With economic and market data being released in mass quantities and at a rapid pace, trying to evaluate it all can feel like drinking from a fire hose.
To maintain an informed perspective, Abby Joseph Cohen, professor of business in the Economics Division at Columbia Business School, has developed a process to sift through it all.
“Some of it is heavy-duty arithmetic, and some of it is an art form,” she says.
For Cohen, the first key questions involve determining the underlying assumptions about the data. For example, she starts by considering: What are the data actually sampling? And is the time frame in which they were collected still valid? From there, Cohen can analyze the pertinent information and develop better-informed views on the economy.
In the Q&A below, Cohen distills and shares key insights, including the importance of maintaining a global mindset, the effects of aging demographics on the economy as a major structural change, and how to protect yourself from making mistakes.
CBS: When you look at the whole set of economic data that one can gather and analyze, what is your advice about how to sort through it and to be thoughtful about what to focus on?
Abby Joseph Cohen: This is going to sound flippant, but I’m quite serious: Many investors will expectantly wait for 8:30 in the morning when some data are released by the government, and the markets will respond promptly, and by 10 a.m., it’s old news.
What bothers me more is that when the economy is fundamentally shifting — and we have an economy that is shifting now — the data may or may not reflect what people presume. Most economic data are based upon statistical sampling. For example, when the industrial production data are compiled, the information that we see reported out two weeks after the end of that month certainly doesn’t reflect every single action within the economy. It represents a statistical sampling that was weighted based upon what was deemed important in the economy a few years earlier. For example, as our economy became less dominated by metal-bending manufacturing and more dominated by advanced manufacturing and technology in the 1980s and 1990s, it took a while for the government data agencies to collect the information from the newer, faster growing companies and to downplay the information from the companies that had become less dominant. Today, we have a very similar situation in which the service portion of our economy has become more dominant in the United States. And we don’t do a particularly good job of collecting the data from these firms.
CBS: What challenges do you see in that?
Cohen: It’s hard to collect the output or productivity of a service producer. For an airline, you can talk about how many passengers or how many miles flown, but if you’re in financial services, how do you measure that? Do you do it by the number of transactions or by the dollar value of transactions? Do you do it by the number of customers served? It becomes a little more confusing and complicated. It’s a moving target, and it’s hard to keep up with everything — particularly since, from a budget standpoint, data collection has not been a priority in recent years for the federal government. This underfunding was one of the major obstacles in the 2020 Census collection.
Here’s another example to keep in mind. Right now, there is much focus on inflation. Yet, most news reporters talk about year-over-year changes. I understand why they’re doing that, because they don’t need to worry about seasonal adjustment. The problem is that if we look at inflation on a three-month rolling basis, we see that there was an abrupt change a few months ago — a very dramatic and welcome deceleration. But this important point of inflection is not yet easily apparent in the year-over-year data. So it’s very important to consider the frequency of the data. Should something be looked at on a monthly basis, a three-month average, a six-month average, or year on year? Your conclusions may be driven by the frequency or periodicity of these data points.
CBS: What else is being overlooked?
Cohen: There is a tendency among US-based investors not to consider data from around the world. And I certainly understand that most Americans may not be familiar with data in Europe or Asia and elsewhere. The problem is, of course, that we are in a global economy. For the companies in the S&P 500, on average, 30 to 40 percent of their revenues and profits are generated outside the United States. If we’re only looking at what’s happening within the US, we’re missing a good piece of the picture.
CBS: We’re in the midst of a demographic shift globally, with aging populations growing. How are you regarding what that means long term?
Cohen: The aging of the population is going to be one of the major factors when we evaluate which nations do well in this century. We had foreshadowing of this by watching Japan over the last three decades. It’s also been apparent in China and parts of Europe during the last two decades.
From an overly simplistic view, economic growth is related to the productivity of a nation’s workers multiplied by how many workers there are. In Japan, the workforce stopped growing and is now shrinking. And if we think about why the United States was the fastest growing of the developed economies over the last two decades, a critical reason has to do with our growth in population and labor force. From a population standpoint, Europe on average was growing 1 percent per annum or less, and the United States was growing 3 percent per annum. And here’s a very important kicker: Our domestic birth rate has been moving lower. So how have we made up the difference? The answer very simply is immigration. If we look at the 10 years ending in 2019, as the pandemic was beginning, about 50 percent of the net increase in our labor force came from immigrants. That is really quite extraordinary. And it’s not just what some politicians would have you believe, “Oh, it’s those low-end workers.” It’s also high-end workers. Something like 60 to 65 percent of the working PhDs in the United States in science, engineering, and medicine are immigrants.
CBS: Where do you see the latest developments in AI fitting into this picture and how that will affect the economy and jobs?
Cohen: I studied computer science as an undergrad, and even though I try to follow the literature, I’m not quite sure how to answer your question. What we have seen, not just with AI but automation in general, is that jobs which are repetitive are most likely to be revamped or eliminated and it doesn’t matter whether it’s at the low end or the high end of the labor scale. In manufacturing plants, for example, 20 drill press operators can be replaced by one computer operator who’s running 20 drill presses. We’ve also seen big changes in the service sector where whole categories of white collar jobs have been redefined.
The question about AI is whether these applications can replace creativity and imagination. We know they can replace the functions that are rote. If it is spitting out a regurgitation of already known facts, AI can do a great job. It can probably improve those chatbots we all hate dealing with when we have a problem with the telephone company. The thing I’m watching with great interest is the use of AI for applications such as language translation and teaching.
But will it create new things? Can it improve the creative process in writing? Well, maybe — but the examples thus far are mainly in straight essays drawing on existing work. Will AI write great poetry? Will it come up with new theories in economics or in science? Can it generate new ideas, new insights? I haven’t seen that yet. Never say never.
CBS: As a close observer of the markets for many years, when there are these new developments where it’s hard to know where it will land, what is your advice on how to analyze it as it unfolds?
Cohen: In the early stages, it’s not so easy to separate the wheat from the chaff. Number one: Sometimes what gets people very excited turns out to be fads that don’t really have staying power. And number two: In the markets, there’s the additional question of how much are you willing to pay for something?
If you go back to the beginning of the 1990s dot-com era, companies accounting for more than half of the market capitalization in the Nasdaq had no earnings and no cash flow. Most didn’t have coherent business models either. They had stories. And sometimes stories turn into something and sometimes they don’t. Looking back at that era, when so many stocks in the Nasdaq didn’t have any financial results behind them, it leads to head scratching as to how investors decided how much to pay for the shares.
For many of these securities in the dot-com bubble, the share prices weren’t driven by earnings or cash flow or even prospective earnings. For many of the young internet-related companies, the valuation metric was eyeballs. It was how many eyeballs were looking at the website. Needing to invent new ways of valuing financial assets can sometimes be a red flag.
Of course, brand new industries may require new measures of success especially in their early stages. If you have to devise a new valuation scheme for something, maybe that’s right. But once a company is no longer just a hope and a dream, but is actually an operating business, investors need to revert to some of the usual metrics and questions. Is there a sound business model? What are the operating margins? What does the balance sheet look like?
I also spend a lot of time looking at flows of funds. Is new money rushing in because other money has already rushed in? This was a very helpful question to ask when looking at the valuations some tech and other growth stocks reached by the end of 2021. Some of these are great companies. The question is, how much are you willing to pay for their shares?
In a momentum-driven market like 2021, the larger stocks in capitalization-weighted ETFs and indices became more and more important. A small number of securities attracted the lion’s share of new investment dollars. If everybody’s investing in a cap-weighted index like the S&P 500, then a big chunk of the new money coming in will be placed in the stocks that have already performed well. Relative valuations of shares became highly skewed. By the end of 2021, the P/E ratios on this handful of favorite stocks was double and triple the P/Es given to the rest of the market. Maybe that was the proper valuation. But you have to be able to justify it, and investors weren’t even thinking about whether these relative valuations were warranted.
CBS: What is your takeaway for your students?
Cohen: There is a list of the seven or eight factors that typify market peaks. I can tell you what they are, that they tend to repeat, and so on. But what they often have in common is investors’ willingness to believe that things grow to the sky. That thing, or company, may be wonderful, but if it’s already priced as being wonderful, where’s the upside?
Eighteen months ago. I was the guest lecturer at a class at Columbia and the students asked me about cryptocurrencies. I asked them in return, “How many of you are intrigued and participating?” And many of the students were. Trying to be provocative, I said, “How is this different than tulip bulbs in the 17th century?” And I explained, “Tulip bulbs were worth whatever somebody was willing to pay for them. But when the bubble ended, tulip bulbs resumed their usual value as a source of flowers. So I don’t understand the valuation behind some of these cryptocurrencies.” Of course, I was viewed as somebody who just didn’t get it. Some of the students have come back to me recently to say, “Maybe we should have been paying more attention.” But that’s okay. I’ve made my share of mistakes, too.
What I love about teaching at Columbia is that the students are so well rounded as a group. Within each class, there are people with incredible experience from many different industries and functions, and they have come from so many different countries. They all come together as an amalgam, bringing it together as a team. When we approach thorny questions about what’s happening in the global economy and financial markets, what will the future bring and so on, engaging with people who look at things from different perspectives is an important way to learn and enhance the outcome.