Transcription
Hello friends, our friend Neelkanth Mishra ji is with us again today. He has a very good office. He is in the Axis Capital office itself. He is our old friend. He is an advisor to Narendra Modi ji. He is on the board of Axis Capital in the Economic Council. There are so many things about him, if I start, it will never end. So I will write that in the description. Neelkanth ji is with us for the third time. I would wish for him to come more. Sir, I will come six times, as many times as you say. We will do just that, and first of all, the fight between the rupee, dollar, and oil, and the internal conflict that is happening, which is difficult for India. Let's start from there. What is happening? See, a very big vulnerability that India has, and I think will remain for the next 100-101 years, is that we do not have access to dense energy. So in 2018, we wrote a report, I read a book by Vaclav Smil, "Energy and Civilization". After that, I felt a desire to know what is happening in India on this. Then I understood that our dependence on the import of dense energy is increasing. Absolutely. So, since then, whenever anyone asks me what is the most worrying issue for India for the future? I say it's energy dependency. The day energy prices surge, our economy will take a big hit. And that's exactly what happened. I don't think this is a very long-term risk. We will discuss that too. But for now, we can assume that if oil is at $X, then it's a headwind of about 2% of GDP. Meaning, it's a terms of trade shock. Basically, our calculation is that last year, oil was around $X per barrel. An increase of $Y per barrel has an impact of 1% of GDP. Yes. In this, we have also included gas, fertilizer, and edible oil. Uranium and metallurgical coal have very indirect linkages, so we are not going into that. So, if our economy was growing at 7-8% until February-March, which I think it was, because cars were selling, cement was selling, credit demand was good, footfalls were good, even FMCG finally achieved 6% growth. So the economy was doing well. We were at this stage, and suddenly a headwind of 2% of GDP came. So the airplane will slow down by 2% on the ground versus the ground. But there is a fiscal buffer in this because the government will not let fertilizer prices increase, no matter what. Oil prices have finally started to increase. So our estimate is that for every rupee increase in oil prices, the government's fiscal deficit is increasing by about 1.2% of GDP, which is about 40-45,000 crore every month. Right? So, 1.2% annualized. Meaning, if there is a 2% headwind, then the government was pushing back by 1.2%. So, the 7-8% became around 6.7-7.2% GDP growth. Now, when oil prices are being increased, for every ₹3 per liter, which is about 16,000 crore. Yes. So, 160 billion liters of petrol and diesel consumption in India. So, from ₹3, it becomes 48,000 to 5,000 crore. So, that's again around 0.1-0.12%. So, every time it increases by ₹3, you move from fiscal to growth. Right? But this is a one-time growth shock. Because once your terms of trade are adjusted, it means you are buying the same amount of oil, fertilizer, and gas, but you have to pay 2% of GDP more. Yes. Meaning, you don't have money to buy 2% of GDP worth of domestically produced goods. So, its production falls, just like GDP falls. But next year, because you have to grow at 7% on that base, you will come back to that 7% growth. So, this is a one-year growth headwind. Therefore, it is not a very worrying issue. And secondly, as we will discuss, if this war ends, and the Strait of Hormuz opens, if oil prices are going to fall, then I am not too worried about this. But the pressure on the currency is a bit worrying. Although today, May 25th, oil prices have fallen. It seems the Strait of Hormuz will indeed open. So perhaps that worry will also end. But for now, the currency market looks a bit panicked. And in currency, there is no near-term anchor, meaning no earnings, no book value, no cash flow. So, what happens is, whatever everyone is saying, happens. Yes. And currency traders, because most of the market is for 1-month or 3-month hedges, currency traders do not look beyond 1 month or 3 months. So, if this has happened, that the rupee has gone from 83 to 96. Absolutely. Then the economy will adjust significantly. Imports will become expensive, exports will become cheaper. Foreign investors should also lose this fear that the rupee will fall further. But the opposite happens: as the currency falls, people start getting scared, so they start hedging. So, there is a lot of evidence in the data that if you do accrual basis balance of payments, meaning, if you assume that the money for trade goods, exports, imports of goods and services, or FPI money, FDI money, arrived on the day it was announced. Right? Then from October to March, October 25th to March 26th, there should have been a shortfall of about $24 billion. That's called the balance of payment deficit. There was a demand of $24 billion more and supply was less. But during the same period, the RBI had to intervene by more than $70 billion. So, why was this happening? According to me, it was happening because importers who did not hedge before, are now hedging. Foreign portfolio investors, I was in Singapore a few days ago. There is such a recession regarding India that people were going into crowds to hedge. So, now, whoever hedges, effectively increases dollar demand. This increases pressure further. This causes the currency to fall further. So, this is basically a stampede. Well, this stampede will either stop in such a way that as oil prices have now fallen, perhaps the stampede will reduce, and then it's not a problem. Then, the rupee's demand and supply might suddenly reverse. Or, if this continues, suppose that even after today, the conflict continues and oil prices do not fall, then India might have to take some emergency measures. Because whenever there is a stampede, when crowd control is needed, then giving logic like "the rupee is undervalued" or this and that, doesn't work. In that, you have to show that we are strong, like "Look, for the next 3 years, we have secured an inflow of $100-150 billion. Now come and short it, show some courage." Those who are scared, we tell them, "Okay, now there is no need to short." And then the economy gets the time it needs to adjust. Because you see, if the rupee has fallen a lot, an exporter will not set up capacity here in a single day, nor will an importer be able to substitute. No, absolutely not. But in 6 months, 12 months, 18 months, that substitution and export response starts coming. So, once the rupee has fallen so much, it should be given some time to adjust. For that, perhaps some emergency measures might be needed. But look, if this war ends, perhaps even that will not be needed. So, this, but the war you are talking about ending, before the war, oil markets were almost at a peak. 105-108 million barrels per day were needed, and 115 million barrels per day supply was already starting to come. What do you think will happen if the war ends and Iran's oil also starts coming into the market, subject to sanctions being lifted? What equation forms? Before this, you see the war that is finally between the US and China, which you call the grand war, the great war. And this happens, this war happens for 20-40 years. And this war, this war happens on different dimensions, like political continuity, economic models, whether it's market capitalism or socialism with Chinese characteristics. This means the internal government throws a lot of money, competes internally, then two-three warriors emerge from it and go out and conquer all markets. So, will this model work, or will the US model work? But there is another dimension, which is the dimension of energy. So, a very suitable example in this is, when Genghis Khan wanted to defeat the Southern Song Dynasty, they just had to fight with sticks, meaning they went, were very strategic, and made many economic innovations, but they couldn't defeat them. Then his son, Ogedai Khan, tried, and he also failed. Finally, his grandson, Kublai Khan, finally defeated the Song Dynasty. To defeat them, he made such arrangements that first he adopted Chinese culture and customs, showed their people that "Look, we are not your great enemies, we respect you a lot." Second, and in that war, he lost, but strategically it was beneficial. So, Japan, because sulfur came to China from Japan, which they used to make explosives. So, he first said, let's stop that supply. So, he went to attack Japan. Now, there is a very famous story that a storm came and all their ships sank. So, the Japanese still believe that they were saved by God, and therefore God favors them, etc., etc. But after that, what happened was that the Japanese said, "Look, you tried hard, you tried twice to kill and defeat us. We said, okay, we accept your terms. We will not do anything to China." So, after that, you can understand that after 30-40 years of effort, the Mongols took over the Southern Song Dynasty. So, energy supply is a critical resource in the same way. Yes. Because without energy, you cannot reduce manufacturing costs, material costs, transportation costs, and now you cannot generate intelligence. Absolutely. What a point. So, now look at the energy mix. America gets 72% from oil and gas. Yes. And their grids have not seen investment for 40-50 years. And the rule of the American economy is that no one invests until prices rise. So, first there will be a shortage, prices will rise, people will see money, then investment will come. So, the government does not decide this there. Yes. So, it will take time. So, now, if you want to set up a data center, you are taking an old aircraft turbine, which they are already doing, and running it on gas to generate electricity. In how many cities is it now banned to set up data centers? Because electricity prices increase. So, reducing the cost of energy supply is very important for them. And they have gas and oil. Yes. So, they will keep gas and oil prices low. So, whether it's Venezuela, or Iran, or Abu Dhabi. So, people say that ADNOC is saying that they can produce 6 to 7 million barrels per day. They are currently producing 3. They have exited OPEC. Their known capacity is perhaps 4 to 4.7. But they are saying they can do 6 to 7. So, think about it, they were at 3, and they went to 7. Here, Iran was around 3-4, and it can also go up to 5, perhaps if they get the right investment, because they haven't received technology for 30-40 years, so they can also increase it. So, if our luck is good, oil prices can fall significantly. So, this is not a strategic problem. But in the near term, we have to manage the currency. Okay? So, I was reading a Goldman report, they had a pre-war estimate that oil would remain below $60. If 5 to 6 million more barrels come, perhaps it will even be $50, which is strategically very good for India in the long term. So, you are saying that this 6-9 month difficulty, or 3-month difficulty, is always a 3-month or more difficulty. Because what happened is, to support oil, as you saw initially, and we also said this, that if the war ended by April 15th, there would be great chaos, a deluge, but it didn't happen. Because China started releasing 3-4 million barrels per day from its inventory, and OECD and America also started releasing 3 million barrels per day. So, due to these inventories coming out, the total demand destruction was not the 12-13 million barrels per day that was being talked about, but 1.5 million barrels per day. Right? But now inventories are decreasing. Think about it, just yesterday in April, China released 105-110 million from its inventory of 1.2 billion barrels. They will have to restock. It's not like ships are empty and they can fill as much as they want. Because think of a ship as a pipe, that pipe has a limited capacity. So, people will try to fill it. So, I think it will take 8-12 months to stabilize. Meaning, this is not a 3-month problem. But it is certain that the oil curve, which is currently at a level, meaning, the futures for 1 month at $X and March '27 at $84. Correct. So, if we look at the previous few sentences, after the Strait of Hormuz opens, that entire curve comes down. So, perhaps the price of $84 for March '27 might even come down to $70-75. Then our problem will be more or less over. Let's come to our 30-year-old Brahmastra, which I call IT Services. Right? Our revenues were growing, employment was growing. There is a debate in the market about IT services employment, token usage, and revenues. What are your thoughts on all three? One is that this is a very fast-moving technology. So, whatever views you ask on this, I will give views. But these views keep changing. And when forming views on complex things, it is also very important to know your limitations. Like oil prices, I don't try to forecast oil prices even if I want to. I look at oil futures, that whatever it is, right or wrong, this is the current consensus. Similarly, in AI, everyone's expectations keep changing. And at the speed at which investments are going into this field, because of the AI boom that has happened in the last two-three years, and the hundreds of billions of dollars invested in it, what will be the impact on innovation and what will be the impact on AI, no one knows. Absolutely. Absolutely. For example, inference chips, because AI has many steps, but two big ones are training and inference. Now, what is happening is that no, specialized chips for inference will be more productive. With the same silicon, you can generate a lot of computation, consume very little power. So, efficiency chips of that kind will come in 2028-29, so perhaps computational costs will fall further. So, many things will change. So, there are many caveats in this. But as of now, you see, there are two big problems, in my opinion. Yes. One is that some people, some companies are making huge profits, and some people are incurring huge cash losses. Yes. So, look at Nvidia, Samsung, SK Hynix, TSMC, Broadcom. If you add up their free cash flows, it's $800-900 billion in free cash flows. Yes. And on the other hand, Anthropic, OpenAI, Google, Microsoft, Amazon, all these are investing heavily, giving them money. Yes. It's not a game of giving money back. Yes. So, in every funding round, they give some money to them, saying, "Keep the shop running." But, well, perhaps Anthropic has claimed that they have made profits. But in my opinion, at the level they are burning cash, the way they are investing, and the semiconductor shortage that is being talked about, due to which their money is being made. Yes. Like, if memory chips are not made, or GPU rental rates have increased by 100% in six months. Meaning, there is a shortage of compute. If OpenAI shuts down Sora, which is their image generation tool, because one, they were far behind the cutting edge, and second, it didn't have a monetization model. So, if you have a shortage of compute, then put money here, put money there. Yes. So, there is already rationalization of compute. People are already saying that Apple's strategy is edge computing, because a lot of compute also happens on the iPhone device. So, because server-side computing is constrained, we will compute on the edge, and those are also very powerful chips. So, when this is the situation, that $200 per month rate is not going to fall for some days. Okay? So, what we assume is that in tech, what happens is, what is $500 today, will be $200 after two years, and then $50. That is not happening for the next two-three years. So, this is a big observation. Okay? But there is a disclaimer that after three-four years, perhaps such innovations will come, or algorithms will come, that due to efficiency, it will start falling again. So, that's a discussion. I will interject here. I am asking Neelkanth ji this question also because he has covered these sectors, Taiwan, China, etc. So, he has very good knowledge of semiconductors. Well, we used to cover that, right? We remember. So, we meet. I know, I know. No, no, I also track it with a lot of passion. It's a very interesting sector. And second, you see now that in many places, there are problems of token overuse. Yes. That people's budgets have been exhausted in April itself. People said that a lot of AI slop is being generated because it will be easy to develop now. Earlier, what used to happen was, for example, if you wanted to develop software.
It used to cost a lot because now people will write code. Before that, it would take so much time, so much expense, what would be the benefit, what would be the loss? Now, as soon as the desire arises, you press a button, and the code is generated. What is happening now is that it is going into production, and people are finding it to be a very incompetent type of code because the efficiency that used to be is completely absent in it.
Let me tell you, for example, C++ is a high-level language. But computers are given instructions in assembly language, like moving data from this register to that register. Move it from here to here, and then the next instruction is to add these two registers. This is assembly. Ultimately, the computer understands only that. Right?
So, a revolution came around 1970, called the compiler. You write in a high-level language like "for i = 1 to 10 do this," and that compiler converts it into assembly language. Right? It does, meaning that system is still running. So, we took a course in IT called Advanced Computer Architecture. There, our project or assignment was to take compiler-generated assembly code and reduce it to one-third of the lines. This means that compilers were also very inefficient.
Indeed. Indeed. If you write it in assembly language, you might be able to do it in, say, 100 instructions. If you write it in C and compile it with a C compiler, which is a very efficient language, it will come out to 300 lines of code. It was inefficient anyway. Now, on top of that, these people are introducing such dangerous inefficiencies that a good coder would write it in 200 lines, but these people are writing it in 10,000 or 1,000 lines. So, obviously, the computational load will increase significantly. That also has a cost. So, these people are now struggling with this too, saying, "This is too much. Please control it a bit." It's not very beneficial to apply AI everywhere.
Secondly, you see that entry-level engineers in India are available for $300 per month. Even today, those with 5 years of experience are available for $500. So, if a good person, you see, there are many IT engineers in three places in the world: America, China, and India. Now, if I replace an American engineer earning $10,000 per month with an engineer earning $500 and code worth $200, there is still a benefit. Absolutely. Before, this was not possible because of domain expertise, and this and that. Now, you can do all of this. Absolutely. So, I don't think that the service industry here, in the product industry, the headcount will decrease everywhere. That is also justified. It is also justified because that is their core cost. If you are a software company, people are raising questions like, "Will you be able to survive or not?" So, if a software company is reducing its headcount by 30-40%, it is completely understandable because that is their job. That is their basic cost.
But if you are a retailing company, or a bank, or a manufacturing company, your IT spend is one or two percent of your revenue. Right. You can't do much in that. For example, if you are a car company in Europe and the US, you are currently facing an existential crisis about how to stop Chinese cars, how to stop Chinese cars, and how to change manufacturing, transition to electric vehicles, do autonomous driving. In this, changing my ERP will save me half a percent. That is not their priority at all. Right. So, this transition will be very slow. Okay.
Secondly, the trend of India moving things will continue because a lot of intelligence is now not just in coding but also in other areas. Now you can work from India because all the expertise is available in India. Because you can run that engine from anywhere. There will be a lot of change, and that's why I still worry about the stocks of IT service companies. Indeed. Their business models will change, but time will tell. But one change that might happen is that the trend of outsourcing will start to reverse. Right. Because 30-35 years ago, when Jack Welch started the outsourcing trend with GE, it led to an organization in society that the tech industry's expertise was needed for a project: sometimes a C++ programmer, sometimes a Pascal programmer, for UI/UX, for HTML, for databases. Yes. Now, as a manufacturing firm, if I have to keep all of them, keep them motivated, and give them work every year, it's impossible. Because I'll start a project, the software will be built, and it's done. The work finishes in two months. Where will I get jobs from? Correct. Yes.
So, for society, the most efficient organization was to keep them in a specialized technology firm, which you call TCS, Infosys, Wipro. Exactly. And they would provide you with shared resources as needed. Shared resources. And because those shared resources were very cheap in India, that's where the most growth happened. Hmm. And telecom networks increased, so data transfer etc., happened. Hmm.
Now, what will happen? Because two things happen. One is that the cost of coding has reduced significantly, and the cost of testing has reduced significantly. Now you don't need headcount. The second big change is that your end-to-end process has now shrunk. They call it sprint time in agile computing. What you used to build in three to six months, you are now building in a week. From specifying business requirements to deployment, you can do it in a week. So, in such a scenario, software deployment can become part of your business process. Hmm. For example, if a bank wants to launch a new product, a loan product for a particular geography. Right. Earlier, it used to be that its workflow needs to be changed, core banking needs to be changed, this needs to be done, that needs to be done. So, first, there would be discussions with vendors, then this, then upload this in six to eight months. Yes, that was a very long process. Now, it's like, "Okay, this project's architecture will be this, we'll do it this way, and the product will be ready in three days." So, now you will want to do it yourself. Do you need people internally or external help? Exactly.
So, what used to happen was, for instance, if you want to get the interior of this house done, the interior of this room. Now you need an electrician, a carpenter, someone to install tiles on the floor, an expert carpet installer, and so on. And an air conditioning expert. No. So, as a banker, if I need to find out who is a good carpenter, who is a good electrician, I don't know. So, I'll tell the interior designer, "Look, do this. Agree on the design. Here's the door, here's the window, and you build it and take the money." The same thing used to happen in software. Now, to build this system, I need Oracle, I need C++, I need HTML, I need JavaScript, and whatever else. Hmm. So, where do I get them from? Hmm. So, I'll hire a vendor. I'll say, "Look, okay, I need all this. Tell me how you'll build it." And they'll suggest something. I'll say, "Okay, now deliver it. I'll pay this much money. And you find the engineers yourself from somewhere."
Now, what if it happens that I just design it, and it gets built? Then why would I outsource? And I can change it every day if I want. So, now I will try to insource. Because now the cutting edge is business requirements. Because now I can translate my business requirements into software much faster. So, insourcing might increase. So, this will happen, or is happening, that global capability centers, where people feel, "Oh, we can do it in-house now." And if we do it in-house in India, it will be cheaper. Indeed. So, you will see that service exports are not slowing down. If IT service company exports have slowed down a bit, but overall service exports from India are not slowing down because this is increasing and that is decreasing. So, effectively, the trend of outsourcing is reversing a bit. Now, where this will go in three to five years is also a question, because IT service firms are also, for the first time, this is a rather big transition. For example, when I joined Infosys in 2000, it was the mainframe to dot transition, Y2K work was ending, and it was starting, and hundreds of learning and development courses were happening, everyone was being retained in large numbers. So, something similar will have to be done because now coding and all that, testing, all that will be done by systems or models. And the volume, both of business requirements and deployment, will increase. Just like in the industrial revolution, when weaving technology was automated, there was a shortage of yarn and tailors because a lot of cloth was being made. Yes. So, similarly, if your code development and testing are automated, then the demand for architects, designers, business requirements, business specialists, business analysts, and deployment engineers will increase because there will be so much code to deploy, which will go and which will not, what risks are involved. So, more people will be needed here, and there will be so many business decisions to make. Now you can implement them quickly. So, these technology solution providers, you see, they are not coders and testers. This is a misunderstanding people have. They are people who provide business solutions using technology. It's possible they might have to hire consultants, business consultants, or retrain people to become domain experts, or deployment experts, or data specialists. Some retraining will be necessary. And that retraining will take three to five years. The larger the organization, the more challenging it will be. But it is also true that their revenue loss will not be as extreme as thought. But one thing will definitely remain because of this uncertainty: how quickly will which firm be able to repurpose itself? How many can be retrained? And secondly, where will this organizational boundary of insourcing and outsourcing settle? It will take three to five years to discover this. Until then, there will be uncertainty. So, the PE multiple, 24-25, has been considered very expensive for a long time. Indeed. Indeed. But 15, 16, 17 is also not particularly cheap. Cognizant has come down to eight or nine. Yes. Correct. So, therefore, I think that due to this uncertainty, for some time, maybe a little growth and a little free cash flow yield, adding both of them, at whatever multiple it comes, they might settle at that multiple. But there doesn't seem to be much macroeconomic risk for India at this point. But just like with the energy vulnerability. Yes. Yes. That one day, they could just switch off access to large language models. So, this is a very big vulnerability. So, we need to be prepared for this and be suitably diversified so that we are not over-dependent on a single model or models from a single country. Indeed. You are talking a lot about China. One question I have is that we are doing a lot of internal analysis, not Rupee vs. Dollar, but Rupee vs. Chinese Currency. That depreciation has been even more. And many sectors will benefit from it. Chemicals, pharma, and so on. Tell us about China vs. America on this.
So, two things are happening. We feel that there used to be something called China Plus One. This should improve for India due to the currency difference between China and India. What is your opinion on this? Then I'll come to America. See, one thing is that after crises like this, people become more cautious. Yes. And what used to be there is efficiency, and there is resilience. Just like in our personal lives, one is that you push yourself to the maximum, get work done, and earn money. And the other is that if you do so much, you'll fall ill in two years. After that, you'll survive. So, resilience is about surviving for a few days, and then seeing what happens. Hmm. Similarly, in supply chains, the first shock came during COVID, and now the second shock has come due to this war. So, if someone is still getting 70-80% of their manufacturing done in China, and they remain there, it's foolishness. Rather, if this kind of war has started now, it happened in Venezuela, it happened in Iran, it could happen somewhere else. So, that needle that was moving from efficiency towards resilience will touch resilience a bit more. Indeed. Indeed. So, that argument that, "Oh, China is very efficient. Yes. It makes things very cheaply. It makes them very quickly." That's good. But if you get everything made in China, you'll get into a lot of trouble. Hmm. So, you have to make it elsewhere. So, I think, in my opinion, this will get another momentum, another push. But I think it will require some effort with the state governments to provide land quickly, get them permissions, and address their specific needs. For example, a semiconductor packaging company's equipment was coming to India, and the customs officer said, "Open it and show us." Oh, brother, it opens in a clean room. If you open it like this, it will become useless. "No, no, we won't let it go." Finally, people from the Ministry of Electronics had to fly down to explain, "Sir, this is a different thing. You haven't seen this before. But let it go. And if you want to see it, you can go to their plant and inspect it inside the clean room." So, so I'm saying that we have to do all these things because there will be new industries, so a lot of local effort will be needed. On this, you always say, I remember, "State capacity." Yes. China vs. India, which has been your big argument, that state capacity also needs to be increased. Correct. And it's good that I see people trying to work hard, but it requires effort. Secondly, a very big change needs to happen in our thinking that manufacturers from China can come. For example, if there are Chinese manufacturers. For instance, they have become the biggest FDI providers in Europe now. Absolutely. Because they understand that they are dumping their cars in Europe now. It's a matter of time before a big barrier comes there. So, before that, set up a factory. Set up a factory in Hungary, in Czechia, wherever they are setting them up. In Germany. So, why shouldn't we get them to set up here? In fact, I even asked this in a previous conversation. In childhood, we used to say, "Hindi-Chinese Bhai Bhai" (Hindi-Chinese Brothers). Now it has come to "Bhai" (Brother), well, maybe not, but at least get them to invest. Yes. Because one thing is that in diplomacy, they say "absence of binaries." Now, this is not a kindergarten relationship where it's either friend or enemy. Yes. So, even if there's a fight somewhere, they can also work together somewhere. So, you gave a very good example of Japan. Exactly. So, I think some progress is happening on this, so this is also beneficial. But a currency weakening does provide some benefit, but one thing should always be kept in mind. Yes. That the nominal exchange rate of the currency is obviously easy to track and you can see it daily, so people track it quite easily. But for economic competitiveness, the currency exchange rate that should be looked at is the real effective exchange rate. Because, for example, if someone asks me, "How is the Rupee doing?" I know they are talking about USD INR. But USD INR is one cross. For instance, we see that CNY INR is a different cross. So, when you take the currency exchange rate weighted by the trade share with all the trading partners of a country, that is called the nominal effective exchange rate, NEAR. But what happens is, for example, look at China. China's consumer price index has been flat for 5 years, actually flat for 6 years. Now, the consumer price index in the US and Europe has increased by 30%. Yes. So, if a US-based manufacturer, whose costs increase according to CPI, which is a fair assumption. So, even if USD CNY is flat, then it means China has become 30% cheaper. And it has become less competitive. So, therefore, if you look at the inflation-adjusted effective exchange rate, it is called the real effective exchange rate. Yes. China's REER has fallen by 20% in the last 3 years. Wow, interesting. So, people look at USD CNY and say, "Yes." It has appreciated by 2% against the dollar. Wow, wow, brother. Like-to-like things. It's down by 20%. Correct. So, all these things need to be considered. So, therefore, if our currency is depreciating against CNY, the Chinese Yuan, or against the Yuan, then to some extent, it is also necessary to counter the inflation differential. Okay.
Then there is a question that I feel that just as you hear stories about Canadian leaders, what happened in Europe, and now Modi ji has traveled extensively. Will "America Plus One" also emerge as a new concept, meaning that America is needed, but an alternative to America is also needed, which will be our alliance, our partner? And I don't think America will be very happy when people do all this, but they will also understand that they are fighting their own battles. Everyone has to fight their own battles. Now, for example, with hyperscalers. Yes. With hyperscalers, no matter what happens, if you put India's critical infrastructure on hyperscalers, like what happened with Nayara Energy, where Microsoft shut them down. Shut them down. Then in the US, there is the Patriot Act, the Cloud Act. So, at any time, anywhere, the government can say, "Bring the data." Yes. So, one, they will block access, and second, they will say, "Share the data." So, sovereignty is meaningless. The lights might go out while giving an interview, they might shut it down. Whatever happens. Yes. So, so you understand that all these things, as they say, the weaponization of interdependence is happening. Yes. Yes. Yes. Everything has been weaponized. If the data line is going from the Strait of Hormuz, the internet line is going, they will say, "Pay us." That used to be a $3 billion market for rare earths, which China weaponized. Yes. Now the US is weaponizing EDA tool licenses, which are chip design tools. Hmm. So, in such an environment, everyone will want to build a certain level of resilience, so as not to be overly dependent on any one country. So, if you think from India's perspective, to reach Developed India, we need four things. We need a lot of energy. Energy is needed. Energy is needed. Second, a lot of capital is needed because the current account deficit is effectively a savings-investment gap. So, you need capital to fill the savings-investment gap. Third, you need a lot of technology. For example, semiconductors. Even if the world has reached one or two nanometers, we have nothing. We have some five or seven. Yes. So, even five or seven, if 28 starts, that's good. But my belief is that if China started working on the first fabs around 1997-98-99, then we are about 25 years behind them in starting. If we spend 25 years to reach the current level, we will be far behind. So, if we want to accelerate, we will need partners who are willing to work on 5 to 7 nanometers, 12 nanometers in India, so that fabs can start in 5 to 7 years. Then our industry, we can also think about 3 nanometers, 2 nanometers in 10 years. Absolutely. But the start has to be made today. The start has to be made today. So, this is the third. And the fourth is global value chain linkage. Yes. For example, if you look at car sales, and this is a very scary chart. It's going like this. If you look at the global car market, remove China and remove America. Because America has imposed many barriers on Chinese car imports. So, one out of four cars is now Chinese. In the whole world, outside of the US and China, one in four cars sold is Chinese. If you are an auto component manufacturer, if you are not designing for Chinese OEMs, you have no business. If you look at it from the other side, like Nike, it buys 500 million pairs of shoes annually. And our Indian shoe manufacturers, who run factories, according to them, they say, "We have made 5 lakh, 10 lakh." Brother, for Nike, that's small. Nothing at all. Absolutely. Absolutely. So, if you want to manufacture shoes and get into Nike's valuation, you need at least 50 lakh to 1 crore. At least. Because if they need 50 crore, and you can't even make 1 crore, then why would they buy from you? So, there is no such capex, no capacity, no technology. So, what will happen is that you will have to get shoe companies, assemblers from Vietnam, Taiwan, to invest here. This value chain is not that you have the technology to make shoes. I went to a shoe factory. I asked the owner, "What's the difference between this and a cobbler?" Because they are moving their hands. They are moving their hands. A lady was just applying glue and pressing it. I said, "Sir, what is productive here?" No. The efficiency here is that she only applies glue. And one person only sticks the sole. One person does the stitching. So, it is very efficient. It's such a basic labor-intensive job. So, it can be done easily. It can be done easily. And it should be done because we have the cheapest labor. But it doesn't happen because those value chain linkages are not there. So, we need these four things. So, for energy, we need gas from West Asia, Russia, America, gas from Australia, oil. Capital again, West Asia has a lot. Canada, Australia's pension funds, they have a lot of money. China has a lot of money. They do $200 billion of outbound FDI. Their 10-year bond yield is 1.8%. The US 10-year bond yield is now 4.6%. So, so, and when people say, "We need to diversify from China," it is in Chinese interest. As we discussed, tell them to come and set up factories here. And they will come. So, I think all these things, when you do all this, you will understand that your geopolitical relationships become quite complex. Because now you are part of the US-led pack, which also includes Europeans, Japanese, and of course, Indians. In semiconductor technology, you cannot go near China. You won't. Correct. But you might have to bring China for capital inflow and value chain linkages. For battery technology, the biggest companies have failed in India, like Fisker, Cell. So, okay, it starts somewhere. So, R&D will have to be done. But you will have to bring technology and capability. So, therefore, I think that in geopolitical relationships, the multi-alignment framework that is now developing, the whole world will do it, not just us. Neelkanth ji, thank you. As usual, as usual, I learned a lot. It was a very good conversation. I will try to meet you every quarter and sit in your office and have tea. Thank you so much. Thank you. Save parents from falling. May that companion provide help. If there is love, tell them. Ensure their safety.