would join the firm. But if you now don't tell them that you use technology, people will be wary of joining.
That's a world of change that's happened. The way we consume technology has changed. I mean, I remember standing in railway station queue for buying tickets or cinema theatre queue for buying tickets.
And now everything is in your mobile, just on your top. You just order things. It comes to you, blanket, 30 minutes service, 30 minutes pizza, five minutes thing.
Things have changed tremendously. So the intersection of business process and technology is completely redefining how we operate. So in that context, fresh MBAs have to understand how the world is evolving.
We may not be able to change the curriculum of an MBA institute very soon, but at least we can layer them with what is happening now so that they are more confident than what they were when they enter into the corporate world, whether it's Accenture or anywhere else.
Govindraj Ethiraj: That's a good note to end on Ganesan. Thank you so much for joining me.
Ganesan Ramachandran: You're welcome. Absolutely. Pleasure talking to you.
