Tag Archives: Forecasting returns

What Your Advisor Believes (And Why You Should Question It)

Chances are your financial advisor believes in two related intellectual theories that you should question them about.

In a recent article, the excellent columnist Brett Arends wrote about the two theories governing most financial advisors – the efficient markets hypothesis (EMH) and the capital asset pricing model (CAPM). These sound like impossibly complicated things, but they’re not. The first theory says prices are right, or nearly right all the time, and that it’s, therefore, basically impossible to beat markets. The second theory says historical asset class returns will repeat and that the more risk you take (with risk meaning volatility), the more return you will make. So, for example, stocks are very volatile, but they’ll produce the best returns — something like 10% annualized (or 6.5%-7% after inflation) – over longer periods of time.

Theoretical Problems

But the theories aren’t always right. For example, if stock prices reflect all available information, why are they so volatile, as Arends asks? It may be because that information is always incomplete, and as more information emerges prices change accordingly — and correctly. But extreme volatility may also exist because people are irrational or emotional, and substitute stories or “narratives” for more rigorous analysis or even basic common sense. The rise of the tech bubble, for instance, wasn’t an example of new information being priced in as much as it was an instance of people’s imaginations getting the better of them, andcausing them to inflate the prices of stocks that had no underlying earnings or even revenues.

More problems: why have U.S. stocks (the S&P 500 including dividends) produced a less than 6% nominal return from 2000 through November 2018? Why did they deliver nothing but dividends from the mid-1960s through the early 1980s? And if stocks are such inflation-beaters, why did the S&P 500, including dividends, return only 65 annualized in the 1970s, far underperforming that decade’s inflation?

Future Returns

The facts of the matter are that prices aren’t efficient and asset class returns may not repeat for the, say, 25-year period your retirement plan is counting on them to do so. Bonds returns, for example are easy to forecast. They generally follow the yield-to-maturity. That means a portfolio of 8-year or so domestic investment grade bonds, such as one finds in a fund tracking the Bloomberg Barclay’s US Aggregate Index, now will almost certainly deliver around 3.3%.

And for the S&P 500 Index to return 10% over the next decade, it must trade at a higher P/E ratio than it does now in addition to delivering around a 2% annual dividend payment and 4%-5% earnings-per-share growth. That’s possible, but unlikely, because U.S. stocks are trading at around 30 times their past 10-years’ worth of earnings. They’ve only been that expensive in their runs in 1929 and 2000. Although nobody can be certain, it’s more likely that P/E ratios will decline over the next decade, not increase, cutting into the 6%-7% nominal return from dividends and earnings-per-share growth. Adherents of CAPM, don’t view the world this way, and think prices can keep rising so that it’s almost a long-term investor’s birthright to achieve 10% annualized returns.

If your broker or advisor can’t respond to these objections that their assumed future returns might be off – by a lot – there’s a good possibility that they’re too dogmatic, and have swallowed academic finance without digesting it or thinking about it.

What this means for your portfolio

The problems in these theories mean your portfolio may not be set up to satisfy your financial plan. As Arends mentions in another article, for the decade from 1938 to 1948 a balanced portfolio went backwards relative to inflation. It did the same disappointing thing from 1968 through 1983. With the Federal Reserve taking us into uncharted waters and returns prospects for major asset classes so low, investors should look at cash, real estate, foreign stocks, and commodities, including gold.

None of these by themselves is foolproof. Some of them have performed well in some instances when stocks and bonds have faltered, and others have performed well at other times when stocks and bonds have faltered. The most important thing is that an advisor sensitive to how warped the current market and situation are right now may be your best defense against tepid stock and bond returns. Making sure your advisor hasn’t fallen hook, line, and sinker for the Efficient Market Hypothesis and the Capital Asset Pricing Model may be the best way for you to navigate the next decade or so in the markets.

DIY Market Forecast II – Our Forecast

A week ago we published DIY Market Forecast. This article was unique for us as we did not offer you our opinions and forecasts. Instead, the article provided data and trends but left it to the reader to create their own five-year S&P 500 forecast. This approach appears popular based on the many favorable comments we have received. Of the reader comments, quite a few have asked us what our DIY forecast is.

In a nutshell here is our forecast:

We selected to use the 3% GDP growth table. While we could have easily opted for the 1% table, we give stronger economic growth the benefit of the doubt. Keep in mind, however, if a recession occurs and growth is minus two or three percent for a year or two, a one percent average growth rate will be the better bet for the next five years. Given the probability of that scenario as this economic cycle ages, consider our range of estimates as a top end of possible outcomes.

We then chose a range of profit margins and CAPE valuations that we think are reasonable. In both instances, we believe they will slip from current levels towards longer-term averages. We do consider the possibility that profit margins and CAPE go below average but think the most likely scenario as highlighted in red, is between average and slightly below current levels. We remind that you that current levels in both instances are extreme and unlikely to continue, let alone rise from here.

The table below highlights our DIY forecast. The red shaded area is our “probable” forecast and averages to an S&P 500 forecast of 1707, with a range of 2420 to 1164. The yellow shaded area allows for the two factors to drop slightly below average which historically has been common during market corrections. The average S&P forecast in this scenario is 1172 with a range between 864 and 1512.

In prior articles, we have estimated what fair value for the S&P 500 index would be. In general, our estimates tend to converge around 50% of the current level, which would be approximately 1350.

Accordingly, the forecast detailed above is in the ballpark and further supports our conviction that investors should not take the possibility of a 50% decline lightly.  These are startling estimates, but considering the extent to which current S&P 500 levels are the product of manufactured liquidity-induced speculation as opposed to healthy organic growth, investors should keep an open mind. If we recall prior occasions in March 2000 and October 2007 when market levels were similarly over-extended, the reaction to a 50% decline projection then was met with equal skepticism.

For more on our expectations, please read Why Another 50% Correction Is Possible.

DIY Market Forecast

Quite often, our articles follow a similar format. We start with a “hook” to grab your interest and then offer a summary of what’s to come. Next comes the meat of the article, with data, graphs and a discussion that supports our view on the subject matter. Our closing summary typically encapsulates the main takeaways along with investment implications if applicable. A quote or two along the way never hurts.

In this article, we take a different tack. We present and analyze the factors that drove the bull market to record highs over the last nine years. It is these same factors that will also determine where the market will head in the coming years. However, instead of stating our opinions and giving a market forecast, we leave that to you. This approach will not only allow you to estimate the future price of the S&P 500, but importantly prove valuable in helping you understand the forces that drive market prices.

Foundations of the Bull

Valuing a stock or an index may seem complex, but there are only two factors that account for the price and its performance – estimates of a corporation’s future cash flows and the factor, or multiple, investors are willing to pay for those cash flows. While this does not occur neatly in a program or spreadsheet as the description might imply, the performance of every stock and index can be decomposed into those simple pieces.

With that in mind, we turn to the current U.S. equity bull market which started in the shadow of the financial crisis of 2008/09. The 315% rally, which might celebrate its tenth anniversary in March of 2019, is the longest uninterrupted equity expansion in modern U.S. history. Given the extended duration of this rally, it is more important than ever to look forward and not assume yesterday gains will continue tomorrow.

The following two sections look at corporate cash flows and valuation multiple trends on the S&P 500. This historical attribution analysis offers context and perspective about how those trends may or may not change going forward and ultimately what that means for the price of the index.

Cash Flows

Corporate cash flows that accrue to investors should be dissected into two components, revenues (sales) and profit margins.

Not surprisingly, corporate revenues are highly correlated with economic growth. Since 2010, aggregate revenues of the S&P 500 grew by 35.7%, while GDP grew by a similar 38.6%. The graph below shows the tight correlation. Historical observations going back decades further support this data.

Data Courtesy Bloomberg

Given the recent and longer-term correlation, it is sensible to assume that expectations for future economic growth, or GDP, are a solid proxy for future revenue growth.

The following graph of the long-term trend of GDP provides guidance for future revenue expectations. As we have written all too often, demographics, the increasing debt burden, and declining productivity growth will continue to impose a heavy and growing toll on economic growth. That does not mean there will not be periods of stronger than average growth, but we believe the 30-year trend lower is intact.

Data Courtesy Bloomberg

Revenue is only half of the cash flow story. Net earnings, which is what investors are ultimately paying for, account for all of the expenses required to produce revenue. Net earnings as a percent of revenues, better known as profit margin, is the common metric used to express this.

Aggregate profit margins historically vacillate above and below the historical average, but they have always been mean reverting.  To wit, here is the wisdom of an investing legend:

The following graphs provide historical context on profit margins over the last two and seven decades respectively.

Data Courtesy Bloomberg and St. Louis Federal Reserve

In both graphs, the profit margin post-financial crisis is at or near all-time highs and has failed to regress to the mean despite the wisdom of Jeremy Grantham.

While not an extensive list, consider the following chief factors responsible for elevated profit margins:

  • Lowest interest rates in recorded history
  • Minimal wage growth
  • Low input costs
  • Unchanged shipping, trucking, and freight costs

Review the following table of major corporate expenses to understand current margins and formulate expectations for future ones. The table compares some key proxies for expenses over the last year versus the prior three years.

Data Courtesy Bloomberg

Two aspects of corporate expenses omitted from the table are taxes and tariffs. Recent tax reform boosted margins and helped more than offset the negative effects of the rise in costs shown above. The consequences of the on-going “trade war” are yet to be seen, but they are likely negative.

Valuation Multiple

Since 1877 there are 1654 monthly measurements of Cyclically Adjusted Price -to- Earnings (CAPE 10). Of these 82, only about 5%, have been the same or greater than current CAPE levels (30.5). Other than a few instances over the last two years and two others which occurred in 1929, the rest occurred during the late 1990’s tech boom. The graph below charts the percentage of time the market has traded at various ranges of CAPE levels.

Data Courtesy Shiller

Valuations are a function of investor sentiment. When sentiment is exuberant, as it has been recently, investors are willing to pay more for a series of cash flows in expectations that revenue and earnings will rise at a heady rate in the future. Conversely, when investors are concerned about future earnings and economic growth, valuations tend to decline.

Looking back, there are many factors that drove investors to pay a higher multiple for cash flows over the last ten years. Consider a few:

Historically Low-interest rates

  • Lower discounting rates made the value of future earnings higher.
  • Resulted in a push towards higher returning, riskier, longer duration securities like equities and long maturity bonds.

Heavy Monetary Stimulus

  • Record low-interest rates and burgeoning central bank holdings of financial assets here and abroad.

Corporate Share Repurchases

  • Since 2013, S&P 500 companies have annually bought back 3% of their outstanding shares in aggregate.

Margin debt

  • Since 2012, net credit balances have been larger than those seen before the market drawdowns of 2000 and 2008.
  • Currently, balances are 3x larger than any peak seen in at least the last 36 years.

The proliferation of passive investment strategies which tend to ignore valuations

The expansion of corporate leverage to record highs nominally and as a percent of GDP

To that list we submit one important factor – inflation. The following graph demonstrates that valuations have only been well above the norm during periods when annual inflation is running between one and four percent. Outside of the “sweet spot,” CAPE valuations tend to peak about 25-30% lower than current levels.

Data Courtesy Shiller

How The Market Got Here

With an understanding of the factors that account for price performance since 2010, we now turn to the graphs below which decompose the gains of the last eight years into the components: revenue growth, profit margin expansion, and valuation expansion.

As shown, durable organic growth only accounted for 26.96% of the gains in the S&P 500 index since 2010. In other words, without multiple and margin expansion, the S&P 500 would stand at 1587, a far cry from the current 2790.

DIY- Forecasting the S&P 500

Now we let you forecast where the S&P 500 might be headed over the next five years based on your expectations for revenue, profit margins, and valuations. To formulate a personalized forecast, you will need to complete a two-step process. First, answer the three questions below. Next, feed your answers into one of three tables provided below. The result will be your forecast.  To help with answering questions two and three below, we provide current levels along with minimum, average, and maximum historical levels. We also urge you to go back and consider the graphs and factors that drove recent trends.

  1. How much will GDP, and therefore corporate revenues grow over the next five years?
  2. Will margins stay at current levels, expand further or contract back to or below historical norms?
  3. Will valuations stay at current levels, expand further or contract back to or below historical norms?

Once the questions are answered, the data can be used to generate a forecast. The example below offers a guide to the process. In the example shown, the question responses are that GDP growth will average 1% a year for the next five years, and that profit margins and valuations will stay the same for five years. The resulting output of 2838, as highlighted, is the expected value of the S&P 500 in five years.

Choosing among the three tables below is based on your forecast for future economic growth: low 1%, average 3% or high 5%. Once you select the appropriate table, find your expected profit margin for five years from now on the horizontal axis at the top and your estimate for CAPE valuations on the vertical axis on the left. The estimate for the future level of the S&P 500 lies at the intersection of the two forecasts.


The framework described above can be used for something as simple as finding one answer as we did in the example, but it also allows an investor to conduct scenario analysis to arrive at a range of possibilities. By assigning various probabilities to one, two or all variables, one can calculate a weighted average outcome. For example, hold CAPE and Profit Margin constant and assign a 30% probability to 1% GDP growth, 60% probability to 3% growth and 10% probability to 5% growth. Repeating that process produces a range of answers which could effectively be used to gauge risk versus return.

Although relatively simple in its construction, the ability to customize your forecast and apply multiple scenarios is a powerful risk management tool.


Projections Are Imprecise, But Not Useless

It seems banal to say, but financial planning requires return projections or estimates. If you’re saving money for a goal like retirement, sending a child to college, buying a home, or taking a vacation, you need to know three things (at least) — how much to save, how much of a return that savings will earn, and the distance to the goal. Without any of those three things, there can’t be a plan. And all of this doesn’t take into consideration your own temperament or how you react to volatility and the potential for permanent loss.

Of course, the return projection won’t be precise if any part of the capital is being invested in stocks. It’s not easy to forecast how much stocks will return over a given time, and the shorter the distance to the goal the more unpredictable and random stock returns are. And that’s one reason stocks shouldn’t be used for short-term financial goals. They can do virtually anything over one- or two-year periods of time. However, over longer time frames — 7-10 years or more – forecasts can be more reasonable, though still not precise. But one is never absolved from making an estimate or a range of estimates.

Unfortunately, some prominent financial planners, who often double as pundits, denigrate all forms of forecasting. Financial planner and sometimes New York Times columnist Carl Richards recently tweeted that the only thing we know about projections is that they are wrong. He applied the hashtag “projectionfreeplanning,” which, of course, is an oxymoron. There’s no such thing as financial planning or projecting a future value of an investment, after all, without a return estimate.

Similarly, prominent advisor and pundit, Barry Ritholtz, has argued that forecasting is “almost useless” and that we “stink at it.” Ritholtz says assertions like “stocks tend to go higher” are vague enough to be exempted from his critique, but “The Dow will hit 25,000 by the second quarter of 2018” aren’t. What’s frustrating about his writings is that they don’t say anything about the ordinary forecasting of long term (say, 10- or 20-year) returns financial advisors must do to satisfy future value calculations for their clients.

The pundits like to say that the Shiller PE isn’t a valid metric anymore because it’s been well over its long-term average – around 16.5 – for over 25 years. But the annualized return of the S&P 500 Index, including dividends has been 5.4% from 2000 through 2017, and the Shiller PE was over 40 in 2000. In other words, in 2000, it did a good job of telling investors future returns would likely be tepid. Moreover, that return has depended on the dazzling 15% return of the index since the financial crisis that has driven the Shiller PE up again to the low 30s. And, as Rob Arnott has said, we can have a reasonable argument about whether the new normal for the Shiller PE is 20 or 22, but not whether it’s 30.

Advisors are rebelling so much against forecasting because they don’t like to deliver bad news to clients. Bad news can be bad for business. Clients will choose the advisor with the highest future returns projections because they want to be soothed. But delivering optimism when it’s unwarranted can lead to projections that border on malpractice on the part of the advisor. Investment professionals usually know this when it comes to bonds. It’s difficult for a bond or a portfolio of bonds to return more than its yield-to-maturity. However, when it comes to stocks, advisors often resort to using the longest term return numbers they can find. Those usually come from Ibbotson Associates, now a division of Morningstar, which popularized a stock market return chart dating from 1926. But most investors aren’t investing for a century, and there have been enough 10- and 20-year periods of poor returns to give investors and advisors pause. More importantly, those periods are associated with high starting valuations.

And now it has become clear that estimating, say, 7%, for a balanced portfolio over the next decade is a stretch. Bonds are likely to deliver less than 4%, and that means stocks will have to deliver more than 8.5%. Advisors are becoming increasingly pessimistic about that possibility for stocks, but they aren’t responding by expressing that pessimism clearly Instead, they are responding by bashing forecasting altogether. It’s not the most mature response, but the possibility of losing clients because of poor forecasts has its bad effects. And it’s true that the Shiller PE — or any other valuation metric — isn’t perfect in forecasting returns, but it can’t be prudent to count on stocks delivering 8.5% for the next decade with a starting Shiller PE in the low 30s.

Investors should question their advisors about returns, because they need to know how much money their current savings rate will leave them with to spend in retirement. The return assumption is just that — an assumption — but that means investors can ask for a range of assumptions to see what different returns will deliver. That doesn’t mean the optimistic assumptions are truer ones though, but it helps investors understand what they’re up against without being precise. And that is far from useless. The second thing investors should do is something financial journalist Jazon Zweig discussed in an old column – they should ask their advisors how much return the advisor would deliver to the client in a “total return swap,” whereby the client hypothetically hands over their entire portfolio and gets an annualized return on that portfolio in exchange. There’s no better way to put the screws to your advisor when it comes to getting his or her opinion on future returns.