The S&P 500 is the World’s Largest Momentum Strategy

In many ways the stock market makes no sense. You would assume that half of all stocks would outperform a market index while the other half would underperform. Then all you would have to do is pick from the top half and avoid the bottom half, make massive amounts of money and go buy an island somewhere.

Unfortunately, the stock market doesn’t follow a normal bell-shaped curve. Active investing may be a zero-sum game, but picking individual stocks is not. It doesn’t work out that half of all stocks outperform and half of all stocks underperform. There are huge tails when you look at the extreme over- and under-performers in the market.

This is exactly what Eric Crittenden at Longboard Funds did in a recent blog post. This visualization of the winners and losers in the stock market since 1989 is fascinating:

Screen Shot 2016-05-18 at 12.40.35 PM

Here are the stats behind this excellent chart:

We analyzed 14,455 active stocks between 1989 and 2015, identifying the best performing stocks on both an annualized return and total return basis.

Looking at total returns of individual stocks, 1,120 stocks (7.7% of all active stocks) outperformed the S&P 500 Index by at least 500% during their lifetimes. Likewise, 976 stocks (6.8% of all active stocks) lagged the S&P 500 by at least 500%. The remaining 12,404 stocks performed above, at or below the same level as the S&P 500.

I think this data may come as a surprise to many. Over 40% of all stocks during this period ended up with a negative return. And the winners in the stock market are a much smaller group that account for the entire gain over this period:

Screen Shot 2016-05-18 at 12.40.43 PM

These numbers are staggering — 20% of all stocks have accounted for the entire gain over this time frame, meaning the remaining 80% have collectively given investors no return. From 1989-2015 the S&P 500 was up almost 1200% in total. The majority of that gain came from a small number of stocks while the rest were more or less worthless.

There are a few different ways you could choose to look at this data:

  1. Try to avoid the losers and pick the winners.
  2. Own the whole thing and benefit from the huge impact that the winners give you.
  3. The S&P 500 is basically a momentum strategy.

Momentum investors cut their losers and let their winners run. In a roundabout way, that’s exactly what the stock market has done over time. My friend Lawrence Hamtil recently shared a stat on his blog that illustrates part of this equation:

Compared to 2004, more than a fourth of the companies then in the S&P 500 have “been acquired, taken private, or gone bankrupt,” though there have been hundreds of IPOs since.

The very nature of a market-cap weighted index means that the performance will be driven by something of a winners-take-all scenario. The cream rises to the top and the losers tend to fall by the wayside.

Now, is it possible to avoid the losers and only invest in the winners? Sure, anything is possible. Is it probable? Technically, based on this data it’s a low probability strategy. Obviously, many of the companies in that 80% group were smaller, more risky companies that you would expect to fail. Not every business is meant to succeed over time. And there are ways to screen for quality and valuation that can give you better odds of success at avoiding the worst of this group.

But one of the reasons the S&P 500 is so hard to beat is because it has a built-in mechanism to let the winners ride by the way it’s constructed. By our very nature we have a tendency to hold onto our losers and sell our winners too quickly. The S&P 500 or any systematic strategy doesn’t have these same issues holding them back. Can you imagine owning stocks like Apple, Amazon, Google or Facebook as they continued to rise over the years? How many of us would have had the ability to hang on as those gains compounded and the market caps kept rising?

Buying during a bear market is probably one of the most difficult things an investor can do, but staying invested as stocks rise may be a close second. There are constant temptations to “take money off the table” or “de-risk your holdings” after a nice run-up. It’s not easy to allow compounding to occur without getting in the way and screwing things up.

I like to say that index funds are nothing special — they’re systematic, disciplined, rebalanced occasionally, transparent, low-turnover, low-cost, and low-maintenance. Probably the biggest benefit an index has over a human is the fact that it has a disciplined process by default. It’s hard to compete in the markets if you’re not disciplined.

Sources:
Defense Wins Championships (Longboard)
You Can’t Have the Creative Without the Destruction (Fortune Financial)

Further Reading:
Why Momentum Investing Works

 

Download PDF

Full Disclosure: Nothing on this site should ever be considered to be advice, research or an invitation to buy or sell any securities, please see my Terms & Conditions page for a full disclaimer.

  • W_D_1

    Thank you for a wonderful blog. Just starting to learn about this stuff and I really appreciate the advice. Can’t wait for your book to arrive. Greetings from Toronto,
    – Will Dubinski

    • Good article. Interesting fact that over 40% of the traded stocks ended up with negative returns. While doing our own research the returns distribution for the S&P 500 gives us the number of 52.39% positive up days from 1962 – now. We also have a negatively skewed distribution, which means a greater chance of extremely negative outcomes. When doing the math one can come to the conclusion that being on the short side of the markets seems to be the easier way to the promised land.

  • Eric Weigel

    Ben – looking at the S&P 500 (or any other cap weighted index) as a momentum play benefits investors on the way up (as the best performers increase their index weights) but hurts tremendously in a down market such as 2001-2002 when the index was loaded with former high flyers on their way to much lower valuations (CSCO, AOL, MSFT, INTC are examples of this). One of the lessons I learned as a PM was the necessity to lower tracking error on the way up (ride the wave) and actually increase the tracking error on the way down (being more different). It is a bit counter intuitive for portfolio managers to think this way.

    • Ben

      True, momentum can work in both directions.

    • John Richards

      Your comment feels similar to some things that Ken Fisher has written that despite seeming simplistic when I read them (expressed in a less defined way than you have above), have been very influential to my thinking process, and returns. I’m just curious how you decide that the future direction is down?

      • Eric Weigel

        John – generically speaking there are two ways to figure out the prospective direction of markets. One is use a multi-factor model that employs concepts such as valuation, profitability, growth and momentum (a TAA model). The other is to simply use a momentum rule such as what Meb Faber has popularized. No measure is perfect but we employ such a system along with our assessment of investor risk aversion (separate model) to figure out our target tracking error to the benchmark. Simplistically if the trend is up lower your tracking error. If the trend is down increase your tracking error (typically by holding a bit of cash and lower volatility/lower valuation stocks). I hope this helps.

        • John Richards

          Eric, thanks for the follow-up! Very helpful, it connects the dots on some of the reading I’ve done.

          It took me years to finally derive a static 401K portfolio I can live with, and to settle on a rebalancing plan. It’s asset-class based, no access to individual equities. Since then I’ve been looking at what you can do (besides owning LT Treauries or Gold – my circle of competence excludes Options) to limit the downside exposure during large drawdowns. I methodically sold equities in 2008 and bought back in 2009 (1st time I deviated from Buy & Hold), so good results but lousy process – I had some level of conviction, but there was a huge element of luck which I’d like to nibble away at.

          I have since explored variations on the ideas Faber expressed, knowing that it trades one set of risks for another, and found one promising approach based on my specific portfolio modeled over about 30 years. However, the timing and trading limitations in my 401K killed the apparent advantages, regardless of any modifications I could conceive. I have only begun to explore TAA in a very shallow manner, and it was helpful to find that this is a reasonable place to explore. Thanks again.

        • John Richards

          Eric, thanks for the follow-up! Very helpful, it connects the dots on some of the reading I’ve done.

          It took me years to finally derive a static 401K portfolio I can live with, and to settle on a rebalancing plan. It’s asset-class based, no access to individual equities. Since then I’ve been looking at what you can do (besides owning LT Treauries or Gold – my circle of competence excludes Options) to limit the downside exposure during large drawdowns. I methodically sold equities in 2008 and bought back in 2009 (1st time I deviated from Buy & Hold), so good results but lousy process – I had some level of conviction, but there was a huge element of luck which I’d like to nibble away at.

          I have since explored variations on the ideas Faber expressed, knowing that it trades one set of risks for another, and found one promising approach based on my specific portfolio modeled over about 30 years. However, the timing and trading limitations in my 401K killed the apparent advantages, regardless of any modifications I could conceive. I have only begun to explore TAA in a very shallow manner, and it was helpful to find that this is a reasonable place to explore. Thanks again.

  • Eric Weigel

    Ben – looking at the S&P 500 (or any other cap weighted index) as a momentum play benefits investors on the way up (as the best performers increase their index weights) but hurts tremendously in a down market such as 2001-2002 when the index was loaded with former high flyers on their way to much lower valuations (CSCO, AOL, MSFT, INTC are examples of this). One of the lessons I learned as a PM was the necessity to lower tracking error on the way up (ride the wave) and actually increase the tracking error on the way down (being more different). It is a bit counter intuitive for portfolio managers to think this way.

  • WEEKEND!

    “These numbers are staggering — 20% of all stocks have accounted for the entire gain over this time frame, meaning the remaining 80% have collectively given investors no return.”

    Pareto strikes again! 😉

    • John Richards

      In a sense yes, Pareto is relevant – but I think what really makes this remarkable is collectively we’re so used to “normal probability distributions”, e.g.. the Bell Curve. The work Ben references clearly shows that market returns are anything but normal. That’s a huge departure from what I expect most people assume, and it also puts most people way outside their zone of competence (perhaps why so many people do so poorly as investors). I took some statistics way back in my college days, it wasn’t my primary field of study, but I have used that information successfully in my everyday career. However, the market return distribution as noted above means all the probability assumptions I’m used to making, if applied to investing, are very suspect and likely wrong.

  • Matt

    This hits on something I have thought about recently regarding factor regression analysis, specifically with the momentum factor (UMD or WML). Most factor regression models/tools use the CRSP cap-weighted market index as a proxy for market beta, however, as described in your article, a cap-weighted index is essentially a momentum strategy. If the goal of factor regression is to help explain return variation and to estimate betas vs alpha, an equal-weight index seems like it would be more appropriate to evaluate momentum exposure. Could this help explain why momentum exposure is low relative to other factor exposures in most regression models?

    • Ben

      Never thought of it that way. Not sure I could say definitively w/o testing that out your theory but seems to make sense at first glance.

  • Bane Bear Dont Walk

    There are a lot of layers to the onion that may make the S&P 500 the single best investment in the world. I personally like to feel like a major conglomeratist (not a word I know) by owning a piece of all of those companies.

  • Mark Massey

    Maybe I just suffered a mini-stroke but I cannot seem to process the meaning of this sentence of yours, Ben:
    The remaining 12,404 stocks performed above, at or below the same level as the S&P 500.

    • Ben

      I think that statement was just showing the stocks that weren’t at the extremes either high or low. So it was the stocks that weren’t at the tails.

      • Mark Massey

        Roger that. Thanks.

  • Prof. Jeremy Siegel took the track of researching “yield” momentum in his work but nothing says that an average investor couldn’t extrapolate forward results using absolute price change from the Wharton data set representing the original and evolving S&P 500 constituents over time constructed since 1957. We use the Nasdaq 100 for research as it represents the “new” growth constituents and derive lists of momentum candidates ( the likes of which the Apples, Amazons, Googles have landed on the lists ) that are consequently held for a 2 year period and then recalculated. These 2 year periods have produced many 100%+ returns.

    Yet, as simple as it sounds, it is still difficult for an average investor and manager to comprehend this type of analysis without sound quantitative finance and investment experience ….

  • Reply ForYou

    Hi Ben–love the blog posts. Meaty food-for-thought. I have a question about the stats here. I went to the original blog post, but they didn’t explain either. I’m wondering what “total return” is–is it simply the sum of annual total returns, or is it the CAGR? I bet that sounds pretty ignorant, but I just don’t know enough about the scales to make sense.

    My reason for asking was because the total return of the S&P 500 during 1989-2015 was 1199% (the original article verified that dividends were reinvested), and the CAGR was 9.87%.

    The universe of stocks is different; however, I can’t tell whether the S&P 500 has already caught the gains in the “overperformers” or if the advice really should be “to beat the S&P 500 by oodles, just avoid the 25% of massive underachievers”.

    Thanks,
    respond

    • Ben

      My stat is the total compounded returns w/dividends. Many different ways to look at this data, but my conclusion is stock picking is very difficult.

  • Steve

    If 20% of the stocks account for all of the gains then why
    would RSP, an equal weighted sp500 etf, outperform SPY since its inception in
    2003? It would seem that if RSP was reallocating money from the 20% high growth
    stocks into the 80% laggards that RSP would have underperformed. If I keep
    selling the high growth stocks to keep them at an equal weight, how could RSP
    outperform?