End of an Era?

A Fed economist argues we should expect a slowdown in net income growth
 

By: Minje Kwun

The past 30 years appear to have been a golden era for net income growth in corporate America. While the real (net of inflation) growth in EBIT has been consistent with its long-term average of around 2% per year, real net income growth has nearly doubled over the past 30 years to average 3.8% per year.

Figure 1: Annualized Real Growth Rates (1962–2019)

Source: Michael Smolyansky (2023), based on data from Compustat.
 
What explains this enormous divergence from the 2% historical average, and is it sustainable? In a new paper, Michael Smolyansky—a principal economist at the Federal Reserve—argues that US net income growth has been anomalously high over the past 30 years because of historic reductions in interest rates and taxes. Since net income represents the earnings that are left over after deducting interest expenses and taxes from EBIT, any reduction in those two costs would structurally increase net income. As shown in the figure below, interest and tax expenses in the US have been declining precipitously as a share of EBIT over the past 30 years, from 54% of EBIT in 1989 to 27% in 2019.

Figure 2: Interest and Tax Expenses as a Share of EBIT (1962–2022)

Smolyansky believes that these structural tailwinds for net income growth are unlikely to continue in the future. For starters, corporate interest rates appear to have bottomed out in 2022 after nearly four decades of marching downward in tandem with Treasury yields since the 1980s. As Treasury yields have increased since 2022, so have corporate borrowing costs, with the Fed consistently messaging that it intends to hold interest rates higher for longer. Moreover, the decline in interest rates since the 1980s coincided with a reduction in effective corporate tax rates, starting with the Reagan administration. By the time Reagan’s second term ended in 1989, the US effective corporate tax rate had declined by 10 percentage points to 34%, compared to an average of 44% over the twenty years from 1962 to 1982. The next major step down in corporate taxes occurred with the passage of the Tax Cuts and Jobs Act of 2017, which reduced effective corporate tax rates from 23% in 2016 to 15% in 2019. At least in the near term, it appears the 15% tax rate could be a floor because the Inflation Reduction Act of 2022 imposed a 15% minimum corporate tax. And over the long term, Smolyansky argues that the near all-time high ratio of US debt-to-GDP will make it challenging for Congress to pass deficit-funded corporate tax cuts going forward. These two structural dynamics are illustrated in the charts below.
 
Figure 3: US Corporate Interest Rates and Tax Rates (1962–2022)

By accounting definition, growth in net income can only come from three sources: EBIT growth, changes in interest expenses, and changes in tax rates. As noted in Smolyansky’s paper, these mathematical relationships are true by definition, without relying on any assumptions. In Smolyansky’s view, looking at how these sources have affected net income historically can offer us a guide for what to expect in the future.
 
In the charts below, we summarize Smolyansky’s attribution of US net income growth over the 57 years from 1962 to 2019. Panel A shows the sub-period between 1962 and 1989, which was marked by a 6 percentage-point rise in corporate interest rates and a 10 percentage-point decline in effective tax rates. While rising interest expenses were a drag on net income growth over this period, this was offset by lower tax expenses. As a result, net income grew at roughly the same 2% rate as EBIT over the 1962–1989 sub-period. But over the next sub-period, between 1989 and 2019, interest expenses and corporate tax expenses both declined, resulting in positive contributions to net income growth. Corporate America’s good luck in having both of these expenses cut simultaneously over the past 30 years explains the near-doubling of net income growth relative to EBIT over the period, and relative to its own history of growing at 2% per year before 1989.

Figure 4: Contributions to US Real Net Income Growth (1962–1989)

Source: Michael Smolyansky (2023), based on data from Compustat.

The crux of Smolyansky’s argument is that interest expenses and tax rates are unlikely to fall below 2019 levels, so investors should expect lower net income growth in the future. He points to 2% real growth in net income going forward as an optimistic scenario, assuming that interest rates and corporate tax rates don’t rise any further from 2019 levels. For context, the 10-year Treasury yield in 2019 was 1.9%, the pre-pandemic low. Fast forward to current day, and the 10-year Treasury yield has risen to 3.8% and, in our opinion, seems unlikely to fall below 2019 levels. As for corporate tax rates, when the last major tax cut was passed in 2017, the cumulative federal deficit over the first nine months of fiscal year 2017 was $520 billion. Today, the federal deficit is nearly three times higher at $1.4 trillion over the first nine months of fiscal year 2023. Smolyansky argues this makes it difficult for Congress to pass another deficit-funded cut in corporate taxes anytime soon.

We believe lower growth expectations going forward also have important implications for investor returns. As Smolyansky points out, in 1989, the trailing P/E ratio for S&P 500 nonfinancial firms was 11.8x; by 2019 this valuation ratio had almost doubled to 20.3x. Therefore, it appears investors are expecting a continuation of the 30-year trend of near-double net income growth. But if future net income growth comes in at 2% annualized or below (instead of the apparent expectation for 4% growth per year), then S&P 500 multiples would likely contract going forward, according to Smolyansky’s analysis.

A counterpoint to Smolyansky’s projections is that his focus on the mechanical relationship between EBIT and net income doesn’t account for the factors that drive EBIT growth. For example, it’s possible that wide-spread implementation of AI could materially reduce companies’ labor costs, thereby causing EBIT to grow at a higher rate than 2% going forward. Indeed, the recent AI boom among a handful of US tech stocks in 2023 seems to be predicated on an assumption that productivity gains will boost earnings growth going forward. But, of course, this AI rally comes on the heels of a 10-year tech bubble, so the companies that are becoming more expensive were not cheap to begin with. Moreover, the US Congressional Budget Office projects real GDP growth of 1.9% annualized in the US over the next decade. As Smolyansky points out, it is difficult to imagine a scenario where the largest US companies, which have declining returns to scale, can grow earnings faster than GDP over the long term.

A separate counter to Smolyansky’s analysis is that he doesn’t account for the possibility that companies could deleverage going forward in response to today’s higher interest expenses. That is because Smolyansky believes that “reducing leverage is costly” because it would “require either issuing equity, which would dilute existing shareholders, or paying down debt, which would involve lower payouts to shareholders in the form of either dividends or buybacks.” We think Smolyansky is correct on the first option, regarding the dilutive cost of share issuance. But we believe he overstates the “cost” of the second option by missing the fact that paying down debt from free cash flow has the accretive effect of mechanically increasing a company’s equity balance. Much like paying down a mortgage, a company’s equity balance mechanically increases in an accretive way for shareholders when debt is paid down from free cash flow. In the past, when US corporate leverage levels had reached cyclical peaks, companies proceeded to reduce leverage levels over the next 10–20 years. This was the case in 1972 when US corporate leverage levels declined by 17% over the next 10 years, from 0.30x Debt/Assets in 1972 to 0.25x in 1982. Similarly in 1992 when corporate leverage levels reached 0.35x Debt/Assets, leverage declined by 23% over the next 20 years to 0.27x in 2012. Today, US corporate leverage levels are near their 1992 levels at around 0.35x, potentially setting the stage for another decade or more of deleveraging.

Given these dynamics, we think the key takeaway from Smolyansky’s research is that investors should be wary of expecting (and paying for) high growth rates going forward, since the good luck of the past 30 years may not be repeatable. And among levered equities, we believe investors should focus on companies that are already prioritizing deleveraging in their capital allocation policy to capture the accretive benefits of paying down debt from free cash flow.

Acknowledgement: This piece was authored by Minje Kwun, a rising sophomore at Dartmouth College. She majors in applied mathematics and is an attacker on the Dartmouth varsity women’s lacrosse team. She is looking for an internship in finance next summer.