Bazaar & Beyond: Investing Beliefs, Global Diversification and Market Narratives

Vivek Ananth:  Hello everyone and welcome to another episode of Bazaar & Beyond.  Today, we have with us Devina Mehra, Founder, Chairperson and Managing Director at First Global.  She has spent more than three decades in capital markets, with deep experience across research, investment strategy and fund management. She is also the author of Money, Myths and Mantras, a book that challenges many commonly accepted beliefs around investing, asset allocation, diversification and long-term wealth creation.  In this episode, we will talk about the investing beliefs that retail investors often follow. Whether it is reducing equity exposure after retirement, investing only in Indian markets, or blindly following popular market legends. 

We will also understand how a disciplined investor should look at portfolio construction, risk management and diversification in today’s market environment, where global uncertainty, valuations, earnings growth and changing sector leadership are all influencing investor decisions.  Most importantly, we will understand why investing is not only about learning new ideas, but also about unlearning some old assumptions.  Thank you, Devina, for joining us on Bazaar & Beyond and for giving us your time.  I want to start with one of the points I came across during my research. You have mentioned before that we should move away from the theory that as we approach retirement, our equity allocation should also automatically go down.  You have also said that we may need to rethink how much equity allocation should be reduced, considering life expectancy and how equities perform over a long period of time.  Could you shed some light on that? Because it goes against conventional wisdom. 

Devina Mehra on Retirement and Equity Allocation 

Devina Mehra:  First of all, I think two pieces of financial advice that are often given by experts and professionals are completely off when it comes to asset allocation.  One is that you should be 100% in equity when you are young. The other is that you should have zero equity when you are retiring.  Both are incorrect.  You should never be 100% in equity. Sometimes people send me messages saying, “I am 25 and I have a good job. Should I be 100% in equity?”  The answer is no. Equity is not predictable. You should never have anything in equity that you may need within the next six, seven or eight years.  Even if you are in a good job, you may lose your job. You may want to study further. You may want to make a down payment on a house. There may be a medical emergency.  So, a certain portion should always be outside equity, in more predictable assets.  But now, people also think of retirement as something close to the end of life. If you look at the data, the average Indian at the age of 60 has a life expectancy of around 19 to 20 years. And that is the average Indian.  For the kind of people watching this podcast, the number is likely to be much higher.  So, your retired life can be 30 or 35 years. It can be as long as your career. That is the part very few people consciously think about.  It also means that your money has to last for 30 or 35 years. If you put all of it into 100% safe assets, you are probably not even going to beat inflation on a post-tax basis.  Therefore, a reasonable portion of your retirement corpus should be in equity.  The caveat, again, is that it should not be money you are planning to withdraw within the next five to seven years.  What happens in equity is that the order of returns matters. We know equity may go up in one year and go down in another year. If the down years come early, and suppose you have put 100% in equity, your ₹100 may go down to ₹80 or ₹85. But you still need to withdraw ₹5 the next year.  Now, that withdrawal comes directly out of your capital. So, the money that was supposed to last for 30 years may run out in 20 years.  That is what you want to avoid. Still, a reasonable portion should be in equity because over the long term, equity does give you higher compounding.  And of course, I always say that the big missing part in most portfolios, including retirees’ portfolios, is global investing. But I suppose we will come to that later. 

Why Global Diversification Matters 

Vivek Ananth:  Actually, that is my next question.  When I was reading your book, you spoke about how the Asian crisis, and what happened to the Asian Tigers, convinced you that you have to offer global equity to clients.  Right now, we are in the same discussion again. If you see Twitter or social media videos, a lot of people are talking about global investing. You have built your business around providing global expertise to your clients.  I want to understand how a lay investor should approach this. There are certain limitations in terms of how much one can invest because of RBI rules, whether it is through mutual funds or direct equity.  Direct equity can happen through LRS, and RBI rules specify how much one can invest. For mutual funds, there is also an industry-level overseas investment cap.  So, how should a lay investor think about this exposure? You have been working extensively in this area, and you have also built expertise in investing in global stocks. In your book, you have written extensively about the research expertise required for this.  How does an investor get there in terms of the exposure you are talking about? 

Devina Mehra:  First of all, I will tell you why you should be global.  Suddenly, everybody has discovered global investing simply because India has not done well in the last year and a half.  One piece of data I always give is that when I started working, the dollar was ₹12. So, in less than a career span, we have seen around 90% depreciation.  When you are planning for financial goals that are 10, 20 or 30 years away, you cannot forget that the rupee depreciates. That is the basic thing.  Then, of course, you mentioned the Asian crisis.  This is the book, Money, Myths and Mantras. That was one major lesson because in all those markets, the stock markets fell and the currencies also fell.  In dollar terms, all those markets fell between 50% and 90%. That was a wake-up call for me.  Indonesia fell 90%. During my Citibank days, I had spent a few months in Indonesia setting up the Citicorp Securities subsidiary. So I could really imagine that if I had been Indonesian, 90% of my net worth would have been gone.  That was one part of it.  The other part was that I wanted to learn more, and I was bored of doing the same old Indian stocks. But the main lesson was: do not put all your eggs in one basket, because this is what can happen in a crisis.  And remember, those were not basket-case economies. Those were the highest-growth economies of the 1990s. So, that was the wake-up call. You cannot have all your eggs in one basket.  Nobody had dreamt of going global at that stage. We got the RBI approval, set up the company subsidiary and then the bank, and we became not just the first Indian, but the first Asian members of the London Stock Exchange. 

First Global was the first Asian member of the London Stock Exchange, other than the Japanese. That was in 1999. The US broker-dealer took another year or year and a half.  So, we have been following global stocks since then.  And it scares me, the kind of people I see talking about global investing today. Many of them actually do not have a clue.  Global companies are more complex. Understanding them is not easy. The list of companies itself changes. If I go back to the early 2000s, when we had started covering US stocks, the tech majors were IBM, Dell, Motorola and Cisco. People forget that at one point, Cisco was the highest market-cap company in the whole world. Then it fell 86%. 

So, even in the current AI boom, you have to remember that this has been the history. You have to be careful when going overseas. Unfortunately, a lot of people offering global products in the Indian market today do not actually have the expertise.  I am in an AMC market, so I understand that. But overall, in India, and perhaps elsewhere too, the asset management industry often becomes an asset-gathering industry.  The reason a product is offered is that today assets can be gathered under that head, not necessarily because it is the best thing to do at today’s price, or because the manager has the expertise to do it. If I look at all the GIFT City products, most of them are down over the last six months, even though global equity markets have done well. Over the last one year, nobody is even close to the benchmark. 

Our multi-asset global product has done far better than these equity-only products. And a multi-asset product means it has lower-yielding fixed income and other assets too. Still, it has done better because you cannot do this without expertise.  I have posted a caricature before, and this is not just for global markets. It is of somebody driving a car while looking backwards. That is what many people do. They think, “This is easy. I have set up a global product. I know the names of ten US stocks. I know the names of four Asian stocks. I will invest in those.” That is what causes the lag. The story of the Magnificent Seven in the US is a good example. They drove the S&P almost entirely in 2023 and 2024. More than 50% of the S&P move came from just those stocks. 

But in 2025, that bull market had already started to tire. Five of those stocks underperformed the S&P. Only Google and Nvidia did well. This year, they have completely lagged. Now, the Russell 2000, which represents US mid and small caps, has done much better than the S&P. The Magnificent Seven have contributed only around 6% of the S&P movement.  Now the theme has passed to semiconductors.  Again, people think semiconductors are easy to analyse. I saw someone saying, “Look at the P/E to growth for semiconductors, it looks so good.” I said that you have no idea how this industry works.  We have followed these stocks for 25 years. It is a highly cyclical industry.  Semiconductors and semiconductor equipment are highly capital-intensive and highly cyclical. This is both because they are capital-intensive and because they are supplying someone else’s capital expenditure. 

It is too deep to go into just now, but basically, if you do not understand the entry cases, you should be careful.  So, how should Indian investors do it?  Of course, this has been a cause for me. I have not started talking about it in the last six months or one year. I have been talking about it for years.  We have also been offering a product that starts at a few thousand dollars, and it covers all geographies and all assets.  Going global does not mean just buying the US. That is another myth. People think the US is enough diversification. No.  There are long periods when the US underperforms. Just because that has not happened much in the last 10 or 12 years, barring 2022, does not mean the US never underperforms.  So, you need all assets and all geographies.  Our products were, say, $10,000 not long ago. Earlier, that translated to around ₹7 lakh, and now it is around ₹9 lakh.  If you simply want to buy a US index, I would not necessarily recommend it. If at all you want to buy an index, buy a global index. Those ETFs are available.  You can open a broking account and buy it yourself. There is no reason to pay somebody a fee just for holding an ETF on your behalf.  For example, ACWI is the All Country World Index. Something like that works. Or even if you want to buy a US index, buy it yourself. There is no point having an intermediary for that. 

I saw one GIFT City product from one AMC which is an index product. If you want to buy that, you can buy that. But otherwise, be very careful about whom you are trusting with your money.  They should have expertise and some accountability.  That is the other issue with fund-of-funds structures. If it is being managed elsewhere, the fund manager here has no accountability because they are not managing it. They are just adding another layer of cost.  So, if you want to buy a US mutual fund or any other global mutual fund or ETF, you can open a broking account and buy it yourself. 

Semiconductors, AI and Indian IT 

Vivek Ananth:  You mentioned the semiconductor theme being cyclical.  In India, we have also seen that the whole AI boom has impacted a lot of IT companies here. A couple of weeks ago, if I am not wrong, over the weekend the whole Twitter sphere was talking about how Indian IT companies never innovated. A lot was written, and then there was pushback from IT sector veterans and past leaders as well.  Where do you see this cyclicality impacting Indian markets in terms of the valuation compression that has happened?  How do you see this playing out? Over the last six or seven months, we have seen this impact. You have said your base case is that you are positive on India, so how do you see this playing out? 

Devina Mehra: Semiconductor cyclicality is a different thing.  Let me explain why that happens. If you are selling toothpaste and you sell ₹500 crore of toothpaste this year, next year you are not going to suddenly go down to ₹100 crore. It will be plus or minus from ₹500 crore.  Whereas if you are supplying something that is capital expenditure for someone else, you really start from zero every year.  If you look at this year, AI expenditure is crazy. There are all kinds of estimates, from $800 billion to $1 trillion.  So, not only semiconductors, but even Dell, IBM and other hardware suppliers have come out of nowhere. You have demand and you have pricing power.  But two years later, will AI capex still be $800 billion? That is the trillion-dollar question. 

Two years ago, the total capex of the same companies used to be of the order of $50 billion. That is why it is highly cyclical, apart from the fact that it itself is capital-intensive. Any capital-intensive industry is cyclical because capacity comes out in batches. Whether it is hotels, cement or steel, that is why these sectors are cyclical. Coming to IT, I have not started saying this now. This has been my bugbear with Indian IT companies for 20 years. They had tons of cash. They had trained people. But they did not try to move up the value chain. In contrast, if you look at China, not just in IT but in anything, they started with mass manufacturing and then moved up to technology. That is one part of the story. Having said that, I have not been saying that this is the end of IT services. If you look at their obituaries, they have been written many times. The earliest was Y2K. People said all the business was Y2K, and once Y2K was over, these companies would be left with no business. Then came cloud transition, digital and SaaS, which is software as a service. Every time, people said these companies would be finished. But within their sphere, they have been able to pivot their business. I have seen that time and again. In this case also, remember, AI is not something where the CTO, or chief technology officer, of any large enterprise will just hand over the keys to AI. I had to write it very briefly in my last column in Mint. There is a paper called Agents of Chaos. It talks about giving rights to these models and how they created chaos, from wiping out entire systems to databases and other failures. There are so many points of failure and security breaches. So even if an enterprise goes towards AI, it will need somebody to intermediate that. But I am more concerned from the macro point of view.  For the last 25 years, the big driver of India’s employment has been IT. That may not be the case, at least for a few years. With most technologies, when they first come in, everybody says they will kill jobs. And they do in the first round. But eventually, jobs come back, often in greater numbers. 

Look at banking in the 1980s. There were so many strikes saying that if banking was computerised, everybody would lose their jobs. Today, banking employs far more people than it did then, at least in India. That is the story of every technology, right from the time spinning and weaving were mechanised. Ultimately, it increases productivity and increases opportunities. But for now, hiring and the number of people employed are more of a macro concern. For the last 25 years, there has been direct employment through IT and IT-enabled services. Plus, there is a multiplier effect. For every person working, there is a driver, food delivery, real estate and so on. That is more of a concern for me. But I do not think IT services are going to zero. NASSCOM, on the employment side, thinks there may be a decline and then a sharp increase. 

Dangerous Investing Assumptions 

Vivek Ananth:  I want to change track a little bit. Investors often have certain beliefs that may be contrary to their goals. You also write about this in your book.  What are some dangerous assumptions you would like to point out to our audience? These may feel okay today, but in the future they could actually harm the portfolio.  For example, one idea that sounds very good is “buy and forget”, that you should simply hold on. 

Devina Mehra:  You should hold on to your overall equity allocation, but do not hold on to the same stocks forever. 

To give you an example, the original Sensex list had a lot of paper, textile and shipping companies. There were companies like Premier Automobiles and Hindustan Motors, which used to make the old Fiat and Ambassador cars.  In terms of business groups, there were the Mafatlal, Thapar and Scindia groups. Many of those companies have gone nowhere.  You hear stories like someone’s grandparents bought Hindustan Lever and it is worth so much today, or someone’s parents bought HDFC Bank. But you forget all the failures.  At that time, there were many other banks that got licences and then went bust. Global Trust Bank, Times Bank, Yes Bank, Centurion Bank and many public sector banks went through major problems.  That is how your mind tricks you.  It is not that you have to hold the same stocks forever. The Sensex stocks at that time were the blue chips. It is not as if something was wrong with the index formation. All of them had long histories. They had been around for decades. These were all very well-established business groups. 

Yet, that was the trajectory.  Barring Indian Hotels, which was a relatively new company at that time, everything else was made up of very old companies. So, that is one way your mind trips you up.  Another thing is that we tend to hold on to our losers. We say, “Let it come back to my price and then I will get out.” The market has absolutely no interest in what price you bought it at. So, get out of whatever is not making sense.  I sometimes say that in your DP statement, the last page is like your exes. You do not want to remember that you ever loved them. So, go to the last page of your DP statement, gather some courage and get rid of all the junk. Put it somewhere sensible. Have a proper risk management system. Have stop-losses. These are things people overlook because investing is a loser’s game. You have to play it not for the sixes, but to avoid getting out. Do not take a big hit on your capital. That has to be your first driving force. 

Home Country Bias and Asset Allocation 

Vivek Ananth: Specifically with respect to Indian investors, many are over-invested in the Indian market. That home country bias is built into most portfolios. Do you think this is also something people are not very aware of? 

Devina Mehra: I started with that. In asset allocation, the biggest missing thing is not having enough global stocks. Asset allocation does not mean just asset allocation within India. That is the biggest hole. 

Vivek Ananth: I am circling back to that because when I speak to many people, they think investing means SIPs. That is also the success of the AMC industry and AMFI. But somehow, asset allocation still seems to go over people’s heads. I think the industry also has a lot of work to do in explaining asset allocation. 

In your book, you also talk about how people only think of large cap and small cap. They do not look beyond that in terms of assets. I think you have also spoken about commodities and gold. 

Devina Mehra: Yes. Take gold, for example. There were some fairly large AMC heads in 2024 who were saying, “Gold is lying useless in your locker. Sell it and put it in equity.”  When gold went up, the same people started offering gold funds.  Then in January 2026, inflows into gold funds in India exceeded equity inflows for the first time, I think. That was the peak of the cycle. I had written about it at that time.  Last year, everybody thought I was silly for talking about the risk in gold. Again, the long-term chart of gold shows that in dollar terms, it has always been a more volatile asset than equities. If you go back, gold made a high in 1980. That high was not crossed for 27 years. The next time that 1980 price came back was in 2007.  In 1999, it was still down 60% from that peak. And that was not an aberration. The next peak also saw a 40% fall. That is the nature of the asset.  Coming back to why Indians do not think about global investing, part of the reason is that when we were growing up, there was no option to invest abroad. 

Your mindset is set from there. When I started working, and well after that, even when you were travelling abroad, you could take only $500, and that too once in three years. If you travelled a second time within three years, you had to leave the country with only $20 in your pocket. Your credit card said, “Valid only in India and Nepal.” So you were travelling with no money. Our grandmothers were not wrong in buying gold because, at that time, it was the only hard currency asset they had access to. 

Vivek Ananth: So, you are saying foreign exchange exposure also came through gold. 

Devina Mehra: Exactly. That was the only hard currency asset available. That is why the rupee chart of gold looks okay. It is essentially a rupee depreciation chart. Even though the dollar price of gold fell 60% over 19 years, the rupee price of gold went up because the rupee depreciated. That is the whole thing. But now, that access exists. The RBI’s Liberalised Remittance Scheme has been around for almost 20 years. Still, it has not really seeped in that this option is available and one should look at it seriously. It should also be significant. There is no point putting 2% or 5%. That is not good enough. 

Why Investors Should Not Blindly Follow Market Gurus 

Vivek Ananth: One thing I read in your book was more specific to Warren Buffett and his performance, but I also want to apply that to India.  A lot of us, and I have also been a journalist before, used to write about what a certain investor is buying. There are also many platforms that track portfolios in real time and give alerts.  You have specifically called out that all the gurus do not always outperform. There are many old adages that people acquire and then try to invest through. You have also spoken in the book about moving away from the biases created by these legends. You mentioned Peter Lynch as well. 

Could you contextualise that for the Indian audience? 

A lot of retail investors get an alert on an app saying that a certain person has bought a certain company, and they go and buy it. It builds up a kind of frenzy. I wanted to pick your brain on this. 

Devina Mehra:  Why you should not follow famous investors is one of the longer chapters in my book.  Some chapters are shorter, and I told the publisher that I wanted to vary the length depending on how much it takes to explain something. There are many levels to this.  One is Warren Buffett. It is not that he is a bad investor. A lot of the fundamentals that you read from his initial writings are useful.  But often, the investor is not doing what you think they are doing. Everybody thinks Warren Buffett made his money in one particular way. For that matter, his guru Benjamin Graham, who wrote the book on value investing that people keep on their desks, also made most of his fortune in an insurance company, not in the kind of value companies he talks about. Similarly, despite what Warren Buffett says about derivatives, he has made a lot of money on derivatives. These were structured deals and derivatives of that kind. Also, in the last 20 years, he has not really outperformed. Often, a strategy may work at a particular time in a particular market, but it does not always work. You mentioned Peter Lynch. His idea was to buy the stocks of companies whose products you like. If you like a toothpaste, or as he mentioned, if his wife liked a particular product, he would go and buy that company. But that worked because that particular profile of companies worked well in the US markets in the 1980s and part of the 1990s. The real superpower of Peter Lynch was that he managed the Magellan Fund of Fidelity, which was the best-performing mutual fund ever in history. I think it compounded at around 29%, though there is another story to that. His real superpower was recognising that this performance and this strategy were not replicable. 

So, he retired in just 14 years. He retired in his mid-40s. Who retires in their mid-40s? But he realised that this was an aberration and that he would not be able to replicate it. That is what I give him credit for. When you are talking about all these alerts, first of all, they may be entirely false. The other thing is that you do not know the entire thing that investor is doing. You do not know where this one transaction fits into their overall portfolio. You do not know what hedges they have put in. You do not know any of those things. So, that is a fool’s errand. Even if you are looking at a strategy level, we often think: “So-and-so has made so much money. What was the strategy they followed? I should follow that.” That is the wrong question in the first place. For example, suppose there is a very high-risk strategy where every year 90% of people go bankrupt, but the people who are left make 10 times their money. You play this game five times. 99% of people have gone bankrupt, but one person has made out like a bandit. Now you say, “This person started with ₹1 crore and now has ₹50,000 crore. So this is the strategy to follow.” But no. There is a 99% probability that you will go bankrupt. People used to ask Rakesh Jhunjhunwala also: “You made your first money in trading. Why do you tell people not to trade?” He said, “I smoke, but I tell my children not to smoke.” That is because it is not a sensible thing to do. 

Circle of Competence and Evaluating Advisers 

Vivek Ananth: Interesting. I also liked the point you made in the book about the circle of competence. Whether it is a fund manager, adviser or someone who is going through an advised investing experience, how should they evaluate whether their adviser is competent? You touched upon that briefly earlier. Could you expand on that concept? 

Devina Mehra: Actually, what I have written about in the book is slightly different. The context for circle of competence is one of the myths. People say, “Invest only in your circle of competence.” That is okay if you are a retail investor doing a little bit of individual stocks here and there, along with your mutual funds. 

But what does circle of competence mean? It is often a euphemism for comfort zone. As a professional fund manager, you cannot say, “I invest only in my circle of competence.” If you only understand banks, FMCG or consumer goods, what are you going to do? Themes always change in markets. What will you do when metals, defence, industrials or something else starts doing well? Will you say, “I do not understand that”? That was the context in the book on circle of competence. What you are asking is how one should evaluate a financial adviser. Again, you have to look at what their track record has been and what they have been saying. The problem is that there is frankly a lot of mis-selling in financial services at all levels. A retiree goes to a bank, and the bank relationship manager tries to sell them an insurance product which is totally unsuitable for them. On thematic funds, we have tracked this since the mutual fund industry started. We looked at when various themes were launched. When were IT funds launched? When were pharma funds launched? When were small-cap, defence or PSU funds launched? 

We found that they are almost timed to perfection. They come near the peak of that theme. After that, the trend often goes down. On social media, someone sent me a message saying, “Ma’am, you keep talking against thematic funds. Look at this technology fund.” I will not name the fund house. It was ₹2 in 2003, and look at how much it has compounded over the years. I said, “Very good. It was ₹2 in 2003. Now look at when it was launched.” It was launched, timed to perfection, one week before the global tech crash. In two and a half years, it went from ₹10 to ₹2. That is the part you are missing. From the time it was launched, it was down 80% in two and a half years. Who would have lived through that 80% fall and survived to see the compounding after that? Unfortunately, this is how things are presented. First of all, be financially literate. I think this should be taught in schools. People should know how to calculate internal rate of return and CAGR. If somebody is selling you an insurance product, at least you should be able to calculate that if you are paying this amount and this much is going as the person’s cut in the beginning, what is the amount actually being invested? 

Take ULIPs. If somebody had calculated the rate of return, they would know that it is a very poor rate of return. It just sounds good that you put in this much and after 20 years you get this many lakhs or crores. You should at least be able to calculate that. Also, make sure you know what is in your portfolio. Typically, anybody who has been investing in the markets for a few years may have 25 mutual fund schemes in their portfolio. Whichever scheme or fund house was giving the incentive at a particular time, every year their adviser added three more schemes to the portfolio. 

AI, IPOs and Corporate Governance 

Vivek Ananth: Interesting. I have read a lot of what you have written in the recent past, and one thing that caught my eye is that you have also written about the AI boom and this space IPO. Just to clarify for our viewers, there is no recommendation on any company. This is only to understand the distinction. You have analysed a thread where you discussed what may work and what may not work. It seems like the IPO has come at a time when interest has peaked. The company also appears to be an amalgamation of two different businesses, but it is still getting a top-dollar valuation. Do you think this indicates a top in AI, in terms of the race to build and then hope demand will come? That seems to be the narrative everyone is seeing. 

Devina Mehra: It is not even “build and they will come”. It is more like “raise money and they will come”. The article I had written was not even about the business and valuation as much as the fact that there are no safeguards. There is no corporate governance possible, because what Elon Musk has done is that he has taken investors’ money and given them no shareholder rights. The shareholder rights are all with him. He can change the board as he wants. He can pay himself what he wants. He can even merge a private company of his with this company at any valuation, and shareholders have no recourse to question that. That is peak stupidity. It is like giving him the divine right to rule. In the past also, he has merged businesses. Going forward, he can do that again. The name may be space-related, but if you actually go through the prospectus, not much of it is to do with space. Then he is talking about how their skill is to identify new trillion-dollar opportunities. If you read the history of financial bubbles, this is always the peak of a bubble. If you look at the South Sea Bubble, today people talk about how there was this company that raised money saying it would put it into a great enterprise, but nobody knew what the enterprise was. And people gave them money. I am 100% sure, and I have said this publicly and posted this too, that some future edition of a book on financial bubbles will have this as a chapter. 

Not only is the valuation crazy, people are giving money with no safeguards and no basic corporate governance possible. In Delaware, Tesla shareholders could sue him for certain things. That is why some structures were changed. There used to be the concept that the emperor had the divine right to rule. That is what you have given to Elon Musk. This is not going to end well. I have written quite a few articles on this. 

AI Narratives and Investment Risk 

Vivek Ananth: It seems like many AI firms are lining up IPOs towards the end of this year and beyond. 

Devina Mehra: Yes. I just wrote a column yesterday, and it came out in Mint. It was on this. All the stories and narratives you hear about AI, realise one thing: there is no independent verification. They are all coming from interested parties, from companies, their managements and their investors. I saw Vinod Khosla some time back saying, “Would it not be a wonderful world where everybody has access to medical diagnosis and prescriptions at zero cost?” What zero cost? You are putting in a trillion dollars. Who is going to pay for it? Already, enterprises are saying AI is not proving to be economical. The Uber CEO has said that they have tried it, but it is too expensive. Others are saying similar things. People are looking for open source. They are looking at Chinese models. And there are many risks. Your data and everything else can go out. These AI firms have no scruples. Recently, it was found that Anthropic had signed an agreement saying they would retain no data, and then it was found that the data was still there for 30 days. There was also something circulating about someone asking AI about a tender and how they should strategise, and the AI pulled out emails from their competitor where the CEO and team were discussing the same tender. That is how crazy it is. Plus, now you have unknowable risks, like the US government saying that access to certain models has to be stopped for all non-Americans. If you look at the history of new technologies, a large number of technologies never do what they promise to do. Look at Facebook changing its name to Meta, saying the metaverse would be the new world. That entire thing did not happen. There are many such examples. 3D television. Self-driving cars have been theoretically around for 20 years. Even what succeeds takes time. Even genuinely transformational technologies, like railroads, automobiles, aviation and the internet, have all been graveyards for investors. There is always overvaluation and overinvestment. The difference with AI is that if a railroad company went bust after building a railroad, the railroad still existed. Somebody could buy it at 20 cents to the dollar and make it work. But AI assets are very high-obsolescence assets. If they do not work, they may just be junk. There may be no value left. Take internet infrastructure after the dot-com boom. People said the internet would transform life, and it did. Then they said, let us invest in internet infrastructure companies. Those undersea cables on which the internet runs have been around for a long time. A large part of that was laid by a company called Global Crossing, which went bust more than 20 years ago. The cables are still in use. But in AI, even the expenditure being put in may not remain useful. And that is only the economic part. Look at the environmental cost. India is happy that data centres are coming here. But in the US, people are protesting against data centres because of power and water shortages. Data centres use huge amounts of fresh water. India is already short of water. In many places, the monsoon has not even fully arrived. Arizona is a desert state, and you are putting mega data centres there. People are now protesting because they realise the consequences. You go to some small community in South America and deplete their water table. That is what happens. 

Can AI Sustain Its Current Growth Expectations? 

Vivek Ananth: Do you think the AI theme, specifically the way it is destroying value in other pockets, like IT services here, will slow down because less capital becomes available to companies raising billions and hundreds of billions of dollars? Or do you think the inherent complexity and scaling challenges of that business will put a stop to the accelerated growth that is being projected? 

Devina Mehra: I am not saying AI will not be a transformational technology or that it will not succeed. That is not the point. The question is: will all this investment ever make returns on capital? Will it ever make enough money to justify this kind of investment? Certain things will definitely be disintermediated. I think coding is one area. I was talking to a small software company and they said entry-level coding is clearly much faster and much easier with AI. But it is not 100% perfect. It still needs to be checked. So, that is definitely one use area. But as I said, any new technology also increases productivity. It creates opportunities and aspects that you have not thought of today. When Instagram started, it was supposed to be a static photo-sharing site. What it later became was not even envisaged by its creators. So, how this evolves and what opportunities come from it is not always predictable. Once things become easier, you start doing more things with them. You do things that you would not have done if the technology had not been available. So, it may be transformational. But in any technology, the question is: who will succeed? In AI, even within this short period, first it was ChatGPT, then Gemini, then Claude, then something else. It changes every day. 

I remember what Chuck Prince, then CEO of Citigroup, said just before the mortgage crisis in the US, which later became the global financial crisis. He was essentially saying that risks are building up, but till the music plays, you have to get up and dance. As a CEO, if there is growth, even if you know it is higher-risk growth, and you do not show that growth for three or four quarters, you are out of a job. So, you have to keep playing the game. In tech today, it is the same. Companies feel they cannot afford to miss out on this bus, so they keep investing just to remain in the game. But that does not mean everybody’s capital expenditure will ever make returns. 

Policy, Manufacturing and the Indian Economy 

Vivek Ananth: For the last few questions, I wanted to take a more 30,000-foot perspective in terms of fiscal policy and how monetary policy is evolving. After the global financial crisis, the global regime was easy money. During the pandemic, it became easy money again. Now that we are out of it, it seems like constrained fiscal and monetary impulses may continue in some way. This could have an impact on certain industries. Earlier, a lot of foreign investors used to come and buy Indian stocks. That is one of the reasons why FPI and FII flows have also gone down. How do you think the world economy is placed right now? The appetite for debt-fuelled growth has reduced considerably after the Covid pandemic. It seems like capital allocation is happening only in certain sectors now, because that is where demand is. I want to understand where we are in terms of monetary authorities having the capability and willingness to rescue us from another crisis. A lot of themes seem to be overbought right now. And there is also not a lot of fiscal space to build. For example, India wants to build semiconductor plants here. We want to build the next set of manufacturing plants here. PLI has been seen as a success, and somehow we have been able to achieve some degree of success in terms of local value addition. Purely from an Indian economy standpoint, where do you see this? 

Devina Mehra: How was PLI a success? Let us see. I read an article today. If you look at manufacturing as a percentage of GDP, we were at about 17.5% for around six years in a row after the global financial crisis. Now, depending on the year, we are at 12.5% to 13.5%. That is actually the lowest ever since records began in the 1960s. Often, when you are spending money on a policy intervention, you have to see what you are targeting and whether you are achieving that target. If you are giving a subsidy to a manufacturing industry, what are you doing it for? Is it for employment? Is it for moving up the technology chain? What is the objective? For example, seven or eight years ago, corporate tax was cut. What were you trying to achieve through that? If you look at the high corporate tax payers, there are two or three categories. There are banks, large consumer companies and PSUs. In none of those cases would the tax cut necessarily help expenditure or investment in India, which presumably was the target. The assumption was that if you cut taxes, more money would be spent. 

If you wanted to give an incentive for investment, you could have given an incentive for investment. That has happened in the past. 

We moved away from investment-led incentive schemes to more production-led schemes. I think it has something to do with the WTO commitments that we have made, that we will not give income-based subsidies. Somehow, the production-linked incentive scheme was woven into that system. With the production-linked incentive scheme, what happened was that very large subsidies were given to specific companies. If you are giving land to someone at ₹1 per acre, or at some very low rate, what is it that you hope to achieve? These have been very chunky subsidies, many of the subsidies under PLI. Maybe some amount is more scattered, but what is the objective? That is where I did not get clarity. Manufacturing as a percentage of GDP is not doing well. What is it that you are trying to achieve? If you want to drive employment, that is mostly driven by MSMEs. One reason unemployment has gone up and manufacturing has gone down is that there were disruptions. Former chief statistician Dr Sen had given data that because of demonetisation and GST, the number of MSMEs came down by 10 million, or about one crore. During that time, it should actually have gone up by a crore. So, basically, there was a shortfall in the number of enterprises. Employment in India is driven mostly by MSMEs because large corporates do not want to employ people. I am not talking only about now. Even 20 years ago, I remember steel plants and auto plants boasting that they had no labour. They would say, “We have 400 engineers and no blue-collar workers at all.” That used to be the boast. They do not want to employ more people. Employing more people is what MSMEs do. So, depending on what you are trying to target, if you are trying to target employment, then you should be helping this kind of category rather than very specific projects where maybe they will employ 5,000 or 10,000 people, and that is it. Any intervention has to be thought through in terms of the objective and whether that objective is being achieved. 

Manufacturing, Deep Tech and What India Is Missing 

Vivek Ananth: Do you think that right now, the way we are conducting policy, there is scope for manufacturing to go up in terms of contribution? I remember that when Make in India was launched in 2015 or 2016, the target was 25%. 

Devina Mehra: Yes. That is what I am saying. For five or six years up to then, it was actually at 17.5%, and it has gone down from there. 

Vivek Ananth: In the future, do you think these specific interventions can make a difference? For example, I remember reading about a ₹1 lakh crore fund for deep-tech investments, where the government is underwriting this and evaluating startups. It does not seem like we are breaking new ground in terms of achieving success. From your perspective, since you have seen global companies across the world, what are we missing here? 

Devina Mehra: One fundamental thing we are missing is something I say every year when people ask me before the Budget, “What is your ask?” I never ask for tax changes and all that. I say: invest in health and education. Look at the latest National Health Survey. It is terrible that so many of our children are malnourished, stunted or underweight for their age. Unless you invest in health and education, you do not have the foundation. Even if your focus is only the economy, unless you have good-quality human capital, how are you going to do anything further? Look at China, for example. China spent on human capital even before it industrialised. By the time it really ramped up industrialisation, it had a literate and healthy population. We are currently at a stage where we are not able to do that. There is no sense in saying that people should have three kids each if you are not able to provide for the kids you already have. That is one part of it. People say there is unemployment, but there are also people who are unemployable. If you talk to many business owners, they will say that even if people have finished a certain amount of education on paper, they do not actually have skills. They are not even teachable or trainable. Human capital is where the real focus should be. Of course, it is not glamorous. It does not give results in one year. Long ago, this scheme was originally introduced by Kamaraj, but I remember when I was very young, MGR was promoting the midday meal scheme. At that time, the pink papers used to say this was fiscally irresponsible and that it would ruin the finances. But that is what built Tamil Nadu. Tamil Nadu today, on many parameters, is comparable to middle-income countries. Unless you invest in human capital, you cannot build long-term strength. We are all very proud of Indian-origin CEOs in the US, but that happened because India invested in the IITs and IIMs back when no other Asian country did. 

Book Recommendations and Behavioural Lessons 

Vivek Ananth: Got it. I wanted to end with your personal recommendations. Apart from your book, Money, Myths and Mantras, which we showed in the beginning, is there a book, show or anything else that you think could enrich our audience’s lives? 

Devina Mehra: No, no. I am this weirdo who does not even have any OTT subscription. So for me, it is only books. I cannot recommend shows, but I have too many book recommendations. That is always my problem. Every year, I do a round-up of the best books I have read that year, and even then it is difficult for me to bring it down to 10 or 12. One permanent recommendation from me is Daniel Kahneman. Both his books are among my top five: Thinking, Fast and Slow and Noise. There are many others. This is something we did not speak about, but it is easy to understand finances and numbers. It is more difficult to get control over your mind. That is where most people falter. If you look at the long-term history of mutual fund investments in India, inflows peak around market peaks and bottom out around market bottoms. 

You see this beautiful little compounding over the years, but SEBI data shows that the majority of mutual fund investors do not remain invested even for two years. We spoke about Peter Lynch. That fund compounded at around 29%. Guess how much the average person in that fund compounded? Around 7%. A whole lot of them lost money. They lost money in the best-performing mutual fund in US history because they entered and exited at the wrong time. Any strategy, including ours, does not outperform all the time. What would happen in that fund also is that one year you would have blockbuster performance, everybody would rush in, and almost invariably the next year would be an underperforming year. Then people would rush out. That is the kind of thing you have to understand. So, Daniel Kahneman is my permanent recommendation. Right now, I am reading a book called Empire of AI, which is about the evolution of AI, mainly OpenAI, but also others. Something very different, if you want, is a Hindi book called Sipiya, which is about dohas. Human nature does not change over hundreds or thousands of years. That is why great literature persists. No matter how much the outer world changes, humans are still the same. Today, we talk about the mind-body connection. Kabir wrote about this 700 years ago: “Chinta aisi dakini, kaat kaleja khaaye, Vaid bechara kya kare, kab tak dawa khilaaye.” It means medicine will not work if stress is what is causing your ailment. So, I like to read a wide variety of things. 

Vivek Ananth: Thank you so much for taking out the time. And thank you for sharing your recommendations. The last recommendation was really interesting. 

Vivek Ananth:  So, that was today’s episode of Bazaar & Beyond.  From Devina Mehra’s insights, one thing becomes very clear. In investing, the biggest risk is not only market volatility. It is also the beliefs we follow without questioning them. Whether it is retirement planning, equity allocation, global diversification or understanding market narratives, the core of successful investing is data, discipline, risk management and independent thinking.  The key takeaway from today’s conversation is simple. A portfolio should not be built only on the basis of comfort. It should be structured on the basis of your goals, time horizon, risk appetite and evidence. If you liked this episode, do subscribe to our YouTube channel. We will meet again in the next episode. Until then, stay safe and invest wisely. 

Disclaimer:  Investments in securities markets are subject to market risks. Please read all related documents carefully before investing.