No, Jesse Had It Right: Owning Stocks Today Has An Unattractive Risk/Reward Profile
My rebuttal to Terrier’s rebuttal. Terrier seems to believe that timing the market is not possible, but beating the market through stock picking is very much possible. Many bulls look at one market over one long stretch of time and believe they’re all clear for 10+ year periods… nope. Terrier Investing posted a rebuttal this morning to Jesse Felder’s original piece : “Owning Stocks Today is Risking Dollars to Make Pennies.” Terrier makes three points in his rebuttal. To quote: Well, according to Jesse, it means stocks are so wildly overvalued that your potential return over the next ten years is miniscule, and your potential downside is massive. I posit this is: A) alarmist and statistically inaccurate; B) overly narrow in its definition of risk; and C) treats “stocks” as some monolithic entity” Each of Terrier’s points are problematic; I’ll handle them one by one. Before I do, however, let me say that Terrier makes many sensible claims in his rebuttal. I dispute his line of reasoning here, mainly because he uses three arguments that I think undergird many bulls’ logic, whether they realize it or not. Someone like Terrier who explicitly makes assumptions is in my view on much firmer soil than the many bulls who are implicitly making the identical assumptions. If Terrier sees reason to modify his explicit a priori, he can. Bulls that are actually sheep have no such explicit framework against which they can base a reasonable shift to their investment thesis. With those disclaimers out of the way, I will now address the problems that I see with Terrier’s arguments. A) Alarmist and statistically inaccurate (sorry to quote so much of Terrier’s piece, but I want to address what he DID say, not what he didn’t): What is the actual likelihood of stocks resulting in a significantly negative 10-year return? Here’s a link to a nice document providing this data from 1926 through 2013 in both tabular and graphical format. Summarily, there were only a very few rolling 10-year periods when investing in the S&P 500 would have resulted in losses in nominal terms. Specifically, you would have had to invest right before the Great Depression or in the late 1990s – two of the larger bubbles of all time. Looking at one market over one stretch of time, even a long stretch, does not give you a statistically robust sense of what that market can do over any 7-15 year timeframe. I’ll grant you that it’s better than a sharp stick in the eye, but the data can easily mislead. I wonder what the German stock market would have looked like over the first half of the twentieth century. After enduring two world wars, a bout of hyperinflation, and political dismemberment, I don’t believe that German stocks performed too well over that meaningfully long timeframe. The German stock market was at the time (and still is) a well-developed market. I wonder what fraction of those 10-year periods had sizeable losses. Whoever said that can’t be us? From an Investopedia article on history of stocks and bonds: At the same time, many other economies suffered great losses. For example, according to Phillipe Jorion and William N. Goetzmann in their article “Global Stock Markets In The Twentieth Century” (1999), the Japanese stock market saw a 95% decline in real returns between 1944 and 1949. The German market also suffered devastating losses. In this context, the U.S. market’s success seems to be an exception, which the previous lack of data for other countries may have obscured. (emphasis added) Japan 1986 to present?…let’s not look there I’m guessing. (click to enlarge) How about the US stock market from 1891-1974? There were many poor return stretches over that time frame, especially when viewed on a real return basis. That’s a long stretch in our own market; how do the total return statistics bear out? While I’ll grant that the percentage of positive ten-year returns would likely still be high, the final results would be substantially more lackluster, particularly for investors who did not reinvest all of the dividends over the entire horizon with no tax implications. In fact, depending on your starting and ending points, you can find periods of negative real returns over a fifty-year time frame if you don’t include complete dividend reinvestment over the entire 50+ horizon. To see that this is the case, check out Political Calculation’s S&P calculator . Enter some periods that end in 1974 or 1983 for instance. I’m not trying to cherry pick here; I am demonstrating that there certainly are periods for even the longest of practical time horizons where equity returns are quite unattractive. There are two other, larger reasons why past may not be prologue for S&P returns. And I’ll address these points alongside Terrier’s point B: B: Risk as volatility, not as permanent loss of capital Moreover, there is more than one definition of “risking dollars” – assuming you have a ten-year or greater time horizon and need to invest to fund long-term liabilities (kids’ college funds, retirement, etc.), then earning near-zero returns by investing exclusively in bonds is just as much of a risk as potential volatility from investing in stocks. Risk, in this context, means you won’t meet your financial goals – and if you don’t invest in any stocks, it’s very hard to see how you will generate sufficient returns with yields on fixed income where they are. Some clarification here first. Terrier goes on to say that he believes that the market as a whole is on the expensive side (which leads to his point C), so he’s not some brainless stock market cheerleader. To that same end, Felder never explicitly says that nobody should have any equity exposure. (As for me, I have plenty of equity exposure: I’m short SPX.) Terrier’s second point essentially makes an assumption: risk as volatility vs. risk as probability of permanent capital loss. If risk is merely volatility – stocks whipping around for short and maybe even violent bursts, only to recover over a reasonably quick timeframe and make new highs – then I believe that he is correct. In his defense, he doesn’t suggest going “all-in” on equities, and even recommends having a decent cash pile. The issue is that Terrier’s problematic analysis from his first point (stocks rarely have negative nominal 10-year returns) leads him to the next conclusion that equity risk is actually only volatility, not capital impairment. This is where Terrier and I truly part company. Many long-only investors believe that strong long-run SPX returns happen mostly as a simple function of time; they’re somehow owed these returns for weathering volatility. I find it amusing that these same long-only bulls don’t feel like Brazilian investors are owed strong long-run returns, or that Greek or Russian or Japanese or South African equity investors are owed long-run returns. This amounts to a personally dangerous form of financial jingoism. Let me make it clear: IF sustained poor equity returns can happen to Brazil (the world’s seventh largest economy), then they can happen for the US. See, investors today aren’t looking at the Greek market and shrugging it off as a temporary bout of volatility. Ditto the other markets mentioned above. Investors correctly see these declines for what they are: semi-permanent capital loss. That is to say that even a strong bounce and even full dividend reinvestment will not bring a buy-and-hold index investor who purchased in, say 2010, back to even for years to come. Bears like Felder and myself believe that S&P balance sheets, investor margin positioning, GDP growth trends, and equity valuations in light of a slowing global economy put the S&P 500 at risk of a vigorous fall that will NOT be recovered anytime soon. (click to enlarge) (click to enlarge) Source: FactSet Why should we expect S&P returns that approximate history when a) GDP growth (global or US) is nothing like what it has been in the past, b) corporate balance sheets are not very healthy and c) valuations for the broad market are MORE expensive for almost every decile than at the March 2000 peak? To conclude, Terrier states: Risk, in this context, means you won’t meet your financial goals – and if you don’t invest in any stocks, it’s very hard to see how you will generate sufficient returns with yields on fixed income where they are. Well, what if the S&P falls – a LOT – and does not recover? Meeting one’s financial goals goes from being difficult to completely impossible. I believe such an outcome needs to be given a very meaningful weight. Terrier’s last point is that we don’t have a stock market, but a market of stocks: Finally, point C: I think it’s unfair to treat “stocks” as a monolithic entity – as if you either own the S&P 500 (NYSEARCA: SPY ) or you do not, and there’s no other alternative. Even if you believe the market as a whole is overvalued, like I do, that doesn’t mean every single component of the market is overvalued. Terrier goes on to say that one can do research and find a basket of stocks that will beat the market. Which is basically saying that Terrier doesn’t believe that investors can beat the market via market timing (“Not owning the market is risking dollars to make pennies”), but that they can beat the market through security selection. I completely disagree. Look at all the “smart beta” ETFs and actively managed mutual funds that are essentially continuously fully invested. How many of those pros beat the market? Not too many. I’m not saying that it cannot be done, but I see no reason – whatsoever – why market outperformance through the security selection channel is so much easier to consistently achieve than market outperformance via the market timing channel. But my objection to Terrier’s point C goes well beyond this first point: In 2013, the most heavily-shorted stocks were some of the best performers . It tends to be sophisticated investors that short companies. Full disclosure: I have never in my life shorted an individual name, and so I claim absolutely no expertise on this process. These securities specialists had their you-know-what’s handed to them, because it was a bad idea to be short any stock in the S&P during 2013. Similarly, it was a bad idea to be long any stock in the S&P between March 2008-March 2009. When “the market” gets crazy (up or down), security selection absolutely will not save you… period. At that point, the macro takes over, and the micro gets buried. That doesn’t mean that security selection cannot help you (assuming that you can in fact do it AND stick to your discipline): better to lose 33% than 40% or 60% instead of 75%… but you still won’t be happy with your strongly negative returns. In conclusion, Terrier states in his point A (in context of negative 10-year returns): Specifically, you would have had to invest right before the Great Depression or in the late 1990s – two of the larger bubbles of all time. My stance, and I believe Felder’s as well (though I’ll let him speak for himself), is precisely that today’s market IS one of those great bubbles. James Paulsen of Wells Capital Management produced the chart below to compare P/Es of the S&P for each 5-percentile increment for year-end 2014 vs. June 2000. The overvaluation of the broad markets is far more severe than it was in 2000, and so when the bottom falls out, there may not be too many great places to hide from the merciless reaping that ensues. Permanent capital impairment from any and all long US equity exposure needs to be treated not as a fringe case, but as THE base case. In that world, long investors really indeed are risking dollars that they won’t recover for years in order to pick up those juicy 3-5% yields or hope for the continuation of a stretched and tired bull.