Tuesday, March 01, 2011

The Outlook for Equity Prices: It’s probably worse than you think

With recent tremors in global markets queued up by concerns about oil supplies, sovereign debt levels, government deficits and the tax increases and spending cuts required to get them under control, I think it is about time to look at Equity Prices in a more holistic – and to my mind, clearer - way. To strip away the noise, and think about what truly long term returns are likely to be.

The proximate cause of this exercise was an interview given by James Glassman, contributing editor to Kiplinger’s Personal Finance and author of the notoriously badly timed “Dow 36,000”, to Henry Blodgett of Tech Ticker. Glassman, it should be noted, is now peddling a new book – encouraging investors (individual investors) to reduce their overall stock exposure and instead to hedge against loss. This only took 10 years of dismal equity returns for him to acknowledge that markets sometimes go down and for long periods. I admit that I have not read either book, though I have read many of his columns over the years.

Blodgett and Glassman discussed the effect of a lower volatility portfolio on long term returns. Both used a 10% number as the long term compounded (geometric) average return to investors from equities. Glassman’s argument was that in saving for retirement, you should sacrifice some upside to ensure that you don’t experience a big, catastrophic drop – in his words “once of the worst things you can do is reach for extra return”. (I agree with this view. Warren Buffett also had some choice words on this in his annual letter). Blodgett countered – but what of the people who *need* 10% to make “the number” they need to retire? Glassman essentially said, you should settle for a lower return, which he estimated at around 8%. When challenged by Blodgett that this might also be optimistic, he conceded that 5-6% returns might be possible if things were “really bad”. Unsaid by both was that with lower investment return expectations, that investors must instead adjust either saving rates (up), working life (longer with a shorter retirement), or their accept a lower standard of living in retirement.

Glassman and Blodgett were talking about a version of this chart, which I confess had a huge impact on me as an undergraduate in the 1990s, and got me focused on saving and investing. But consider how different the chart looks if returns are much lower than 10%, or even Glassman’s 8%. The difference between 8% and 10% is 60% of the total account value for our “young and done” investor, and *only*45% of the account value for our later diligent. What if returns fall to 6%? That $1 million portfolio might now buy our “young and done” investor a small condo in Miami (at post bubble prices). Even our diligent saver has lost 67% of his projected nest egg.

Given the impact that different rates of return have on expected retirement assets and living standards, it makes sense to have a basis for expectation. All of the numbers I see are “rear view” projections. They make assumptions that the future will be like the past.

Instead, I would like to suggest that we face forward, and think about what is probable – not in the near term, but over the long term. The short answer is – we aren’t going to get historical returns.
Let’s start by unpacking the components of equity prices. Stock prices have two drivers – earnings (dividends, or free cash flows, technically, but over an infinite time horizon these converge), and investors’ appetite for earnings (that is the price investors are willing to pay for those earnings).
Earnings have two components, the current rate, and the rate of earnings growth over an infinite time horizon. This is what one might call “fundamental” or “intrinsic” value – the actual money the business would throw off.

Investor’s appetites (valuation) have three components - interest rates, volatility (risk) and expectations of earnings growth.
I want to begin our investigation by thinking earnings and – more importantly earnings growth. I start with a statement – over the long term, earnings cannot grow materially faster than GDP. This might shock you, but it is the GDP identity.

Without getting too technical, there are three methods of calculating GDP: the output method, the income method and the expenditure method. In theory, they should produce the same result. In actual practice, they differ slightly, but are close enough for our purposes. Since we are discussing income (earnings) we will apply the Income Method, which states:

GDP= Employee Compensation + Corporate Operating Surplus + Gross Mixed Income + Taxes – (Subsidies on production and net Imports)

This identity says that Employee net income, plus corporate net income, plus the income of unincorporated businesses, plus taxes (government income) less subsidies (which accrue to one of the other income sources) is the national income. This makes sense. From this identity it follows that for corporate profits to be larger than GDP, either employee compensation, proprietor income or taxes (less subsidies and imports) would have to be negative. In practice, no one component would ever be as large as GDP, even in a socialist country, because people have to eat.

If one component in growing faster than GDP, it is gaining share of the national income. Moreover, that over very long (infinite) time horizons, no component can grow significantly faster than the whole, since it would over time gain enough share to approach 100% of GDP, and either drive up GDP growth with it, or slow to a rate approximating GDP. Over shorter than infinite time horizons, however, individual components can grow faster than the whole, and gain share of the overall pie.

I suggest to you that people who throw 8, 9, 10, 11% growth as the “long run” return of equities are projecting a future in which corporate earnings permanently grow significantly faster than the underlying economy. If you ask these same people to estimate GDP growth you will usually get a number of either 3% or 6% depending on whether you mean nominal or real GDP. (Since earnings are a nominal measure nominal GDP is what counts here). I also suggest that they are likely to be disappointed.

It is true that looking backward over long time horizons, equities have earned 9-10% or so. So how has the market delivered this return “over the long term”? Well, consider two factors. First, earnings may have grown faster than GDP over measurable historical periods. Our information about markets and economics are really quite recent developments. The stock company is also a recent development. The first joint stock company was invented in 1600 when Elizabeth I chartered the East India Company. Stock exchanges are even newer, with the London Stock Exchange having its origins in 1698. Most of the “stock” traded was government paper, not equities and their profits were tiny compared with overall economic activity. Modern industrial firms were really created in the 19th century and have steadily increased their share of economic activity at the outset. Most people were unincorporated small proprietors working at subsistence farming, or small shopkeepers. Wartime cartelization and unionization meant that in the 1950s workers held a much larger share of national income So, looking BACKWARD, corporate profits have grown faster than the economy overall, and have gained share of output, but projecting that growth onto the future is asking for trouble because corporate profits are no longer working off of a low base.

Second - there is valuation: equity investors have simply been willing to pay more for earnings over time. This is particularly true in the period after 1994. A variety of factors contribute to this, including favorable macro conditions, which have driven higher all asset prices (as measured by earnings or interest yields) - and have reduced the realm of attractive alternatives, recent experience which has been for equity prices to increase much faster than overall economic activity and reduce the perceived risk of equities, wider equity ownership (with most of the growth coming from less sophisticated investors, who may be less scrutinizing of changes in underlying fundamentals), and the possibility that investors in early market stages were excessively cautious - that is, that equity prices were simply too low relative to the risks (thus some of the gains were simply a correction of undervaluation).

In short, when we look at equity returns, I believe we are taking as normative something that is in fact a historical anomaly. Like any new and successful technology or approach, corporations experienced significant growth, taking share from other market participants (otherwise, it would have died out. It is tantamount to projecting iPad sales, or Google revenue out into the future based on the first few years after introduction.

Even exceptional cases of long term performance, e.g. Berkshire Hathaway, have experienced diminished returns over time as the base on which high growth has to be earned, expands. At some point, corporation earnings will slow, because attractive investment opportunities will decline, because workers will gain more pricing power, or because governments will raise taxes, or all three. What is certain is that to earn the kind of returns investors expect, they must believe that corporate profits will come to represent 20-30% of economic activity within the course of an adult lifetime (up from 10% now). This is highly improbable, and should therefore not be used as a “conservative” or “reasonable” forecast of returns.

Moreover, the GDP range I have used above is also historic and may also be anomalous.

6% nominal growth seems high, given that inflation is running well below 3% and the primary inputs of real output – labor (in the form of hours worked) and output per hour, both look to be factors that will grow slower in the future than in the past. More people are working in the US than practically ever before. This is mostly due to demography – women’s labor force participation has increased dramatically, the dependency ratio (the number of people divided by the number who work) has hit a low and is projected to rise as Boomers retire. Labor hours worked is set to rise slowly from currently suppressed levels. Productivity growth may continue at 1.5% but with hours worked increasing only 0.5% - 1% per year, real growth will likely be closer to 2-2.5%. If inflation remains around 2%, then nominal output will grow 4-5%, not 6%, and we will see a corresponding decline in the rate of earnings growth.

Consider though that if the variation (the volatility) of the returns stays constant regardless of the mean expected return that we should expect to see many more years with negative stock returns. The standard deviation of returns is around 15%, which means that if mean returns are 11% most years will see positive stock markets. If the mean is 4%, however, there will likely be many more years where returns are negative. Worse, stock returns are heteroskedastic, which means that volatility changes with the time horizon (though I am not sure if it increases or decreases – this is important).

I think it more likely that taxes will increase as a share of the economy, and with it, subsidies. If unemployment remains high, it is possible that corporate taxes will rise, reducing corporate income (though most policy advocates on both the left and the right seem to want to reduce corporate income taxes, which would INCREASE income as a share of economic activity, all else equal). One positive factor, however, is that net imports may decline, particularly if consumption declines as well. Finally, over the next few years, unemployment is likely to improve, and with it consumer wages. Already, we see wages rising in both Eastern Europe and the Far East, places which have supplied the low cost labor that has enabled companies to keep a lid on wages and benefits and expand corporate share of income. With rising wages, overseas and slowly falling unemployment, it seems quite reasonable to believe that wages and benefits will rise, eroding corporate profitability. Thus it seems possible that corporate income could as likely fall as a share of output as rise.

With corporate profits already at high levels, and unlikely to sustain growth faster than the economy, and the economy likely to grow more slowly than in the past, the only hope we have for Glassman’s 8% returns seems to lie in higher valuations. Sadly, the news here is also rather dismal.

Stocks rocket from periods of high valuation to periods of low valuation. This table from John Mauldin, shows what happens when you invest in periods of high valuation – you tend to get returns of less than 4%. More accurately, attempting to live of 4% of your money leaves you in real risk of winding up bankrupt over 30 years.

Today, sadly, our valuation (measured over the 10 year business cycle) is high. Several factors influence this, interest rates (at which investors discount future growth) are quite low. Low rates lead to high valuation. Faced with low yields in other asset classes, such as bonds or cash, investors are turning to risk assets. Similarly, growth expectations are high, because people are still projecting the past onto the future. Further, equity valuation has been high for so long, people have come to see it as normal. Indeed, stocks may look cheap when compared to the astronomical valuations of the late 1990s.

Pity a poor Japanese who at age 50 has been working since 1985. If he has been a diligent saver, putting his money into the stock market in hope of high returns, he has watched his investments wither away as slow earnings growth and compressed valuations have left the market down 75% from its high, more than 20 years after the peak! Real estate returns are similarly bad. That is why he instead purchases assets with no risk of capital loss – government bonds – and earns essentially zero. He is just happy to know that he can get his money back someday.

What we know is that equities tend to rotate from high valuation to low valuation and back again. Despite coming down from very high levels 10 years ago, we are not near to a low valuation level. True, at market lows in 2009 we were near the normal valuation level. Beyond 2012, ten year corporate earnings will be able to escape the poor performance in the 2000-2002 period, and so be a bit higher.

But interest rates will also likely rise, and that will give competition to increasingly risk averse Boomers. Moreover, it will raise the discount rate on stocks. If valuations fall, even as earnings rise, the perception of risk in equities as a class may rise, and the return demanded from investors may rise as well. This will feed into yet lower valuations. But this is again to suggest what might happen to stocks over the next ten years.
What I am saying is that one has to be very selective in the stocks one picks – stocks as an asset class will have low returns, likely in the 4-6% range over time. This may come from a permanently 2nd gear growth rate in prices, but history suggests that stock returns are volatile, thus returns are likely to instead go through a period of negative returns. These may be sufficiently large to reduce valuation to such a point that returns from that point forward can grow faster than economic growth for an extended period. This is not the market we face.

In short, I believe we are looking at 4-6% returns for equities going forward. This result is the combination of an expectation of lower earnings growth coupled with a fall in valuation.

This will have profound impact on the level of savings required to have the standard of living in retirement that people expect. Paradoxically, if enough people attempt to save more – we will push valuations higher in the short term, but subsequent returns yet lower later, as underlying yields could fall to Japanese levels. Pity the guy who has to select from an investment universe in which all values are sky high. In the short term, of course, returns could easily exceed this level. But they could just as likely fall.

Such low rates mean that one has to question whether one is inclined to subject oneself to the volatility of stock returns. These will be negative far more often than investors have experienced over the past 30 years. I continue to recommend solid businesses that can pay steady dividends, fortified by a low payout rate. Stocks yielding 2-4% offer current yields not materially different from the expected return on the overall market with significantly less volatility that the overall market. These stocks will also hold value better during a decline, providing the investor with the option to acquire more volatile stocks at a moment when that volatility results in low valuation. In the event of inflation, they should also have pricing power and the ability to raise yields to protect investors.

Happily, many solid dividend paying stocks, such as CL, INTC, WMT, JNJ and other solid brand-name firms are relatively inexpensive. There are also a host of smaller firms which may be attractive takeover targets, particularly in a world where investors expect strong earnings growth but organic returns aren’t in the cards. Of course, once you sell, you face the same unattractive investment universe everyone else has, but at least you have some cash.

Finally, in a world of low returns, the optionality of cash should be attractive to most investors. You are unlikely to lose big sitting out (yes, I KNOW that stocks have rallied 95% since March 2009. I wish that I had bought more in the Q1 of 2009, but we are in March 2011 now, and have to look at the situation we now face). I confess that given relatively high valuations, I believe we remain in a secular bear market, and that March 2009 lows will be back, though it may be a few years yet.

Protect your downside and adjust your expectations and your plan.

No comments:

Post a Comment