“We don’t buy into stories… we don’t fall in love with stocks”

Devina Mehra is a Wonder Woman of sorts. She is brilliant and accomplished – she bagged a gold medal at IIM Ahmedabad despite being the youngest in her class, joined Citibank at the age of 21, co-founded First Global with husband Shankar Sharma and built it into a globally respected firm on the sheer strength of cutting-edge research. 

Mehra has largely stayed away from the limelight and let her co-founder, a wannabe bahubali in his college days, remain the face of the firm for the longest time. Some part of it may be her introvert nature, but it’s an enlightened choice to make in the world of investing where stardom is transient and having an ear to the ground and independence of thought are key to succeed. 

Over the three decades of tracking stock markets, Mehra has expanded the firm’s canvas from Indian equities to global equities, debt and commodities across multiple markets. She has an enormous capacity to soak in new input and the wisdom to appreciate the role unknowns play in determining the final outcome in investing. 

In part, that wisdom is the result of serious reading. A bibliophile, her favourite reads Thinking, Fast and Slow by Daniel Kahneman and The Invisible Gorilla by Christopher Chabris and Daniel Simons have influenced her thinking and approach towards investing. 

Mehra limits her interaction with the media. But, this freewheeling chat with Outlook Business has more than its fair share of insights. 

How did you manage to outperform this calendar year, when the market moved unpredictably? 

It came out of a systematic approach, what we call a human-cum-machine model. This is what helped us to get 26% return on our global function portfolios YTD upto September this year, as against the MSCI World which is down about 1% and a similar differential for our India portfolios. All this has come at much lower than market risk and volatility!

Coming to how we actually navigated the markets this year: Over a period of 20 years, we have built various models, and some of those had started to signal that something is going wrong, around February. These include indicators on market moves, relative weights and changes in various asset classes, and myriad other factors. For example, even within invested portfolios and funds if movements are unexpected (for instance, an unusual number of stocks hitting stop losses, which should not happen in a well-chosen portfolio). We watch all these factors closely and more so watch them in combination.  Plus, by February, even otherwise it was clear that COVID was going to be a serious problem. I had tweeted around that time that all schools have been shut in multiple countries (Italy, Japan, etc). Tourism headed to zero in many spots. 

The scale of disruption is literally unprecedented — didn’t happen during wartime as well. So, it was clear we were moving to uncharted territory. We just didn’t know how different it was actually going to be.

From the month of February, when our system sensed trouble in markets globally, we shifted our global exposure in fixed income to as high as 80% of portfolio weight and increased our gold weightage to 10%.

The interesting part is that after the markets crashed around the globe, we saw in the last week of March, there were huge opportunities in global equity markets, especially in certain sectors. We went 70% into equities, which increased further in April. 

Then at a certain point, we sensed a weakness in the US dollar and increased our weightage in commodities. 

In India, we followed a similar strategy where we increased cash, bought fixed income and gold in March, and fully invested into equities in April first week. 

Hence, we avoided the big fall of March, and went to near 100% into equities in April. So, we got the best of both worlds. That is how we were able to add tons of alpha, which is apparent in our India and global performance.

If you look at our asset allocation pie charts in February end and June end, they look completely different. In February end, it was 80%-85% fixed income, gold and other commodities, and so on, and now it is 80% plus equities in our global funds. Our equity holdings are spread across several economies starting from US, Brazil and China to Japan, Taiwan and Korea.  

Right now, your India portfolio is in line with market momentum, overweight on a number of chemical, pharmaceutical stocks. Some of them arguably look expensive based on their earnings multiple. How do you reconcile to this?

Now, the key is not to storify your investments. Right now, based on our model which combines hundreds of factors, some of these stocks are coming up – more so because they have been gainers from both the pandemic and China’s tensions with the rest of the world. 

Having said that, in line with our philosophy, we remain well diversified and have invested across a very wide range of sectors from IT services to steel!

And as always, we keep reevaluating our bets. Normally, we rebalance every quarter.

In India, for asset allocation portfolio in general, our equity weightage is less than 50%. We do hold a lot of government securities even now. Plus, we are still not going to the debt side for any kind of risk assets, not even investment grade, because we think there is still lot of bad news to come. That’s also the reason we don’t like the financial sector at all.

So, what does the model throw up on the financial sector? 
The first level model cannot capture big disruptions (like a pandemic or a possible trade war or geopolitical tensions), simply because it depends on historical data. That is where human intervention as well as other ancillary models come in. 

As far as financials are concerned, it is the same movie that plays out every decade and we know how it ends, when the lending frenzy ends. This time it came with the fancy names of fintech, digital lending, etc. but risk assets are risk assets, at times of stress in the economy.

After demonetisation, you had a slowdown in the economy, but it was camouflaged by the lending from the NBFCs and banks. This time there was trouble even before the COVID crisis which has further dealt a death blow to many businesses as well as jobs. 

In general, this is a sector we are always cautious on - more so, at times of disruption.

You can have one HDFC Bank in 30 years and of course, we were very bullish on it back then. But in 1996, even the size of the opportunity was huge and your competition was only PSU banks. Now, that is not the case. We don’t like the sector itself. 

Tell us what is unique about your human-plus-machine model. 
Over the past 20 years, we have developed many models. While learning from those, we have now built a machine model that gives us a portfolio universe. The model literally looks at hundreds of factors. Within this, we have also introduced natural language processing tools, so the model can read a digital copy of an annual report, quarterly conference call transcripts and so on, and find out how the tone is changing, whether it is more confident, less confident and so on. All of it from published data. 

We also did a lot of work on things which signal that there is some manipulation, for example, auditor fees going up over a period of time, or depreciation rates changing or earnings and cash flows not reconciling. We found that this model, almost every time, caught all those manipulations. At times, it catches it a year or two before the scandal breaks but it has caught it almost every single time.

This actually came out of quantifying what we did manually, in the past. We had called out the two big scandals Enron and WorldCom. The problem in Enron was in the subsidiary accounting, which we had flagged well ahead of the scandal.

In the case of WorldCom, we actually did a piece saying that we are discontinuing coverage on this company because we are unable to reconcile their cash flows and capex, and even after talking to the management we were not clear. When the scandal broke, the problem was exactly in that area. It has taken us a lot of time to build this expertise. The challenge was to put it all into a model, which we managed to do. 

We believe our machine model is reliable because, when you have a purely quantitative person building a financial or investing model, their mindset is very different. They don’t know how to prioritise what goes into it, so it becomes like a black box. 

Also, as a pure tech person, you will have a tendency to keep adding more and more factors. We wanted factors that made sense to us - that were logical.

Now the amount of information coming at you is so huge that if you actually want to do a screener for say 2,000 companies in India or 10,000 companies globally, it is physically not possible for anyone to do it. 

So we use hundreds of factors, some well-known but many not so, to get to a list of companies that can be expected to beat the market hands down.  

Our target was to have something that is objective, not biased and is able to process a lot of information. We always say, even God must bring data! 

Investing is meant to be an art and science. Are you saying you are de-emphasising the ‘art’? Or do you believe the ‘art’ is overstated?

What we are saying is that a lot of what is art can now be distilled into science, given the capabilities that artificial intelligence and machine learning have given us.

On top of that, there is still human intervention because every quantitative model uses past data and many uncertainties cannot be captured.

Moreover, when faced with a deluge of data, human beings end up ignoring most conflicting trends and data points, and form easy, under-analysed, oversimplified lazy opinions. 

Humans have a storification bias, which comes from our need to justify each investment with a story, which leads to several other biases such as confirmation bias, anchoring bias, recency bias and then the endowment bias, where you are convinced that your holdings are better simply because you hold them.  

Since the human mind’s tendency is to think in stories, not to think in terms of probabilities or statistics which is what the future is about, we need to change the approach to better the chances of success. If you toss a coin 10 times, you might get a very different result than the expected 50-50 one. Like, you might get all heads or all tails, in extreme examples. 

However, when you toss a coin 100 times, you will get close to 50-50 so you have eliminated volatility or luck. Now, with skill, you can improve this 50-50 probability significantly.  This skill, multiplied over a large number of “largely uncorrelated” bets is the key to the investment kingdom. That’s why our investment model evaluates a huge opportunity set, across markets and asset classes, massively expanding the breadth, and uses skill to better the odds of success. 

Think of a casino. A casino has a very small edge in every bet but by expanding the number of bets to infinity, it always comes out ahead. 

What we are doing through our model is to have a reasonable edge in terms of skill which is replicated across a very large universe to get meaningful outperformance at lower risk. 

Tell us about your portfolio construction.

In our India portfolios, we normally have about 40-50 stocks. We might start with 60 companies from our model, eliminate say 15 based on our judgement and how they measure up on our other models, and finally invest in 45. Invariably, stocks that the model picks display similar characteristics when they come in: improving ROCE, decreasing contingent liabilities, sharply improving free cash flows, stable depreciation policies, improving asset utilisation and so on. 

When we built our model, we tested portfolios with different permutation combinations, and found that the risk reward is best in this type of basket, the return was higher and the volatility was also significantly lower. From this machine generated list, we look for turning points for industries or sectors or patterns in a country and use that to narrow our selection, using our skill and experience. 

But on the risk side, we have no human intervention. If the model says ‘sell’, then you cannot justify and choose to wait it out. We do not interfere with risk management. In fact, the only intervention you can do is to make the risk management tighter.

Then, the model also suggests position sizing, although we might tweak it a bit in the end. We usually start out at 2.5% and it might go to an absolute maximum of 4.5% or 5% weightage in the portfolio, not beyond. We say no to ‘high conviction’ because that is where you tend to go wrong. 

Why do you say that?

Every time you say you believe in the story, in the company, in the management and especially if you talk about it, it becomes difficult to exit. The more you talk about it, the more difficult it will be for you to change positions when things change. 

Take consumer stocks in India. The story goes that it is a very predictable business and generates cash flow and compounds every year, but a lot of companies in that category have been quietly dropped. Once up on a time, ITC and Gillette were talked about in the same manner but are no longer mentioned in the list. 

Bata is talked about today but, even 20 years ago, the Bata story was exactly the same but the stock went nowhere. Look at Hindustan Lever. HUL had no revenue growth for the first decade of this century and the stock also went nowhere. So, the theory that all these companies have very predictable cash flows and therefore the price is predictable actually does not hold.

They have all had five- to 15-year periods of underperformance and the stock going nowhere.

Again, look at the finance sector. Frankly, I am quite appalled at all the fund managers who had 35% to 40% of their portfolio in banks and NBFCs at the beginning of this year. The economy even pre-COVID was in great difficulty, and post-COVID you know anyway where that is headed. You know this sector is highly vulnerable but somehow that risk was just not recognised even by experienced fund managers and their investors paid the price for it.

So, we do not buy into stories. We do not ‘buy and hold’ forever. And, we don’t fall in love with our stocks. 

It is good to invest emotions in your friends and family, not in stocks. Stocks are only your assets. Your assets have only one objective and that is to build your wealth. 

Isn’t that about getting a call wrong? If your research is thorough and you back it with conviction, the rewards could be rich… 
There is always an element of luck, however skilled you may be. No one is so skilled that they will always pick all the winning bets – not even Warren Buffett.

So, first of all you have to make sure you don’t make big losses. Managing our own money is different – one can do many things, take on risks, take concentrated bets and so on. 

I would say, it is almost criminal the way a lot of people manage public money in portfolios, just putting money into 10 or 15 stocks. The risk is very, very high. There are known unknowns and there are unknown unknowns always in the markets. And, you can never be sure. The future is only about probabilities. Who would have thought in January that the business of hotels and airlines would go down to zero for a long period this year? But it happened.

One of the most mind opening articles I have read in my life was The Loser’s Game by Charles Ellis. It is about a very simple concept that every game is a winner’s game or a loser’s game and how it transforms from one to the other. One of the examples is aviation. Some decades ago, to be a pilot meant that you had to be highly skilled, adventurous, make decisions on the go and only then would you be able to succeed - think dashing aviators like Charles Lindbergh and JRD Tata. So, aviation was a winner’s game. Now, it is a loser’s game, there is only one way to play, which is to make no mistake as a pilot. 

Ellis also gives the example of tennis, where he says that amateur tennis and professional tennis are not the same game. Professional tennis is a winner’s game. If you look at the championships, the top seeds must serve great, play brilliantly and they might win. But, if you are playing with your friend at the local club, you only concentrate on keeping the ball in play, making no mistakes – basically ensuring that you don’t lose. Your opponent will make enough mistakes for you to win, so the outcome is decided by the loser. 

Even in investing, earlier it was a winner’s game where you could find many undervalued opportunities and make loads of money. But now, it’s more of a loser’s game. You have to, first and foremost, avoid making mistakes. 

When there are so many smart people competing, you cannot hope to reach the zenith every time. First, you have to make sure you don’t lose, which means avoiding the big mistakes and losses. 

This year, for example, lots of mutual funds and portfolio managers just rode out the decline. Once your investors are down 30%-35%, it will take them a 50% move to even come back to zero. If you go from 100 to 65 or 70, then you have to go up 50%, which is a two- to three-year journey. You have to make sure you don’t take a hit of that sort. Risk management is key. 

I was reading this book on Amazon, about when they were a store. The big takeaway for me is that, looking back we think it is one linear growth story, but the reality is that at every stage they made dozens of bets. Jeff Bezos tried out music, he tried out toys, they bought dozens of these dotcom businesses during the ‘99-2000 boom and in many of these ventures, he wrote off hundreds of millions of dollars, literally. You were in an environment where capital was easily and cheaply available to let you do that — you take dozen bets and maybe one or two work out. That is the way Amazon was built. Even Bezos did not know in advance which of his bets would have big payoffs.

Even though you don’t have unlimited capital, you should still spread your bets and then ensure that you don’t take a crippling blow on any of them.

The takeaway is that ultimately in life, it is better to make large number of small bets rather than small number of large bets. Small number of large bets may at times make you money but that’s not a predictable, replicable strategy. Recently, Wirecard went up in flames in the UK and the funds having it as their number one holding also went down with it. 

So, you must limit your risk. And this year, more than any other, should have made it clear to everyone that life is not completely predictable. No matter how good you are at analysis, could you have predicted how 2020 would pan out for businesses or your own life?

People quote Warren Buffett out of context and say that, if you have high conviction, why not buy more of what you are convinced about rather than hedging your bets. But, it is not about not having skill because the fact is that this is a business of skill and luck. 

It is not that you should diversify because you don’t have skill. Even if you have skill, by diversifying, the outcome becomes better. You can mathematically prove that. Chess, I think, is the only game which has only skill and no luck. Investing is certainly not that. 

In fact, the most rewarding part of managing money has not been that we have been able to generate great return for our investors, it is that we have been able to do it with half the market volatility, reducing risk substantially.

All individual investors should also look at their portfolio. If you have 15 stocks, out of which six are financials and five are consumer stocks you actually do not even have diversification at the level of 15 because all the financials would tend to move together and the consumer stocks would tend to move together. Diversification means uncorrelated or at least less correlated bets. Only then the diversification is meaningful.

You were among the first broking houses to start looking at global markets. Why did you decide to go global?

One was boredom – once I had understood Indian companies, I was looking for more intellectual stimulation.

But, the other trigger was the crisis that hit the Asian Tigers towards the end of the nineties. That demonstrated to us that while all these economies were supposed to be booming and headed for a great future, suddenly you have a situation where anyone who was only in their home market was going through an absolute disaster. Look at Japan, it was the poster child for all that was good in business through the ‘80s and then it took such a hit that it took them three decades to recover. 

When we discussed our intention to go global with our institutional clients, they all said it cannot be done. They said if you want to do it, you go to another emerging market, start covering Thailand or something, because the same fund manager may give you some business there. But, we were not excited at the prospect of adding another small market. 

We decided to go to the biggest markets, the US, UK and Europe and cover the biggest stocks there, which at that time were IBM, Dell, WalMart, Intel, Nokia and so on. Doing things from first principles, we made our name very, very quickly. 

Even a simple thing like the fact that commodities prices are denominated in dollars, so what looks like a commodity price move might only be a dollar price move was not articulated that way at the time and we made it to the front page of The Wall Street Journal in 2003 for discovering this link. 

It has been a very interesting journey over two decades. US, and even markets like China and Europe, are much more complex than India and hence it took time to build expertise. In India, the companies we have today are not very different from the set we had about 20 years ago. In the US, the whole lifecycle of companies has been shrinking. 

So, what does this shortening lifecycle of companies mean for ‘buy and hold’ strategy?

In the US, if you look at S&P 500, for example, in 1958 the average age of the company was 61 years and, in 2015, it was 18 years and it has been declining further. Even the survivors like Proctor & Gamble are no longer the movers and shakers they once were. So, this whole concept of ‘buy and forget’ and ‘buy and hold’ is coming from a different market, from a different era. You have to be much more dynamic today. 

Plus, there is a huge survivorship bias when people talk about ‘buy and forget’, because you are only talking of the companies or stocks from that era that made it big but there are many others that fell by the wayside. Even in the Sensex, about two-thirds of the companies from the original Sensex are no longer there. The ones that are left are the likes of Tata Steel, which again are not of any great significance anymore. 

The other interesting thing when we studied the history of companies is that, it doesn’t matter how long you have been in business, the risk of doing badly or even going out of business remains the same. If you look at the original Sensex companies that went out of business, almost all of them had been around for a long time, they were fairly old but that hardly prevented them from going down. 

In the US, they did a study which found that, in the first five years, the risk of going out of business is higher. But beyond that, whether you have been in business for 10 years or 30 years, the risk is broadly the same. Even though a company that has been in business longer may have seen more cycles and more downturns, it does not really better the odds. 

What’s the case for investing in your global fund? 

The first reason for investing globally is to not subject yourself to Single Country, Single Currency, Single Asset Risks (SCCARS), and as the Asian crisis amongst many others showed, that can be very substantial.

Even in India, equity mutual funds have not even beaten savings account return for five years and have not beaten fixed deposit return for 10 years!

In USD terms, the Indian market has been almost flat for nearly a decade — mere 2% CAGR, even as global markets went up 2.5x.

As an investor looking to create wealth and optimise return, the key is asset allocation. Many things are there right on the first page of investment books, yet it takes all those decades of experience for it to really dawn on you and implement it. One key principle is that asset allocation determines 90% of your return. It is well known, but we do not think about it. And asset allocation is not about just allocating between value and growth strategies for defensive versus aggressive strategies, etc in the Indian market.

That is because, for a lot of people, equity is a source of entertainment and party conversation. If you are only in Indian equities, you are facing SCCARS. 

The next step is people think, “I will buy a Nasdaq fund or something and I have taken care of global diversification”. That is far from being good enough. It only gives you one other country and still only one asset class. It is better than being only in India, but only marginally. 

The fact is, over time, leadership changes every year in terms of which markets do well.

This year Vietnam is No.1 YTD, followed by the Korean Tech Index and then Denmark. Last year, Russia and Brazil were right on top and in the first half of the year, they were close to the bottom.  

By the way, Taiwan has been good because a lot of the business that was going out of China has also been going to Taiwan and South Korea. You have to look at the whole globe, that’s the point. 

We cover the whole world with our global fund, as also all asset classes, which is equally important. This was an unusual year and as I said earlier, we have gone all the way from single-digit equity exposure to 80% equity exposure at various points. So, the asset allocation has to be tactical and dynamic.

In some markets, we buy stocks directly. In others, we may buy the indices. What we provide is a truly globally diversified portfolio across all asset classes that is managed dynamically and tactically. And at the cost of being immodest, the results have been superb!

Polarisation is usually a good indicator of a market peak, but this time the market seems to have stayed polarised for long. Do you agree?

Till sometime ago, this was true. But, some of the earlier polarisation is no longer working. The past few months, consumer stocks are no longer outperforming. Right now, because there is so much uncertainty, a sector which has more predictability should do better, but I think that the whole run-up prior to COVID has prevented that.

Is polarisation a reliable indicator of the market peaking? Does it come through in your model?

We have some models where we track a lot of these relative performance parameters. When a weight for a sector or a certain type of stock becomes very low then you watch that turning point because the market always re-balances at some point. Consumer stocks are not something that we currently like, on that basis.

But, if you look at market action in June, the difference between the US and India was that, in the former, the good stocks and sectors — meaning businesses where you could see that things would be better, which are digital space, certain consumers, pharma, medical-related and so on — were doing better. In India, for some time after the big fall in March, you had one rally in all stocks. Even leveraged stocks like banks and NBFCs bounced back. We knew that was a low-quality rally and did not participate. So June, we underperformed a bit but July onwards the rally became more logical (the way it had been globally) and we did very well.

In your global portfolio where does India figure now in terms of allocation?

We don’t have India in our global portfolio currently. We think India is going to be an underperforming market. And, there are better opportunities elsewhere. If you really look at the economy, the signs are not good at all. Even though, the market is no longer a reflection of the economy, the way we used to read in our textbooks.

Now, economic power is becoming very concentrated everywhere, it’s like a winner-takes-all in many areas. In the US itself, the number of listed companies has halved over 20 years and today the top 200 companies make more profit than the entire listed universe. That means the rest are not even profitable on the aggregate. It is a very different world.

Many companies now are truly global. When you buy a US tech company or a European pharma company, their dynamics may have little to do with the home country economics. They are globalised. And the companies themselves have more concentrated economic power, so that is why we have to really disassociate the economy and the company. But having said that, in India, you see serious economic problems, so you have to be very careful as to what you are buying. Definitely, there are better bets in the world available currently.

Today, most top tech companies in the US trade at price earnings of 30-33x. If you look at the multiples in India, 30x for those companies with fair amount of predictability versus 30x for some of the growth companies here with all kinds of fragility, the former seems a better bet…
Exactly. It’s the other way, India is the outlier. There is no free lunch in India. Any company that generates cash, does not require capital to grow, has some predictability of earnings etc is priced at crazy multiples. Not just in the US, I was looking at the Emerging Market indices, the dividend yield of the indices is around 2.5%. In the US, you get plenty of companies with high dividend yield and high free cash flow. In India, you never get that because anything which is that predictable in terms of dividend or cash flow gets priced so high that the yield becomes very low. It is crazy.

It’s the same story with rental yields. In many others countries, you can buy something with mortgage debt, pay that mortgage and still the rent will get you some positive return. They might be paying 2% on their mortgage and they might be getting 3% rent. In India, that’s never happened. You look at Gurugram or something, prices have gone up but the rentals are very low compared to the prices. In India, things get priced to perfection and then beyond. Like DMart, which has a good business, the valuation just went crazy. On the contrary, most well-known companies in the US are not even at 30x.

To add to that, interest rates in India are still 6% compared to near zero in the US.

Yes, a lot of lazy analysis is floating around. Many financial papers and magazines say that P/Es of stocks are at a historic high in the US. But these are meaningless looking at the fact that interest rates have declined sharply. People don’t do those one-step adjustments.

P/E is the inverse of earnings yield and interest rates have a direct impact on the expected yield.

Does it mean that there is a high degree of confidence that interest rate will sustain at these low levels and hence these valuations may be justified?
It may not be these low levels but they are going to be somewhere there. And as always, we are willing to make adjustments if the outlook changes.

How does your portfolio look now?
There are broad categories. In India, in line with global trends, pharma/other medical and chemicals pack have had relatively high weightage. But many others have also been great performers, from the IT pack to metals. We picked up MCX for instance, as part of our global theme of exchanges being cash flow businesses which have attracted more volume this year. And similarly, food delivery stocks globally and in India.

Globally, also there have been stocks from the medical side, which is pharma as well as other things such as testing. But besides the obvious stories of digital stocks that benefited from the change this year (Zoom, Netflix, DocuSign, JD, Netease, PayPal, etc), there have also been consumer stocks such as Domino’s and Clorox, home improvement stocks like Home Depot, tractor companies etc. There have been stock exchanges around the world from LSE to Hong Kong Stock Exchange as trading volume and participants went up. The range has been quite wide.

The market has also been quite discerning. For example, Domino’s did very well as people sitting at home are ordering pizza, whereas most other food/restaurant stocks did not do well. We may have some of the FAANG stocks, like we have Amazon and Netflix but we don’t have Google. So, our portfolio is not like a basket.

Consumer stocks in India have seen the kind of exuberance we are seeing in FAANG stocks in the US. Do you think the market is eyeing this as an extension of the same theme – the idea of long-term durable earnings requiring very little incremental capital, with a long runway of growth?
If you look at the Nasdaq universe, most of them have not required a lot of capital. In India, there are many companies which have been very cash-rich for quite a while. The theme in the US for the past few years has been returning money to shareholders – lot of dividend, stock buyback and so on. Some companies have obviously overdone it as what happened with some of the airlines, but generally that has been the theme. India is a capital-starved economy but the whole world is not so capital-starved.

We have not liked the consumer stocks here as besides the valuation part, we also see pressure on margins and receivables for them as retail gets even more organised in India. Again, the patterns are very clear if you see how the relative dynamics between these manufacturers and organised retail works elsewhere in the world.