In an earlier article, I argued that the S&P 500 (SPY) is more expensive than you might think if you only compared prices to trailing earnings. I recommended a different approach that suggests today’s valuation is similar to that of 1997 or 2006, a few years ahead of the respective bubble peaks of 2000 and 2007.
But what about credit markets?
While bubbling assets are a major part of the history of the Greenspan/Bernanke economy, so too is unsustainable borrowing. It seems wise to keep an eye out for another borrowing binge, especially as policymakers are encouraging all forms of financial risk-taking. And one place to check is the Fed’s quarterly “Flow of Funds” report, which recently took the fancy new title, “Financial Accounts of the United States,” but still goes by the nickname “Z1.”
Z1 data for the second quarter was released last Wednesday. It showed there’s little to worry about in the mortgage sector, which continues to shrink, whereas federal government debt stands out for having astoundingly doubled in the last five years. Soaring government debt may be our biggest threat today, but I’ll leave this topic alone for now. I’ll look instead at nonfinancial corporations, where I’ll check for froth by calculating two types of debt-to-profits ratios.
First, I’ll divide debt by nonfinancial corporate profits as found in the National Income and Product Accounts (NIPA), while making two adjustments:
- To eliminate the problem that debt ratios tend to skyrocket when profits temporarily shrink in a recession, my denominator is the maximum profit in any prior period of four consecutive quarters.
- To account for credit market instruments on the asset side of the nonfinancial corporate balance sheet, these are subtracted from the total “credit market instruments liability.”
Here’s the chart:
At first glance, you might say: “Okay, we’ve reached a 9-year high, but the last bubble was concentrated in the household and financial sectors. The second quarter reading is lower than it was the early 1990s and early 2000s, and that’s not so breaking bad.”
But just as the most recent earnings measure doesn’t tell the whole story when you’re assessing stock values, the same goes for the debt-to-profits ratio. Consider, for example, that interest rates on corporate debt reached extreme lows in the second quarter when the Z1 data was recorded. Back-of-the-envelope estimates (which I’ll post separately) suggest that profits are more than 20% higher than they would be if interest rates were closer to normal. And that’s just one of the reasons to question the permanence of an outcome that’s followed six consecutive years of ultra-strong stimulus.
A “smoother” debt multiple with a knack for flagging recessions
For a more revealing look at corporate debt, I’ll borrow a page from Professor Robert Shiller’s book. The earlier post I mentioned included stock market multiples based on Shiller’s recommended 10-year averages for corporate earnings. By smoothing the ups and downs of the earnings cycle, this approach offers a better picture on long-term valuation. And I’ll use nearly the same method for the corporate debt ratio. Instead of dividing by trailing four quarter profits, I’ll divide by 10-year averages.
(Note: I’ve modified Shiller’s method slightly by converting the profit figures in each 10-year window to constant dollars as of the last year in the window. This makes for more meaningful comparisons between periods with different inflation rates.)
Here are the results:
As you can see, the chart shows a persistent upwards trend. It also shows unusually high volatility in the Greenspan/Bernanke era, consistent with the Greenspan/Bernanke puts that encouraged overreaching in expansions. And with today’s policies setting new standards for “whatever it takes” central banking, Bernanke’s soon-to-be-named successor may yet ring in a new record for the debt-to-profits ratio.
But there’s a cautionary note in comparisons of today’s leverage ratio to the last three expansions. The last three times the ratio jumped above the current reading of 7.2 were Q1 1990, Q1 1999 and Q2 2007. And from these points in time, the economy fell into recession about a year later, or less, in each case. (The respective times to recession were two, four and two quarters.)
The last chart maps out three years of S&P 500 performance from just before to well after the historical breaches of 7.2:
The stock market picture isn’t all doom and gloom – at least the 1990 downturn was short-lived – but it isn’t exactly encouraging, either. What’s more, it adds to other signs of a new corporate credit bubble:
- Year-to-date covenant-lite loan issuance has already shattered the previous record, set on the eve of the global financial crisis in 2007.
- Leverage multiples based on EBITDA (similar to the ratios in the charts above but without depreciation, amortization and today’s unusually low interest expenses) are also well beyond 2007 levels, and even higher than the temporarily swollen multiples of the low-EBITDA years of 2008/09.
- PIK Toggles are bubbling up once again.
It seems to me that the Fed isn’t quite capable of detecting imbalances that accumulate in periods of über-cheap money and moral hazard. Or, maybe they just don’t think much about the potential consequences of those imbalances. And as Breaking Bad’s Walter White has been heard to say: “those consequences, they’re coming.”
In my opinion, the question isn’t so much whether the Fed’s approach leads to another financial bust, but where the bust originates and when it occurs. Evidence presented here suggests that the corporate credit markets are one a few possibilities for the where, and the when may be sooner than you think.
Hi great article on ZI, but I just wanted to ask one question:
How do you reconcile your chart on the growth of corporate debt with the chart on total debt to equity of the S&P.
I understand that you exclude financials and that you compare corporate debt to profits instead of equity, but I was wondering how you could explain that total debt has fallen so much versus total equity, especially if the new debt has been used mainly for buybacks. If so, debt (the numerator) would be increasing and equity (the denominator)decreasing under share buybacks, no?
Thanks David.
I’m not sure where you’re looking at debt-equity ratios, but maybe the equity is market cap-based and rising with the market?
Thanks for the reply & I guess I can’t really get my head around whether the investment grade space (S&P 500 and other large companies) are really more leveraged or not. To me it appears the corporate sector (investment grade) is still flush with cash but if I am wrong this is very important.
Any input most gratefully received
Got it. Here are links with debt/EBITDA for high or investment grade nonfinancial corporates. One (Citi) shows us above the last peak and the second (Barclays) just below the last peak, so not that much different from the first Z1 chart above.
http://www.zerohedge.com/news/2013-09-14/record-high-grade-leverage-means-pik-toggle-lbo-debt-back-and-worse-ever
http://www.zerohedge.com/news/2013-10-01/carl-icahns-nightmare-or-credit-bubble-4-simple-charts
From this and also because investment grade is such a large portion of total nonfinancial corporates, imo the Z1 gives a decent approximation. Also, I netted out asset-side credit instruments in the charts above, partly to capture any unusual cash holdings, and this didn’t seem to make much of a difference. Credit market assets are about $200 bln above the pre-GFC peak (in 2005), while credit market liabilities are about $1.5 tln above the prior peak (in 2008).
I’m happy to look at any data that shows this differently.
Beautiful thank you sir