Rob Arnott is not a typical manager who manages money for investors, instead the 63-year-old founder and CEO of the California-based Research Affiliates runs a research-heavy outfit that offers investment strategies to fund management firms and index providers. Advising over $160 billion in institutional investment assets, including the likes of Pimco, Invesco PowerShares, and Charles Schwab, Arnott’s tryst with finance began when his parents, both PhDs, bought him a stock when he turned nine. Arnott soon turned out to be a force to reckon with when he demonstrated why market cap-weighted indices made for a poor alternative to fundamental indexing, in what came to be known as ‘Smart Beta’. Arnott believes it’s not market cap but a blend of fundamental metrics such as sales, cash, book value and dividends that deliver value for investors. The avid collector of vintage bikes, who considers those models unique, believes that though the US is heading into a long-term bear market over the next five years, the downside risk for EMs is limited.
In an article you argued that we are in a bubble phase. Can you take us through the evidence to suggest that we are, indeed, in a bubble?
We just released a paper titled, Yes. It’s a Bubble. So What? I think the most important point of that paper was to offer a definition for what is a “bubble.” The term bubble is used by lots of people in lots of different circumstances. Everyone has their own thought about what constitutes a bubble. So, the first thing we did was to create a rigorous definition. That definition has two constituent parts. First, in order for something to be a bubble, it has to be very difficult to construct a set of expectations for future profit optimistic enough to deliver a positive risk premium relative to bonds or cash by using any reasonable projection of expected cash flows. The second part of the definition is equally important and that is to do with the marginal buyer. Investors in a bubble market are almost all momentum investors who like the “story” of the asset with little regard for the underlying fundamentals, willing to buy thinking they can sell the asset to someone else for a higher price tomorrow.
Looking at today’s market using that rigorous definition, we see a lot of bubbles. Now, that’s not to say that the market as a whole is a bubble. But if you look at cryptocurrencies, both conditions are met. There’s no valuation basis for owning these assets, and the marginal buyer is buying because he thinks that somebody will pay more later. It’s the greater fool theory at work.
Growth stocks, especially the beloved new technology companies are in a bubble because while you could justify the price of any single stock, collectively it is really hard to justify the aggregate price of the technology sector. Technology would have to become a far larger part of the world economy than today, in order to justify the current prices. While technology is growing in importance, existing companies are facing constant competition from new entrants. But there are marginal buyers of a Tencent or Alibaba or Tesla or Netflix who believe that these stocks have such a glorious future that they are worth buying at just about any price. That’s a bubble.
What we don’t see is that the market in aggregate is in any sort of bubble. We’re not predicting a major market crash or anything of that sort. We’re just saying that certain segments of the market have become stretched relative to the market itself. The spread between growth and value has become unsustainably wide.
Jeremy Grantham in an interview with Outlook Business said that while valuations were expensive compared to historic levels, the psychological manifestation of a bubble is still not visible in the market. He says the reason for high valuations is because of the way fundamentals are today, which is that earnings are at a historic high as capitalism is virile. Do you agree?
I agree with much of what Grantham has to say. But I disagree with him in a couple of areas. Firstly, he says that robust earnings in an economy justify a high multiple. Well, if earnings sustainably double, does that justify a doubling in the market price? Yes. Does it justify a doubling of the valuation multiple? In other words, does it justify a quadrupling of the price? No, it doesn’t. To justify a higher multiple you have to assume a higher future rate of growth. When earnings are enjoying monopolistic power, and are at peak levels as a share of GDP, well that’s a ground for higher prices to commensurate with higher earnings, but not higher multiples. So, that’s one area where we would disagree.
Now do we agree that there are certain reasons that would justify current high multiples? Yes, there are. One reason is demography. When you have a surge in the number of mature working age adults, from ages 40 to 60, who are increasingly concerned about their ability to support themselves in retirement, they become valuation-indifferent buyers. In fact, the higher the price of the market the more aggressively they have to save in order to make sure that they have enough to get by in retirement. So, demography is a source of higher valuation.
Another flawed argument that we hear is that low interest rates justify high valuation multiples. Let’s accept for argument’s sake that when interest rates are so low, valuations should be high, because in the Gordon model if you lower the interest rate you raise the price. The reason for low interest rates is important. Low interest rates may be symptomatic of lower growth. In fact, if you lower growth and lower interest rates in the Gordon model, by an equal amount, valuation doesn’t change. The whole argument that high valuations are here to stay is dangerous because they are all based on circumstances that could turn out to be temporary.
But what I find interesting about the arguments regarding high valuations is that most of them are temporary and mean-reverting. And most of them should apply just as strongly in Europe, where your P/E valuation levels are half of that in the US. The Shiller P/E in Europe is about 16-17x and in the US it is about 32x. Consider the question, if low interest rates justify a 32 P/E in the US, why do zero interest rates in Europe not justify an even higher multiple? It doesn’t make sense that Europe should be cheaper than the US if their interest rates are lower.
As for demography, the soaring roster of people who are worried about retirement and are now valuation-indifferent buyers of stocks will become valuation-indifferent sellers when they retire. They have to sell in order to buy goods and services during retirement. So, as the population shifts from being dominated by people between 40-60 to people in their 70s and 80s, as is happening in Japan and a few other places, the pressure goes in the opposite direction. It is likely to put downward pressure on valuations, even when interest rates are low. What if interest rates go back up? Which at some point they probably will, then you’ve got a problem.
All of the major arguments for sustained higher valuation multiples are temporary, and don’t seem to be applied equally around the world. It’s hard not to be a bull in European and emerging markets (EM) when the Shiller P/E is half of that in the US.
Do you see a dramatic reversal at any point? And what would be the triggers, if at all?
I don’t know when, but it’s very likely that we have a serious bear market sometime in the next five years. It seems likely to me that we are going to see valuation multiples converge. That is, European and EM multiples go higher as US multiples go lower. That could take the form of either a continuation of the bull market in EM or a bear market in the US, or a bit of both. My long-term view on the US is pretty bearish because I do see demography switching from being a tailwind for valuation, as it was for the past quarter century, to a headwind in the coming quarter century.
As baby boomers retire they need to convert financial assets into goods and services that they can consume in retirement. That is going to be bad news for stock prices. India is actually in a very benign demographic circumstance because your population is still young and there is going to be a steady surge in the number of people in the ages of 40 to 60 which is the prime savings age. It’s going to be at least a half century before you have a serious problem with too many retirees. India is in a very benign demographic circumstance for the capital market. It is likely to enjoy a tailwind both in economic growth and from rising demand for financial assets.
You have been pretty consistent in advocating in investing in EMs. What drives your conviction? You spoke about India’s demographics, but if you look at the valuation, it is far from a bargain.
In general, India has enjoyed a tailwind from a rising population of mature working age adults, collectively seeking to own more financial assets. I think that the tailwind will persist for many years. But you are absolutely right, the valuation in India is high by EM standards. This doesn’t mean that the market will go up from current valuation multiples, but that the relatively high valuations are likely to be sustained.
The other element at work is, of course, relative valuation within EMs. The spread between growth stocks and value stocks is enormous. The pricing of Nestle India is more expensive than Nestle Switzerland. That doesn’t make sense. Alibaba and Tencent are priced at huge multiples and yet the Chinese state-owned enterprises are priced at 4-5x cash flow. That doesn’t make sense. Within EM, I think the best opportunity is EM Value. The spread between growth and value in EMs has almost never been wider than it is today. That creates a very interesting opportunity.
The valuations in the FAANG [Facebook, Apple, Amazon, Netflix, and Google] stocks don’t make sense either. What’s your take?
I think FAANG-type stocks are likely to prove to be very, very disappointing. Right now the seven largest market capitalisation companies in the world are tech stocks. You’ve got Apple, Amazon, Alphabet, Microsoft, Facebook, Alibaba and Tencent. We have never seen that before. Even during the tech bubble it was five tech stocks out of seven. The only time it came close was at the peak of the Japan bubble when the top seven stocks in global market capitalisation were all Japanese companies and five of them were banks. We know what happened after that. Those stocks crashed.
While I see a significant risk of a major bear market sometime in the next five years in the US, I also see a significant risk of a crash in the highest-multiple tech companies. Some of them such as Netflix and Tesla have cash flow that is not even enough to cover the servicing of their debt. This is the basic definition of “zombie companies.” The only way for them to survive is to continue issuing new stock and new bonds. That is not a sustainable business model on a long-term basis.
But the Fed could continue to keep interest rates low and they could find lower rates cheap enough to roll over the debt?
Low interest rates prop up zombie companies. Low rates are not a good thing and are available to companies viewed as very safe or companies that are politically well-connected. They serve to hold down rates across the credit spectrum. Towards the end of last year and even now, European junk bonds yield less than US 10-year treasury. How is this possible? Well, the yields are pulled down by the zero yield on government debt in Europe. Now, this also props up companies that correctly should be allowed to fail. Reciprocally, you see EM sovereign debt yielding more than US junk bonds. Well, that also doesn’t make sense because much of the sovereign debt is far higher quality than the US junk bonds. So, why should it yield more? It shouldn’t. And that means that buying EM debt is also a very interesting investment today.
What kind of strategy would you suggest to ride the bubble phase until it lasts?
Firstly, diversify into markets that aren’t bubbly and are priced sensibly. Secondly, look for segments of any market that are priced reasonably. A vivid example of that would be value stocks all over the world priced very cheap relative to growth stocks. Third, ratchet down your return expectation. Investors all over the world think that if they put money in stocks, they are entitled to a 10% or higher rate of return. That is just not true.
While FAANG-like stocks are overvalued, where do you see undervaluation, relatively speaking?
In terms of sectors, the attractive ones are resources, energy stocks, and the higher yielding parts of the market such as utilities and financial services. Broadly, value stocks still represent a good opportunity. In the US, value stocks have a Shiller P/E ratio of 25x. That’s still too high. But EM Value is priced at less than 10x. Rather than worrying about cherry picking individual opportunities, you should buy a broad-based EM basket and enjoy that diversification at low prices. If the US sees a bear market, there will be a sympathy decline in EMs as well. But unlike past cycles where EMs magnified the US stock market swings, they face less downside risk than the US. In fact, during the US bear market, EMs could actually fare pretty well.
What parameters would you monitor to detect turning points for the broader market?
There are a lot of things that can happen. One would be rising interest rates, and the other could be increasing economic uncertainty. We published a paper last year in which we showed that there’s a very powerful link between macroeconomic volatility and valuation. If you look at the trailing three-year volatility in GDP, and trailing three-year volatility in CPI inflation, when that volatility is low, valuations are high. But that tacitly implies a risk if the volatility of economic conditions goes up. So, a rise in interest rates, or increasing economic uncertainty could be a legitimate catalyst that could lead us into a bear market.
But looking to figure out what catalyst would cause the market to turn is dangerous. By definition, a catalyst is going to be a surprise.
You are called the ‘Godfather of Smart Beta’. Could you explain why Smart Beta really works?
Any strategy that is transparent, inexpensive, low turnover, and that breaks the link between the price of a stock and its weight in a portfolio is called Smart Beta. The idea behind Smart Beta is simple: why should we weigh stocks in an index proportional to its market capitalisation? In a capitalisation-weighted market portfolio, the higher the price of a stock, the greater its weight in the portfolio. It doesn’t make sense to give a stock more weight just because it’s expensive. What’s interesting is that an array of strategies that don’t link weight to price have outperformed the market all over the world. All of them add value for the same reason: long-horizon rebalancing. Harry Markowitz [Nobel Prize winning economist] said there are only two free lunches — diversification and rebalancing.
That’s why the Fundamental IndexTM (often called RAFI) that we developed in 2004 beats the market. If a stock soars, the Fundamental Index, or an equal-weight index, or any index that doesn’t weigh in proportion to price, will rebalance away from that stock and reduce exposure. If the price tumbles, we increase the allocation, to get back to our target weight. The result is that you’re contra trading against the market’s most extravagant bets — you’re buying low and selling high. So, you earn money based on rebalancing.
Isn’t contra trade not what value investing is all about? Then, why is it that the Fundamental Index has outperformed the value index over the past few years?
Capitalisation-weighted value indices have very little rebalancing. The price of a stock goes up and down and its weight goes up and down too. So, in a capitalisation-weighted value index, stocks become less inexpensive or more inexpensive. Because you are not contra trading against the moves, you’re not earning the rebalancing alpha. The only rebalancing alpha that we earn is for stocks that move out of value into growth, or vice versa. Hence, we find that the Fundamental Index consistently outperforms value.
Were there any periods when the Fundamental Index did not work well or exhibited weak performance?
Absolutely. The Fundamental Index takes high-priced growth stocks and re-weights them down to their economic footprint, and takes inexpensive stocks and re-weights them up to their economic footprint. You’re introducing a value tilt relative to the market, and what that means is, when the market is rewarding growth, the Fundamental Index will be facing a headwind. Now, if growth is winning by a small margin, then the Fundamental Index can still win. The rebalancing alpha can be strong enough to overcome a modest headwind. But if growth is winning by a wide margin, then the headwind becomes too strong and RAFI underperforms the market. It rarely underperforms value indices though. You can think of it as a value strategy that happens to be relentlessly better than conventional value indices.