What if everything you knew about earnings was wrong? To start, you would probably reevaluate your view of the stock market. In the past few months, some analysts have been asking you to do just that. They claim that the aggregated earnings reported for the S&P 500 companies, which conform to generally accepted accounting principles (GAAP), may be misleading investors. 1 1 Close It’s pronounced “GAP.” The second “A” is silent. They point to an alt-earnings measure put out by the Bureau of Economic Analysis (BEA) known as national income and product accounts (NIPA) earnings. 2 2 Close We think it’s pronounced “KNEE-PA.” Unlike GAAP earnings, which are used to evaluate individual companies’ performance, NIPA earnings are used to calculate national income and other macroeconomic data. It figures that some macro folks would think their macro measures are better than what the bottom-up stock pickers use. 3 3 Close So typical. One note is that there are also the earnings reported to tax authorities which are different from both. NIPA partly draws its data from tax sources, but they do make adjustments. From a philosophical perspective NIPA earnings measure current production of the entire economy while GAAP tries to get an idea of how much companies are really earning across everything they do. From a practical perspective the differences are that GAAP earnings allow companies to depreciate expenses over long periods, while NIPA has a stricter schedule. NIPA earnings are calculated both before and after taxes and do not include changes in pension valuations because those do not affect current production.
Despite these differences in calculation, most of the time the two earnings measures look similar. They tend to show earnings growth and contraction at around the same time. One exception is that during big cyclical moves, such as during recessions, GAAP earnings tend to have larger swings as companies take bigger write-offs. Still, it is unusual to have sustained differences in the relative growth rates during periods of macroeconomic stability. One such time was in the late 1990s when GAAP S&P 500 earnings outperformed NIPA earnings for several years. This immediately preceded the tech bubble burst and bear market of the early 2000s. At the time, some economists attributed the difference to aggressive accounting by companies – implying that NIPA earnings were more accurate than inflated GAAP earnings. As a result, some folks think that NIPA is more reliable and that any divergence is a good leading indicator. 4 4 Close Of disaster.
This leads to our current controversy. A few months ago, the BEA announced some downward revisions to past NIPA earnings data which created a wide gap between NIPA and GAAP earnings. 5 5 Close There was already a divergence. If I hear another “mind the GAAP” pun… , 6 6 Close Bureau of Economic Analysis: “2019 Annual Update of the National Income and Product Accounts,” 7/26/19. These routine revisions indicated that companies’ top line revenue from production was lower than previously thought. As a result, NIPA earnings have been essentially flat for the past three years. It’s unlikely that aggressive accounting is the explanation in this case because the rules are much stricter than they were twenty years ago.
But if companies aren’t cooking their books, what is going on here? Unfortunately, the BEA does not publish a line by line comparison with GAAP data. Without any conclusive data, it seems that we researchers are forced to do what we do best: put a bunch of popular current market stories on a roulette wheel and give it a few spins. A few have landed on stock buybacks. 7 7 Close It always lands on buybacks. An analyst might look at S&P 500 earnings per share vs. NIPA earnings and theorize that the reason for the divergence is that buybacks have reduced the number of shares thereby artificially increasing EPS. This theory is easily dismissed because the difference is still there if you look at aggregate earnings rather than EPS. 8 8 Close Did you really think we would blame it on buybacks?
Another possibility is that taxes are driving the wedge. This theory is attractive because the NIPA/GAAP gap seems to coincide with the implementation of the tax plan. Sadly, removing taxes from the earnings doesn’t entirely close the gap. Next stop: the difference in expensing. GAAP companies depreciate costs over many years while NIPA assumes they take the expenses all at once. This is more difficult to control, but we compared NIPA with S&P 500 aggregate EBITDA. 9 9 Close EBITDA refers to earnings before interest payments, taxes, depreciation and amortization. This also does not fully close the gap. We are aware that the comparison is not perfect because NIPA earnings include expenses for projects not included in EBITDA. It is possible that there has been a sudden increase in these expenses, but it seems unlikely considering we haven’t seen any crazy investment numbers. If anything, investment has been moderate in the past few years, so it’s unlikely that companies are generating massive expenses that don’t show up in GAAP.
There’s not much left on the roulette wheel. GAAP shows gains and losses from the funding of pension plans, but the difference persists in S&P 500 operating profits which excludes pensions, though it does shrink it a bit. The other main difference comes from industry composition. The S&P 500 is made up of a small number of big companies. The industry composition of NIPA is much broader; it has smaller companies and private firms – most importantly it doesn’t have the heavy weighting on a few technology companies. These companies have had some pretty good earnings growth. To correct for this, we looked at what earnings growth would be in a reweighted S&P 500 which more resembled the NIPA industry composition. The gap was still there, though again a bit smaller. So, while none of these factors are the smoking gun, it seems like each (except buybacks) contributes a small amount to the divergence. However, it is not fully explained, and it is possible that the data will be revised again to close some of the gap.
Several research pieces have made analogies to the late 1990s and more broadly to the gap’s usefulness as a predictor of recession or markets turns. These claims do not seem to hold up on an examination of the past thirty years of data. There have been times when reversals have followed very narrow differences in earnings growth and sometimes larger gaps have not led to any kind of reaction. However, it may be useful as a leading indicator of S&P 500 earnings. Our research confirms findings that changes in NIPA earnings tend to lead GAAP earnings by a quarter. The reason is not entirely clear, but we would guess it is because the smaller companies surveyed by the BEA have more cyclical sensitivity than the steadier large companies in the S&P 500.
Our research is by no means concluded on this topic. There are still plenty of gray areas in the data and there may be some pitfalls hidden in other details. We don’t think this gap is indicative of any major structural issue or has an obvious use for market timing. It does perhaps indicate that the economy is not as cyclically strong as the earnings of a few large companies make it appear. So everything you knew about earnings isn’t wrong, but maybe a few of the things you thought are.
What We Are Watching
Canada General Election (Monday)
On Monday, Canada will hold a general election to select members of the House of Commons. Opinion polls point to a close contest between the incumbent Liberal Party led by Prime Minister Justin Trudeau and the opposition Conservative Party under Andrew Scheer. Based on current polling, either party could win a majority of seats. Economic policy differences between the two major parties are relatively small, and market reaction to either outcome seems likely to be modest. It is also possible that the Liberal Party will fall short of a majority, but still be able to form a minority government with the support of one or more smaller parties, most likely the New Democratic Party and/or the Green Party. This might be the most negative scenario for the Canadian equity market, as these parties advocate minimum wage increases and oppose high-profile pipeline projects, a key issue for the market’s large energy sector.
European Central Bank Meeting (Thursday)
Next week, ECB President Mario Draghi will preside over his last meeting, concluding eight years in office. Under Draghi, the ECB drastically changed the way it approaches its mandate, moving from excessively tight policy in 2011 10 10 Close In hindsight, at least. to ultra-easy policy delivered through once-unconventional tools such as Quantitative Easing (QE), negative rates, and forward guidance, the most memorable example of the latter being Draghi’s “whatever it takes” moment in 2012. The ECB delivered a comprehensive package of easing measures in September, 11 11 Close Reportedly not without some internal controversy around the re-start of QE. and the October meeting will likely provide an update on the progress made toward the implementation of the package. The upcoming meeting is not expected to feature any policy announcements, but should set the stage for the hand-off to Christine Lagarde, who will take over from Draghi on October 31st.
U.S. Durable Goods Orders (Thursday)
Durable goods data is released monthly, and provides insight into orders, shipments and inventories of larger, generally expensive, and long-lasting goods. Orders and shipments have softened over the last year, as subdued business confidence, due in part to trade and global growth concerns, may be leading businesses to defer larger capital expenditure decisions. In August, the inventory to shipment ratio of durable goods reached its highest level since 2016, a sign that production is exceeding demand and may decline moving forward. Although recent positive developments in trade negotiations between the U.S. and China could help shift the trend in orders in future months, weak recent releases for the ISM Manufacturing PMI and industrial production data suggest a continued slowdown is likely for now. A weaker than expected durable goods order for the September report could boost Fed rate cut expectations in the coming months.