September 7, 2017
Robert Brooke Zevin, Chairman, Senior Portfolio Manager
Many people, including a number of our clients, have been concerned in recent weeks that the long stock market rise since 2009 has finally reached unsustainable heights that will soon lead to a crash. Our first priority has always been to avoid or minimize losses in bad markets. Financial markets are always more or less unpredictable and subject to sudden declines. However, we believe the likelihood and potential magnitude of such setbacks can be estimated with some degree of accuracy before they happen. Use of such estimates is an important part of our investment process. At the moment, we estimate that the probability of a major stock market setback is low for the next 12 months or longer, and that properly diversified portfolios may reduce to a greater extent than usual the net damage from most of the potential declines.
As long-term investors, we are focused on protecting against market losses that are large and long-lasting, rather than trading in and out of stocks around small and brief fluctuations. Bear markets represent exactly the type of major investment potholes we aim to avoid. A bear market is defined as a decline in value of more than 20 percent. After bear markets occur it usually takes at least a year, sometimes many years, before the previous peak price level is regained. The most frequent cause of bear markets is an economic recession or depression. In fact, every recession or depression throughout the history of capitalism has been accompanied by a bear market. Other bear markets have occurred because of extreme overvaluation or a financial crisis in which people and businesses are scrambling to sell assets in order to pay their debt obligations. These two causes are closely related since most periods of irrationally high prices, or stock market “bubbles,” are fueled with increased debt used to buy stocks and other assets that contribute to the bubble. In fact, all three causes are related and inter-related. The increased debt that leads to financial crises is also typically a driver of the economic expansion that precedes a recession or depression. And declines in the stock market can themselves be not only the signs but also the causes of recessions.
With this in mind, here is our assessment of potential causes of a bear market in the near future. Although the current economic recovery is old by historical standards it has also been weak and slow compared to earlier recoveries. The weakness has extended to both real economic growth and price inflation. We continue to believe, as we have for some time, that the U.S. (along with many other countries) is in the midst of a prolonged period of slow growth and low inflation. It seems paradoxical at first glance, but weak expansions typically last longer than strong ones. Recessions usually start with the collapse of a market in which there has been rapidly expanding activity—internet infrastructure investing in 1998–2001 or home construction and resales in 2006–2008. At the moment almost every category of investment spending is already quite weak, so there is not much further for them to fall. New home construction has been very strong but is currently stalled for lack of resources, including land to build on. However, it is now too small a portion of the American economy to cause a recession if it collapses. The same is true of the decline that has been underway for some time in U.S. auto sales although not in the rest of the world.
Meanwhile, the global economy is experiencing an unusual coincidence of expanding economies in almost every country. This means that each country is able to grow faster and is less likely to experience a recession. Thus, for both global and domestic reasons even a mild recession in the U.S. is highly unlikely for the next year or more. Therefore, the recession/depression cause of a bear market can pretty much be ruled out for at least the next year.
What about a financial crisis? The enormous level of debt in the financial sector of the American economy has been cut in half since 2007. Mortgage debt has also been dramatically reduced. Interest rates remain extremely low. Inflation is also low, suggesting that interest rates will not rise steeply in the next few years. One major exception to this benign outlook is that student loans are approaching the one trillion dollar level while default rates continue to increase. Like many of the financial debts that contributed to the financial crisis of 2008–9, these government-guaranteed private loans are often scams created by the government to benefit banks and other private lenders at the expense of lower-income students seeking their way to higher incomes. The inevitable high rate of defaults on these loans has been harmful to banks and others, but banks also are now significantly less leveraged and better capitalized than they were at the end of 2007. In sum, corporate America has accumulated enormous amounts of new cash since the 2008 financial crisis, and households have about 20 percent less debt and more assets relative to income than they had at the end of 2007, making a new financial crisis very unlikely at this time.
Even without a recession, aren’t stocks still just too expensive? Of course this is possible and by some well-established measures even probable. U.S. stock market aggregate prices are at historically high multiples of earnings and trade at a high level relative to the replacement cost for all corporate assets. In fact the levels are as high as or higher than the starting point for virtually all the bear markets of the past.
But there is a catch. The average ratio of price to earnings (P/E ratio) appears to have been increasing for a very long time, although fluctuating wildly along the way. This could reflect the growing portion of national product for which publicly traded companies are responsible or growing monopoly power and political influence. A second catch is that we are now in a period of slow growth, low inflation, and low interest rates. If a 10-year U.S. government bond provides a guaranteed return of about two percent today, compared to an average of around five percent for the latter half of the twentieth century, then pricing stocks to provide a similar three percent excess return above bonds would imply earning eight percent on your stock investments in the past and five percent today. In the long run, the earnings of your stock in a company are indeed what you earn (or should earn, absent corporate skullduggery). When earnings are eight percent of the price of a stock, the multiple of price to earnings is 12.5, which is below the long-term average. When earnings are five percent of the price, the multiple of price to earnings is 20, close to the present situation.
Catch number three: the modern corporation has become increasingly asset-light. Some companies own little besides patents, trademarks, office furniture, and desktops connected to the cloud. Any necessary manufacturing, transporting, warehousing, and selling is subcontracted to others. The value of such companies often will be enormously larger than the replacement cost of its assets, while the subcontracting companies may often be overseas or not owned by public shareholders. Generally, brand names and monopoly power are not included in most companies accounting of their assets. Since both monopoly power and asset-light business models have grown over time, it makes sense that the ratio of stock market value to measured assets should also increase. Indeed, the one time this measure peaked higher than today’s level was the top of the dot-com bubble, which in retrospect may not have been as irrationally exuberant as many of us thought at the time.
So far President Trump has mostly succeeded in alienating the rest of the world and much of America with bombs, missiles, and words. The courts and many Republicans in Congress have deadlocked almost all of Trump’s agenda, to an even greater extent than they did Obama’s. Failure to raise the debt ceiling or pass a budget might go further and cause more distress than similar failures have in the past, but those earlier experiences suggest that they will not cause catastrophic or irreparable harm to the American or global economies, nor to stock markets.
If the president were able to implement any of his proposed policies, the economic damage would likely be focused in only a few places. Cutting off immigration or imports would hurt the U.S. economy more than anywhere else, with important negative effects outside the U.S. limited to a handful of countries like Mexico, Guatemala, and the Philippines. Implementation of various other threats would typically also do most of their damage in specific regions or to specific industries. One plausible interpretation of the recent turnover in the White House staff is that corporate America has regained some of its customary influence over the U.S. government. As much as many of us might consider this trading one bad thing for another, it will surely not be viewed that way by the stock market.
In this environment, and remembering that countries all over the world are experiencing healthy recoveries, international diversification of investments is a more effective protection against risk of loss than is normally the case, as some regions and types of economies are likely to continue doing quite well even if others suffer. In addition, the foreign recoveries are mostly younger than the recovery here, with more room for profit recovery and stock market appreciation. We continue to favor investments in Europe, Japan, and Emerging Markets, as well as in e-commerce. At the same time, we think that more traditional cyclical stocks are likely to do well in the younger overseas recoveries. Foreign stocks are still less expensive than U.S. stocks despite the gains they have already made, but in Europe the strong appreciation of the euro adds an additional obstacle to future profits for an American investor.
From current levels we expect quite modest returns from most of the world’s stock markets over the next 12 months, but still satisfactory and better than bonds.