In the face of financial markets that we view as steeply overvalued as a result of yield-seeking by investors, an important development in the financial markets in recent months is the gradual recognition by the Federal Reserve that, in the words of several FOMC members, the policy of quantitative easing has “overstayed its welcome”, and that the Fed has attempted to “overplay its hand.” As FOMC member Richard Fisher has observed, the Fed is “distorting financial markets and creating incentives for managers and market players to take increasing risk.” Member Charles Evans – correctly in our view – openly warns “I am very worried about the potential for unintended consequences of all this action.” Esther George concurred last week “these policy settings also contain risk in the long run to financial stability.” FOMC member James Bullard observed that the Fed has set “a high bar” for moving away from tapering QE. Charles Evans and Naranya Kocherlakota, in contrast, remain dovish as ever. Among the moderates, Sandra Pianalto indicated that the first step to unwinding the Fed’s balance sheet would likely be to stop reinvesting the proceeds of maturing Treasury and mortgage debt held by the Fed.
The key observation here is that the Fed is wisely and palpably moving away from the idea that more QE is automatically better for the economy, and has started to correctly question the effectiveness of QE, as well as its potential to worsen economic risks rather than remove them.
We continue to believe that the financial crisis was ended not by QE, but instead by the March 2009 decision by the Financial Accounting Standards Board to relieve banks of the need to mark distressed assets to market, allowing “significant judgment” instead. Though QE undoubtedly helped the mortgage market to recover, nearly a century of economic history had long rendered a verdict on broader hopes for QE even before the policy began. Both theory and evidence reject the idea of a useful “wealth effect” from stock market valuations to consumer spending.
As Milton Friedman and Franco Modigliani demonstrated decades ago, consumers do not alter spending in response to variations in volatile forms of income, but instead consume based on their assessment of their probable “permanent income” over the long-term. Until 2008, variations in home values did exhibit a wealth effect, but that was because those variations were fairly smooth and unidirectional. Equity prices are anything but smooth or permanent. To offer an idea of “effect sizes,” a 40% increase in the value of the stock market would be associated with a short-lived (roughly 2-year) boost to GDP of about 2%, and a short-lived reduction in the rate of unemployment of about 1%. Needless to say, any subsequent market loss would be accompanied by similar, though also short-lived, downside risk to the economy.
Ultimately, wealth is a claim on a stream of future consumption, aggregate wealth creation requires productive investment, and changing the transitory price paid for equities does not materially alter aggregate wealth, however aggressively policy might attempt to prop valuations higher in the short run.
In some sense, the weakness of the “equity wealth effect” is good news – an eventual and economically healthy retreat in equity valuations is likely to have only transitory effects on employment, but favorable effects on the productive allocation of capital. All of this is true, however, provided that the Fed does not respond by encouraging fresh cycles of yield-seeking speculation, and is vigilant about emerging risks in leveraged loans and the now record level of equity exposure financed with margin debt. Fed-induced yield seeking is alive and well, but the desire for new “product” is being satisfied not with mortgage debt, but with low quality covenant lite debt and equity market speculation.
With regard to the debt markets, leveraged loan issuance (loans to already highly indebted borrowers) reached $1.08 trillion in 2013, eclipsing the 2007 peak of $899 billion. The Financial Times reports that two-thirds of new leveraged loans are now covenant lite (lacking the normal protections that protect investors against a total loss in the event of default), compared with 29% at the 2007 peak. European covenant lite loan issuance has also increased above the 2007 bubble peak. This is an important area for regulatory oversight.
Meanwhile, almost as if to put a time-stamp on the euphoria of the equity markets, IPO investors placed a $6billion value on a video game app last week. Granted, IPO speculation is nowhere near what it was in the dot-com bubble, when one could issue an IPO worth more than the GDP of a small country even without any assets or operating history, as long as you called the company an “incubator.” Still, three-quarters of recent IPOs are companies with zero or negative earnings (the highest ratio since the 2000 bubble peak), and investors have long forgotten that neither positive earnings, rapid recent growth, or a seemingly “reasonable” price/earnings ratio are enough to properly value a long-lived security. As I warned at the 2000 and 2007 peaks, P/E multiples – taken at face value –implicitly assume that current earnings are representative of a very long-term stream of future cash flows. One can only imagine that recording artist Carl Douglas wishes he could have issued an IPO based his 1974 earnings from the song Kung Fu Fighting, or one-hit-wonder Lipps Inc. based on Q2 1980 revenues from their double-platinum release Funkytown.
The same representativeness problem is evident in the equity market generally, where investors are (as in 2000 and 2007) valuing equities based on record earnings at cyclically extreme profit margins, without considering the likely long-term stream of more representative cash flows. There’s certainly a narrow group of stable blue-chip companies whose P/E ratios can be taken at face value. But that’s because they generate predictable, diversified, long-term revenue growth, and also experience low variation in profit margins across the economic cycle. Warren Buffett (Trades, Portfolio) pays a great deal of attention to such companies. But looking at major stock indices like the S&P 500, Nasdaq and Russell 2000 as a whole, margin variation destroys the predictive usefulness of P/E ratios that fail to take these variations into account. Similarly, the “equity risk premium” models often cited by Chair Yellen and others perform terribly because they fail to capture broader variation in profit margins over the economic cycle. Even measures such as market capitalization / national income andTobin’s Q have dramatically stronger correlations with actual subsequent market returns (particularly over 7-10 year horizons), and have been effective for a century, including recent decades.
The FOMC would do well to increase its oversight of areas where exposure leveraged loans, equity leverage, and credit default swaps could exert sizeable disruption. From a monetary policy standpoint, the effort to shift from a highly discretionary policy to a more rules-based regime is a welcome development… except for speculators banking on an endless supply of candy.
Small fourth-quarter revisions for GDP and other economic statistics were released last week. Broadly speaking, we continue to observe GDP, real final sales and other economic measures hovering at growth rates that typically delineate the border between expansion and recession. With inventory accumulation the primary driver of growth in gross domestic investment, and households carrying more debt as a ratio of disposable income than at any time in history (supported only in the near-term by suppressed interest costs), we don’t observe a great deal of pent-up economic demand in the system save for the possibility of a weather-related bounce.
As I’ve noted in recent months, the correlation between historically useful leading economic measures and actual economic activity has been disrupted a great deal by quantitative easing in recent years. This has made economic projections much more difficult. So while we observe borderline economic activity, any expectation of an economic downturn would require much stronger evidence on a diverse range of measures, including equity price weakness, widening credit spreads (both on a 6-month lookback), a drop in year-over-year payroll growth below about 1.3%, and further softening in a range of national and regional Fed surveys and purchasing managers indices. As a rule, when the noise increases relative to the signal, the best information is obtained by looking for uniform shifts across several imperfectly correlated measures. We don’t yet observe that sort of uniformity.
Meanwhile, the latest economic data provide no optimism on the outlook for profit margins over the completion of the present market cycle or over the longer-term. From our perspective, the central economic relationships we’ve described previously continue to hold in current, real-world data – see The Coming Retreat in Corporate Earnings. Keep in mind, however, that some of these relationships are exerted with a lag of several quarters and are not immediate.