The Exchanges

Q How would you describe your style back then?

A It's evolved multiple times, because I did not know where to start. I had a little bit of quantitative inclination. My feeling is that you're always good to have some guardrails, so it keeps you out of stupid stuff. I started with building quant screens, what literature was around then. There was one element which I was going to go on, which was top-down models. I was to vouch for that, how wonderful those models are, one of the best countries, and look at the best stocks quantitatively. Martin Zweig, Ned Davis type of stuff. Then came 96, 97 Asian crisis. I was co-managing EM fund. What I found was the top-down didn't work. The reason why it was okay performance-wise was because the bottom of the balance sheet kept me alive. Fundamentals were fine, but the top-down didn't work at all. I became a hundred percent bottom-up investor after that. One crisis after the other, as I've evolved over the years, it's become where top-down is a risk management tool, and we use it heavily. It's a switch off, not switch on. It should help you reduce risk, but not add risk. In other words, if Chinese growth is good and inflation is good, you don't buy China because of that. You still need valuations and corporate earnings. If there's a macro event, the war is a big one today. Maybe you want to be careful about the risk. If interest rates are going up, inflation is going up, what are the impli…

AI assessment note: “I started with building quant screens... Martin Zweig, Ned Davis type of stuff.”

Q What else do you think you're excited about?

A I think energy is a fantastic space because even if Hormuz opens tomorrow, it's going to take some time. If you look at Qatar, they've already said that 20% of thereabouts of their LNG facilities are down. They'll take three to five years to fix them. That means that you can buy companies at double-digit free cash for yield at 75, eighty-dollar oil. We don't see how it goes back to 70 and stays there. We don't need 100 fifty-dollar oil. At a 110, 120 dollar oil, you're looking at 15 to 20% free cash for yields. There are no managers who wants to increase capacity. It has nothing to do with the outcome of the war, but even if you assume the opens tomorrow, most of the models are still assuming 75 bucks. The physical is trading at a 110, 120. We talked to so many oil companies that our realizations are running 10 to 20 dollars, in a lot of cases, above what is trading in the futures market. If you look at jet fuel in Singapore, it's consistently traded at a 150 dollars. Somebody's gonna make a lot of money.

AI assessment note: “I think energy is a fantastic space because even if Hormuz opens tomorrow”

Q How did the various challenging times impact how you adapted thinking about managing the portfolio?

A You want to make sure that there's enough diversity of thinking in the team. You never want a team that agrees with the hundred percent. Huge mistake. Because it simply will cheerlead you. That allows you to at least have the other. The worst, in my opinion, it's the uncomfortable other. The biggest lesson in 2008 was, there are plenty of articles in the press about the mortgage crisis. I remember there's a business week article, how toxic is a mortgage a year and a half before the crisis happened. It's a cover business week. There are plenty of articles of mortgage bubble. Wall Street was in complete la la land. AIG went under, and the lesson was, let's talk to the analysts or journalists who are predicting this, which was the starting point of hiring journalists. Now we have an equal amount of journalists and traditional analysts. Their job is to take the opposite view by default. Journalists are pretty good at that. I've learned a lot hanging out with journalists. Half the team is journalists who basically criticize everything we do, and their compensation is structured that way. Otherwise, if you structure the comp where they agree, it's wonderful, it works, guess what will happen? They'll agree with you. You want to structure the compensation where actually by default, they cannot agree with you. They simply measured based on their calls over the long run. That was the big…

AI assessment note: “The biggest lesson in 2008 was... which was the starting point of hiring journalists.”

Q How do you figure out those position sizes?

A This is a cemented process part, what you're talking about is. The sizing should be based on how a credit analyst would think. You can't have a very large position in monoline business, which operates a very narrow niche. Can't do it. So if you think about how does S&P would give AAA, it would never give AAA to an E&P company. Just can't. No diversity of their asset base, geographically, business lines, but Exxon can get AAA. So the small company, or even a large company, even Monoline, can never be a large position. The top sizing always has to be, as a credit analyst will look at it, can it be AAA? So it is not based on purely on the conviction, but much more on stability of the business, so you don't blow up. Everything you try to do is, just don't blow up. What is not acceptable is market down 40, and we're down 43, we outperform. It's like, no, we think more like long short in a way. We can have very large push in Exxon. We can't have Oxy, which we really like, but we can't have our same size position because Oxy is much more narrow operation, much more risk.

AI assessment note: “The sizing should be based on how a credit analyst would think.”

Q How do you marry the notion of quality as barriers to entry in something like energy, which is also notoriously cyclical?

A Depending on the assets. Some of these are irreplaceable assets. If you have a big footprint in Brazil like Petrobras, those are not replaceable assets. They are profitable at 75, eighty-dollar oil with a decent production growth of two to three percent. You can't replicate those assets. Energy and commodities are particularly those where It doesn't matter till it matters. If you have one middle bell short, that's the only thing you think of. You don't think of semiconductor that way. There's lower barrier to entry business too. In shale, there are a bunch of companies which don't have that high quality assets. Shale depletes very rapidly toward 20 to 30%. So not everything would be high barrier to entry. The business that have long enough tail of producing and low enough cost could be attractive proposition versus something which is like software. If you look at it, how many companies have survived in software business over 30 plus years? Microsoft is an exception, and maybe Oracle, it's a lower barrier to business. Semiconductors used to be, in our opinion, it is not as high quality as it used to be, because everybody is getting into the game.

AI assessment note: “Depending on the assets. Some of these are irreplaceable assets.”

Q What'd you do in the subsequent couple of years to resurrect the ship?

A It's interesting because of the emerging market and Far East like Japan were the two products at that point. This is the early days of the Asian bull market. They had reasonably good performance, which I was running going into that. So we got a lot of traction in those after I became the PM. Almost two thirds of business in the next couple of years was Asian emerging markets, particularly Asia. Asia was more than half of our book. That sort of gave the ability to basically pay the bills. Some of the clients would joke around that you have more products than analysts. There were two analysts, and we had to restructure the team because we couldn't afford the prior ones. I remember one instance where we had shortlisted a couple of people to replace, and one of the other folks who were involved in the hiring process said, who would you hire within two? I said, answer is simple. Who's the cheaper one? Because that was a reality. Gave us the first leg of growth. That gave the first base. Pulkheads came from European clients, banks, et cetera, you know, fund-to-fund business. Then institutions came later in U.S. Then slowly, international took hold as the performance turned around in global and There was one product after the other, which had different cycles.

AI assessment note: “we had to restructure the team because we couldn't afford the prior ones.”

Q What are some of the areas where you've gravitated to as quality growth that other people might not put names in the same bucket?

A There's a laundry list of them. It's fascinating about semiconductors who would have predicted that the whole industry would be 12, 30, 10 forward revenue. In twenty-twenty-two, we wrote a paper, is software the new shale? In last month, we put almost ten billion dollars to work in software. If you look at from the lens of what are the barriers to entry, and is the outlook improving? If you look at steel, steel has become much more high barrier to entry business everywhere. Try to set up a new steel plant in Europe. Good luck. Cold. Very difficult. You won't get approval. Your grandkids might get approval. We are a truly equal opportunity investor. Almost everything is fair game, unless their client restrictions, depending on the bad instrument and forward quality. Today, for example, in 12 months now, we have almost nothing in semiconductors. We have almost nothing in tech. In last month or so, we began to get excited about enterprise software. Everybody feels that the HR system would be wide coded. Good luck with that. Significant energy exposure from time to time. And for 10 years, we didn't have any exposure for energy. Very opportunistic in that context in terms of, because the barriers into some areas are actually going up, and some areas are actually gone down dramatically. If you look at capital cycles, it tells you where the longer return should be. So we let it drive …

AI assessment note: “If you look at steel, steel has become much more high barrier to entry business”

Q and an awakening. If that's not how you would describe your investment management tech, request a demo at ridgeline.ai. And now, back to the show. Under the idea of making sure you have diversity of thought on the team, that's a sufficiently contrarian thought for a quality growth investor. I imagine you have some people on the team that don't agree with that main thesis. What's the dialogue like inside?

A It's been partly debated, and the debate has been around a few different things. Number one is the cloud. The data isn't there. Is that runway still there or not? Now we're getting more clarity that more than half of the backlog, hence probably the revenue too, Is coming from basically anthropic slash open AI. If you look at Google and Microsoft is more than half open AI. Amazon is more than half. So they invest with open AI, but they essentially give them compute credits, which are then utilized at Amazon. We don't know for sure, but that's what really is going on. The debate has been around how durable is that? And can this become a more profitable business? The third is around as the pricing of tokens begin to go up, which it has. Does the demand sustain? Because the whole issue compute shortage, kind of a non-starter. If Starbucks starts selling coffee at 25 cents, there'll be shortage of Starbucks coffee. When you're underpricing everything, if you look at CoreWeave and EBS, they're bleeding heavily. That means they're not covering the cost. The real test of shortage is when you're priced appropriately. It's a capitalistic system, so when you subsidize something, there will be shortage. There should be shortage. Those are things that there have been a lot of debate on. But I don't think so. There's that much debate on longer term, A, from a technology perspective, also in …

AI assessment note: “It's been partly debated, and the debate has been around a few different things.”

Q What's changed in how you think about investing from your early days in the business?

A The biggest thing is that you begin to appreciate how little you know. You become humbler because the conviction level actually goes down. I began to appreciate a lot more of having, making sure that folks are constantly poking holes and have different opinion debate. That's a must. If you want to survive anybody in the business, if you can, internally, that's part of risk management. If you ask me, that's probably the biggest realization. Avoid deep ideologies. We've had a biggest motion of AI. Maybe we'll have it again. The valuations don't make any sense. Plus there are far better risk return opportunities. Exxon is almost similar market cap as AMD. Depending on the oil prices, this year they generate probably fifty billion of clean free cash flow. AMD will be lucky if they generate nine. I'm sure it's going to change the world. But this is going to be nine and 50. That assumes oil is 75 bucks, not 120 bucks. At a 120 bucks, Be careful, AMD. AMD is a fantastic business management and everything else, but the math is just not working in some of these. We're dogmatic about the math rather than dogmatic about our views as such, but with a forward quality. Deep embedded ideologies is what is the most dangerous. As to think that big teams and super specialists is what is needed and they'll do better. We're a 180 degree opposite view. We are total PM slash invest. Everybody includ…

AI assessment note: “The biggest thing is that you begin to appreciate how little you know.”

Q From two to 2016, somewhere along the way, you hit 2008. What happened in that particularly difficult period of time?

A You always learn from mistakes. We had significant banking exposure from 2002. Fannie Mae, Freddie Mac, AIG, in Europe, Anglo-Irish Bank, Northern Rock, large positions. I got nervous in early 2007, so we had exited all our banking exposure and financial exposure. However, we had a lot of energies, very bullish on energy and commodity. The whole thesis about decoupling didn't connect the dots at all. Come September, markets sold off almost double digits post-leaven. Within two weeks, I had fully recovered. I had too much energy exposure, and that became a disaster, because that melted by October. I remember Shulamiji was down more than half in a matter of, like, weeks. I've used it at the end of the day. Relative is fine in an up market, but over the long run, if you don't have absolute returns, nobody needs you. You don't pay bills with relative performance. In a bull market, everybody thinks a relative, but if you want a long-term survival, you need an absolute orientation. That was an unhappy setup, because I recognize the financial issues, but didn't connect the dots on the energy side, that how significant impact would be across everywhere else. Okay, I mean, we obviously lost a lot of assets, and we continue to grow from there on. It was a huge lesson in terms of how it'll ripple through. For example, if you look at today, one thing is fascinating is that cyclical parts o…

AI assessment note: “I had too much energy exposure, and that became a disaster, because that melted”

Q How does management factor into your assessment? If you go back to the energy industry, even if you have an irreplaceable asset, there's a big price fluctuation. And in the boom times, you have to trust management to reinvest capital the right way, which has created all these booms and busts over time.

A Over the years, we've begun to appreciate how difficult it is to assess management quality. I probably meet a lot less management than I used to, because I found that I'm not very good at it. You thought we're very good, but you're riding the tailwind. Let the record talk. Others meet. So we have started to differentiate again, and you don't want to have everybody in the same meeting and everybody in the same Kool-Aid. If I've not met them, chance I'll be more critical. When people say, oh, I met the CEO, and oh, he's so wonderful. Okay, that should be the base. If he's a good salesperson, you probably should be In agreement with whatever the CEO was saying, management quality matters, but at the end of the day, numbers should still be the defining factor.

AI assessment note: “management quality matters, but at the end of the day, numbers should still be the defining factor”

Q How'd you take that and turn it into a professional career?

A There was no industry as such in India at that point. I was in Delhi working for an export oriented firm doing the paperwork for them, which was interesting experience because If you're doing documents for exports, you learn the drudgery of what proper paperwork means. I mean, there are dozens and dozens of bill of lading. You can't have any mistake, and you fill out precisely, and the bank may reject for any given reason. Which was one of the biggest learning experiences you forget as an analyst is that things have to be in a little more precise manner than simply putting in a model, or you grow at 15% next five years, and we all live happily ever after. That's when I thought about coming to US, because there was no real industry at that point in India.

AI assessment note: “That's when I thought about coming to US, because there was no real industry”

Q When you were having conversations with your clients and have a view so different from the market that for at least a period of time now, hasn't been in sync with what's been moving, where do you feel the pressures from those conversations, both on you and the organization?

A There's no question that stress builds up. It might be that stress is not a bad thing sometimes. You need a little bit of stress. Vast majority of our clients understand what they're buying into, and it's our job to ensure that they understand That we do take a lot of relative risk. We try not to take absolute risk. In last six odd ones, we've had some redemptions. We still had net new money last year, the ninth year, but we don't measure success by asset growth. When we went public, in my first letter as the largest shareholder of the company, I specifically wrote two things. We will never have AEM targets, and we never have margin targets, so that nobody's confused. You can grow in the short run by doing other things, but you also start reducing the alpha opportunity in terms of how people think and behave. It's my job to make sure that we stick to our core ethos, which is why we're doing what we're doing. That somebody's retirement is at stake. Some of your kids are not going to college. I'll tell you a story. There was a firm I knew. I went to see them in 2003, and they won't let me go up to the fifth floor. I said, like, I can go up. I said, no, no, you have to wait. They said that armed guards. I asked the guards, why do you armed guards? He said, during the dot-com, they lost so much money, they get death threats. There's an element of not blowing up somebody's retiremen…

AI assessment note: “There's no question that stress builds up. It might be that stress is not a bad thing”

Q How do you counter the difficult behavioral biases that come from, you know, endowment effect to things when you own the name, sell it, and then that ability to flip and come back in?

A Some of this learned behavior in terms of not anchoring to your past as much. If I look at my long-term record, there's not an area that I've not lost money in. You name it, every area. Over time, you also learn that there's a benefit of, if you lost money before, you're probably very endless now on that name. It becomes ingrained over time, so I don't have an issue flipping around at all. If I look at the long-term track record, I used to have a higher hit rate. The reason we consciously try to lower the hit rate in last decade or so, in GQG particularly, versus one total. Even at once only we're going to change that. The reason was because when you have a high hit rate, the problem is you have a high bar on what comes in. So you also miss a lot of multi-baggers for that reason. You actually lower the hit rate because then you have a small position. You know, it is not just like all the box and everything. Yeah, that's wonderful. But you also would miss the best ideas. The one way there's a more doubt. That means if I have doubt, the world has doubt too. So if you lower the hatred, that means you're also taking more chances. It's that early stage investing. If you were an angel investor, there's much more risk, but you'd probably have more home runs, multi-baggers potential. The payoff would be greater, but you can't have a large position in that. Look at software today. We fe…

AI assessment note: “Some of this learned behavior in terms of not anchoring to your past as much.”