Chapter 8
GDP—Overrated
Gross domestic product (GDP) is probably the most universally followed economic data point. It is old and lumbering and frequently wrong. There have always been many shortcomings in the measurement and use of GDP, but it is widely used to measure economic growth and the size of an economy. However, as an economy goes through periods of innovation and rapid changes, its use as a tool to measure actual economic activity gets even weaker. Despite its flaws the securities markets still react when the GDP news release comes out, understanding GDP’s weaknesses and the value of other measurements of economic activity can give an investor an advantage.
GDP is the total market value of goods and services produced in a country. It measures personal consumption (C) plus business investment (I) plus government spending (G) plus net exports (NX) to total GDP, or:
GDP is not supposed to be a measure of the welfare of a country. It is a reasonable measure of the overall size of an economy and the broad trends of economic activity in a country. The GDP report is given in nominal and real terms, the real terms adjusts for inflation and shows it in “constant dollars.” Real GDP is usually what is used, as it separates improvements in the production of the economy from price increases. The seasonally adjusted rate of quarter-to-quarter growth is typically what is focused on. This is supposed to be a measure of the expansion or contraction in the economy and the rate of growth is often factored into the analysis and financial models when examining an individual industry, company, or investment project. Often you would want to compare the growth rate you expect for a company or industry to the overall economy to see if it is performing in line or not and if it seems your expectations are realistic. GDP results and expectations are also obviously an important influence on the policy being set by the government, including central banks.
GDP is also a good tool to use when analyzing the differences between countries if you are making cross-border investments. You can compare the overall size of economies as well as their growth rates using GDP. It is common when comparing data by country to put the data in ratios relative to the size of each country’s GDP. For example, you might compare countries’ level of deficits to GDP ratios or government debt outstanding to GDP. Segment reporting within GDP releases can give insights into the drivers of one country’s economy versus another. For example, reports on exports and imports can be most helpful in examining these exposures in an economy. The degree to which a country’s economy is global can be a good thing in some environments and a bad thing in others, but it is worth understanding each economy’s dependence on these markets
GDP in the United States is part of the National Income and Product Account. The government collects data from multiple sources both private and public using reports and surveys. The Bureau of Economic Analysis releases three batches of data on quarterly GDP, the first is an advanced report about four weeks after the end of the quarter, and then it releases two revisions each about a month later as more data comes in. The first release is the least accurate but seems to get the most attention from the media, politicians, and the markets. Many European countries have only two releases.
The full GDP report in the United States, and most other developed countries, breaks out several sub-categories, and the trends in these categories can be more interesting than the major headline numbers. In the United States one of the most watched pieces of the GDP report is the section on personal consumption expenditures (PCE) as the United States is a very consumer driven economy right now. There are other important categories, for example, there are breakouts for durable and nondurable goods and services. There is also a breakout for fixed investments and private inventories. The inventory number can be a major swing factor and sometimes an indication of slacking demand or the expectation by corporations that demand will increase. GDP includes what is produced in the country and if an item is made but not sold it is assumed to go into inventory.
Examining the subsectors of GDP reports can show important long-term trends and give you a sense of a country’s drivers. For example, if one country is more dependent on service industries than another or one is more manufacturing based it will be evident when comparing subsectors. There are some major items that often skew comparisons between countries. Some examples include dependence on foreign exchange, defense spending by the government, and in some emerging markets remittances can be important and not always captured properly in the data. One factor to consider closely is how rapidly, and in what ways, technological advances are happening in the different countries and how each country is evolving economically from these changes (e.g., new businesses, faster growing business segments).
There are all types of decisions about what gets measured in GDP. It does not count the value of unpaid services, the value of a parent that takes care of a child at home and transports them to school would not be captured in GDP but theoretically a baby sitter that is hired to do the same service would be included in the GDP calculation. Similarly, illegal or gray market activities are not included in GDP so that items like undeclared tips, off the books’ employees, and significantly more dodgy activities are not captured (this is a larger factor in some countries than others). GDP does not include the benefits of a domestic company owning a plant in another country and selling products there. There are many imperfections in GDP. It is very much a lagging indicator. In the United States it is only measured quarterly. The first reported number is released about a month after the quarter ends and the final revised report three months after the quarter ends. Therefore, these reports include data that is as much as six months old. Long-term the agencies may make revisions years later but they often have no noticeable impact on asset valuations. The revisions in GDP can be very meaningful from the initial advanced report to the final release on both an absolute and a percentage basis. Figure 8.1 shows some recent United States GDP reports and revisions with some larger revisions highlighted as well as the expectations of the survey of economists as prepared by Bloomberg L.P. The market usually reacts the most to the first release. In one study on the United Kingdom’s GDP report in The Blue Book it was found that the 1959 GDP was revised eighteen times and went from being reported in 1960 as 2.7% and eventually was revised over time to where in 2012 the revised 1959 GDP was reported at 4.7%.35 A report by the Organization for Economic Cooperation and Development (OECD) found that developed countries tended to have very large revisions to GDP. In the clear majority of countries, revisions increased the initial estimate of GDP; interestingly the notable exception was the United States where the first estimated GDP figures were commonly too high.36
Figure 8.1: Examples of Revisions in U.S. GDP Annualized Q-o-Q
GDP is also often a critical measure of determining when a country is in expansion, recession, or depression. When you think about some of the relatively short recessions in the United States and the lag in getting the final GDP data this measurement does not seem to be very helpful. The 2000–2001 recession in the United States took place from the fourth quarter of 2000 through the fourth quarter of 2001. However, with a three-month lag in reporting final GDP numbers, the country was almost half way through the entire recession before it could officially be declared by these measures.
A Bit of GDP Background
In the early 1930s the United States and Europe were going through a major depression. Congress in the United States decided it needed a good measurement of the economy. It enlisted a Russian immigrant Simon Kuznets and organized a team of economists to estimate the national income. Similar work was being undertaken in Great Britain. Mr. Kuznets treated government spending as a cost to the private sector (it is after all typically funded by private citizens and corporations). However, on the other side of the ocean the very formidable Noble laureate economist John Maynard Keynes argued that it should be included as an addition to GDP. Mr. Keynes used the example that during wartime purchases by the government had to be treated as demand or GDP would fall even though the economy was growing and the crisis of war simply redeployed people’s spending. As GDP reporting become more systematized Mr. Keynes’s model won out. When the United States began its aid program after World War II, countries that received funds under the Marshall Plan were required to develop an estimate of GDP. Then in the 1950s the United Nations drew up a plan for GDP accounting and this plan also followed the Keynes model on government spending.37 It is worth noting that economic theory might have evolved differently and looked at government spending as an expense to the economy rather than an addition to economic activity.
A major concern in the current transitioning economic environment is how much business activity GDP reports may actually miss due to its methodologies in measurement. GDP reports in most countries breakout industry data. One concern is that in the United States the services segment of the economy is more than two times the size of the “goods” segment. However, the services segment is broken into slightly fewer categories (eight vs. seven); this is similar in many other developed countries’ reports as well. In the U.K. it is five categories versus four (but to be different they list them horizontally rather than vertically, or maybe the United States is different). You would think if services were so much larger there might be more nuanced categorizations, or perhaps it highlights how the service economy is harder to track than manufacturing. Another fact that may indicate new industries are not all being captured properly is the relative growth and sheer size of line items in the report that are labeled “other.” In a corporate financial report when anything is labeled “other” (e.g., “other revenue,” “other expenses”) and is a very large or growing line item, you need to question it. Therefore, it comes up as a red flag in these GDP reports when the figures for “other” categories are large and growing.
When the economy and businesses are transitioning rapidly, GDP becomes a weaker measure of the economy. GDP captures and records if one product is being produced and sold more and another is produced and sold less, it does not capture if there are massive improvements in the product that are being made. Consider that at one point a person may have bought a computer, a camera, a device to play portable music on, and a mobile phone. A few years later the sales of all those products are probably down significantly and that same person is simply buying a new mobile smartphone that can do all of the same things on one device—this will, at least initially, cause some noise in GDP unless the prices of the replacement products and the displaced products are exactly equal and all the surveys are properly capturing this change.
The modern economy sees all kinds of changes that distort old measuring tools. In some cases what was once a product has become a service. The likes of Elton John, Paul McCartney, and Robert Plant have been producing music that is ranked in the top of industry sales for decades. However, while they keep making music, what is being bought by the consumer has changed dramatically during that time. Initially when fans of these artists bought their music it was on flat discs made of vinyl. For many people it transitioned to a taped product—cassettes, or the terribly designed, eight-track tape. These products were all linear, but then music switched to digital formats and people bought music on compact discs, more music could be put on this product so in some cases people were getting more for their money (though liner notes became harder to read). Sales were also boosted as people upgraded their libraries to this new enhanced technology (the multigenerational upgrades has probably helped The Eagles Greatest Hits, be the top selling album of all time). Up to that point there probably was not much change in how GDP captured this data. However, then MP3 players and Apple’s iPod were introduced and things really began to change. You could download individual songs or entire albums from the computer, but these were still sales of individual items that could be recorded on the GDP ledgers. Individual sales of music have become more complicated to measure now as much of the industry has transitioned to more of a subscription model where for a few dollars you can get regular streams of, what seems like, an endless selection of songs. This has changed the economics of the music industry tremendously and it is hard to believe there has not been significant leakage in the GDP figures as revenue went from buying a physical product to buying a digital file to paying a subscription. The music industry may be one of the most fascinating cases of adaptation to technology, both on the creative side as well as the business side. In music, even the organizational structures have changed over time from ensemble to orchestras to rock bands driven by technology and financing.38 Similar changes in sales from a product to a stream of service has occurred in other industries such as in mobile telecommunications when plans went from minutes to unlimited plans and usage and revenue shifts have occurred due to the introduction of texting. All these factors cause disruptions between actual economic activity and how GDP measures it.
There are numerous data points available that can be used as a predictor of GDP, or more importantly, as a better gauge of real time economic activity in the economy. Many of these measures can give more insight into trends that can impact business decisions and investments, especially as the world moves faster. Some reports are better about trends and some are good about factors impacting specific economic segments, other reports are good gut checks on GDP. Below are comments on some other reports, but it is far from exhaustive.
One very good report to monitor is a monthly release from the Federal Reserve Bank of Chicago, cleverly titled The Chicago Fed National Activity Index (CFNAI). The index is a weighted average of 85 indicators focused on economic activity and growth. It is divided into four categories: (1) production and income; (2) employment, unemployment, and hours; (3) personal consumption and housing; and (4) sales, orders, and inventories. Monthly data can swing around but the report includes a three-month moving average, even more cleverly named, CFNAI-MA3, this is a more valuable tool to track. Some of the data that is used is lagging and some leading, but the report and its segments can be very helpful trend indicators.
The Federal Reserve Bank of New York puts out a monthly report titled U.S. Economy in a Snapshot. It gives a brief overview and then has break out sections on economic activity, business activity, households, and inflation, and so on. It also usually includes a special focus section. Both this report and the CFNAI are heavily dependent on government data, much of which is not very timely. However, both reports try to incorporate more real-time data and are helpful as they incorporate so many different key aspects of the economy in one place.
Two of the most widely followed, nongovernment issued, indexes are the Institute for Supply Management’s (ISM) manufacturing and nonmanufacturing (i.e., services) purchasing managers index (PMI). The releases come out monthly and are both timely and highly responsive to changes in expectations of economic conditions. The PMI is a single figure based on a diffusion index. More details of data at the medi-level of the economy are in the Report on Business, which accompanies both releases. IHS Markit is a commercial entity that prepares PMIs for manufacturing and service segments for the United States and economies around the globe. This can be very helpful in having comparably prepared data across various national markets. Securities markets tend to react to major moves in the PMI figures or a major change in the direction of the numbers.
Many survey-based indexes that fall into the category of “soft data” utilize diffusion indexes; these are often misused by the media and politicians. It is important to understand them. Diffusion indexes show the dispersion of change. A simple calculation would be to take a survey of a group of sales people on their expectations for the next year’s sale. Any one that expects sales growth above +0.3% would be given a value of one and any one that expects sales growth to be between +0.3% and zero would be given a value of 0.5 and any that expects sales to decline would be given a value of zero. You then would average the results and multiply by 100. If the number is above 50, more people surveyed are expecting growth above 0.3%.
If a survey asks a 100 snake oil salesman if they are selling the same, more or less snake oil each month and the data is put into a diffusion index, the index will show if those sales people think sales are going up or not. It will not necessarily capture if a sales person is selling three times the amount they sold last month or just barely more than they sold last month. It may also not capture if the salesperson that says they are selling more makes up 20% of all snake oil sales or 1%. More sophisticated diffusion indexes can capture some of these nuances. The important thing is to realize that a diffusion index moving up does not automatically say there is more snake oil being sold it simply says more salespeople are expecting an increase in sales. One of the popular uses of diffusion indexes is also to analyze trends in gainers and losers in the stock market for technical analysts, this type of index, of course, is just based on numbers not a survey question, but can be skewed by which stocks are included or excluded.
A key to a good diffusion index is that any survey question is constructed well with little room for ambiguity; even a slight amount of ambiguity can change the results. As with all surveys the sample size and the composition of who is surveyed can make a huge difference.
There are also surveys and overview pieces about economic growth that are published and readily available, often these are less statistical but give good color. In the United States there is a publication by the Federal Reserve eight times a year known as the Beige Book that is based on regional surveys. In Japan there is the Tankan Survey published by the Bank of Japan quarterly that is broad reaching in its scope. Similarly, in Germany the Bundesbank puts out a monthly report that gives both commentary and statistics. The list can go on from the Reserve Bank of India to the Bank of England. As much of the global economy is dependent on energy sources, it is not surprisingly that one of the best sources for global economic trends is the OPEC Monthly Oil Market Report, which focuses on energy but includes good sections on the broader world economy.
Of course, stock markets, earnings reports, and interest rates can be hugely helpful indicators of what is happening in the economy. However, the reactionary, short-sightedness of daily or hourly focused market traders and technical flows must be considered when using this data. Market action and earnings over a short period of time are notorious for giving false positives and negative indications, looking at short-term trends can be more valuable than single intra-day data points. Stock markets are notoriously focused on near-term growth results.
Trying to compile detailed data on a massive complex economy, like GDP attempts to do, will always result in data that is flawed. The methodologies need to be based on samples and surveys that will be imperfect and by the time they are tabulated they will often be dated too. You can build baskets of corporate financial data that can give a clearer picture of the private sector of an economy. For example, you can build a model that tracks the revenue, cash flow and capital spending trends of the five largest public software companies and use it as a monitor for that industry, this can be repeated for each major industry sector. Corporate financial data on an individual company basis, or aggregated, can be a valuable tool in analyzing economic trends, this data will be more accurate than the broad economic information, but much narrower in scope. Growth helps drive returns for many types of investments and GDP is typically used as a measure of broad economic growth. While previous parts of this chapter have raised concerns about GDP as an effective tool to measure near-term economic activity, it can still be very valuable to use, especially when examining long-term growth trends. However, you must understand what it might be missing. Not all growth is created equal and where growth is coming from must be considered. If a considerable amount of growth is coming from government spending it can raise many concerns. Most government’s primary source of funding is debt or taxes—though some own means of production (e.g., a nationalized airline or toll roads). If a country is seeing growth but it is driven by government spending, you must investigate how it is being funded and how sustainable it might be. You will also want to look at where the growth in an economy is coming from in the private sector. If an economy is getting all its growth from a few narrow industry segments it would be weaker than if an economy was seeing growth from a more diversified basket of sectors. Also, be cautious if government subsidies are driving an industry’s growth, this is an insidious way for the people’s taxes to be directed to benefit a sector that cannot attract enough business in the market place to support itself.
It is Not All about Growth
There is a bit of an obsession in the investing community, and in the media, about growth. It is worth remembering that there are investments that are much less dependent on growth. This is particularly true for income driven investments, bonds, preferred shares and higher dividend stocks can all post quite reasonable and relatively predictable returns during periods of slow growth. It is not that these types of investments benefit from declining economies, but in a slow growth environment their returns can shine on a relative basis.
As an investor you need to decide what GDP means for you. It is a number that is not timely but over several years can give a reasonable indication of trends. It is not a strong tool to capture near term trends that are driving economic activity, markets, and businesses. The release of GDP numbers often does move financial markets, shifts political policy, and influences consumer expectations. Therefore, it can not be ignored. However, it is such a lagging indicator to the real economy that any surprise in the figures that cause an initial reaction in asset valuations can create opportunities to enter and exit investments especially if it is an aberration from what the long-term trends are indicating.
It is no easy task to measure an entire economy. The economists that measure GDP have obligations to manage changes to the measurement tools in a very thoughtful, rules-based regimented fashion. This is an important job and data integrity is critical given how broadly this data is used. This makes it difficult to keep up with the level of innovation in the economy. The same is true for many other key economic figures that do not try to capture nearly as broad a spectrum of the economy. The faster an economy is developing new and different businesses the less likely GDP is going to be an accurate measure of the economy, therefore, GDP may be the weakest at giving insight into the newest segments of an economy.