Stock Market Timing Using Historical Moving Averages

In the wake of the recent stock market drops, there has been a increased amount of interest in certain market timing systems that would have told an investor to be “out” of the market during late 2008 and early 2009. Below are a sample of the more popular mechanical systems, which have clear parameters and are easy to follow using basic investing tools.

100-Day Moving Average (FundAdvice.com)
FundAdvice.com is run by Merriman Capital, which provides money management services using both buy-and-hold and market timing models. I’ve mentioned them before for their index fund model portfolios and the related “Ultimate” Buy-and-Hold Portfolio. The website has an entire market timing section, and I am specifically referencing the articles here and here from 2001 and 2002.

Merriman recommends something called the 100-day moving average. For any given asset or mutual fund, you calculate the average of the most recent 100 days of closing prices. Each day, you’ll have a new average by adding the newest price and dropping the oldest price from the average. Then, you compare the current price of the fund with this simple moving average (SMA).

Each day the market is open, if the current price is above the 100-day SMA, you should buy the fund or hold it if you already own it. If the current price is below the 100-day SMA, sell it. After you sell, place your proceeds in cash (money market fund) until the fund price is once again above the average.

200-Day Moving Average (Faber)
The 2006 paper A Quantitative Approach to Tactical Asset Allocation by Mebane Faber explores a very similar but even simpler mechanical system. Here you take the 200-day simple moving average of any index, for example the S&P 500. But this time, you only calculate this at the end of each month. If the current price is greater than the 200-day SMA, then you buy. If the current price is less than the 200-day SMA, then you sell and move to cash (90-day T-Bills). This results in some impressive backtested results, but with much fewer trades than other systems.

By the way, you can track the 100-day and 200-day SMAs easily using most stock quote websites like Yahoo Finance. Here’s the chart for the S&P 500:

As you can see where the S&P 500 is below the green and red lines, you would have been told to “sell” and stay in cash during some of the big drops.

10-Year Trailing Average P/E Ratio (Shiller)
Prof. Shiller of Yale University is well-known for his book Irrational Exuberance. From Wikipedia: “Published at the height of the dot-com boom, it put forth several arguments demonstrating how the stock markets were overvalued at the time. Shiller was soon proven right when the Nasdaq peaked on the very month of the book’s publication, and the stock markets collapsed right after.”

Part of his argument was based on historical price-to-earning ratios, or P/E ratios. This indicator went throw some iterations and became known as PE10, which takes the current price of the S&P 500 and divides it by the average inflation-adjusted earnings for the past 10 years. Essentially, it is a 10-year moving average of the P/E ratio. If the current P/E ratio is significantly higher than the 10-year average, the market is overvalued.

You can find updated numbers and more at his Yale website.

Warnings about Market Timing
I haven’t taken the time to deeply analyze any of these systems, I am only presenting them here for debate. However, I think I should throw out a few quick warnings.

Depending on which stock market index you track, and the time period you track, all of the methods above reduce risk (volatility) without significantly reducing overall returns as compared to buy-and-hold. However, there have also been long stretches where timing underperformed buy-and-hold, which can make it a hard strategy to implement over the long run. The main reason we are talking about them now is only due to recent hot performance. In addition, market timing requires regular attention to stock market activity, increased trading costs, and tends to be much less tax-efficient.

Even Merriman states: “we have concluded that relatively few investors have the tenacity, discipline and faith required to be successful market timers.” But perhaps the same is true for Buy-and-Hold? It is possible that these market timing systems will outperform in the future as well. Or they may not, and you’ll be jumping in only after a recent hot run. I don’t have a well-composed argument either way right now, but I can say that I currently am not using any of the above systems.

Comments

  1. hey Jonathan, I have been a big fan of your site for a long time and I am glad you have finally started thinking about technical analysis! the only advice is the following.

    1) following the green line may look easy , but it is incredibly hard to do in practice.

    I think a good way to get started on technical analysis is to treat it as a form of asset diversification. in order world, allocate half of your stock portfolio to technical analysis based strategy and another half to buy and hold strategy.

  2. Maybe I’m being totally dense here, but those 1- and 2-year moving average methods seem to indicate buying high and then selling low– if the current price dips below the moving average, selling only makes sense if it continues to plunge, which doesn’t really seem predictable? Maybe it works just because of herd mentality, i.e. if the price begins to dip below the average, everyone is going to sell so you might as well get out before it goes any lower?

  3. I may be in the same boat as Madame X:

    “Each day the market is open, if the current price is above the 100-day SMA, you should buy the fund or hold it if you already own it. If the current price is below the 100-day SMA, sell it. After you sell, place your proceeds in cash (money market fund) until the fund price is once again above the average.”

    Seems like I’d want to buy when it’s _under_ the moving average since the average seems to indicate a price it’s most likely to return to and then sell it when it’s over the 100-day SMA.

    Am I backwards or is the world?

  4. One problem with the Yahoo charts is that they don’t include dividends.

  5. PE10 vs. moving averages is like astronomy vs. astrology. PE10 is a ‘valuations matter’ indication. MA analysis is follow the squiggle.

    Madame: that is the downside of using MA analysis. It works great in a unidirectional market, but in a churning sidways market whipsaws are frequent. The only reasons we’re talking about moving averages these days is because 1) it avoided the big drop 2) it’s CAGR approximates buy and hold over the past 20 years – but there’s enormous recency bias.

    While CAGR between MA and bnh is now similar, subtract fees and taxes and bnh is still superior.

    One of my favorite articles on valuations matter: http://bear.cba.ufl.edu/karces.....J_2005.pdf

  6. Jonathan, I’ve been a long time reader of your blog (probably 3 years). Personally, I’m a little disappointed to see articles about market timing, because it’s not fundamentally sound like all your other advice. When I look for personal finance books or blogs, I look for market timing advice. If they advocate market timing, then I run for the exit, because I don’t know if I can trust any of the other info.

    Market timing is a gamble. You can do better than average, or you could do a lot worse than average. Fundamentally sound investing does not rely on timing. You just invest in the whole market and ride its up and downs over time. If you start have articles on market timing, then you might as well start adding CNBC-style hottest stock tips too.

  7. It is true that market timing can do you wrong. I like the Jason’s idea. If you want to play it – only allocate a small portion. Then when your funds are gone – think of it as tuition well paid for your learning experience. And if it isn’t gone – then keep studying!
    I liked this post and thought it was very informative. I really like that this post describes using really long term SMA’s because it results in a much more passive investing style rather than the short SMA’s which are always switching buy/sell indicators and result in more trading and require more attention. Some people like that and do well. That is cool. For now, I’m sticking with my monthly deposits into a buy and hold diversified allocation strategy. However, I do have a small account that has proved invaluable as a learning tool for me.

  8. @Jason – I agree, I think it would be too hard for me to follow the green line as well.

    @Madame X & Jay – A common explanation given for these types of systems is momentum. If the S&P 500 starts going above the average, then it has “momentum” and will keep going for a while. If it starts going below, then it has “momentum” and will keep dropping.

    @Bucky – I appreciate your opinion, but I’m looking for more definitive evidence against market timing here than simply “it doesn’t work”. I try to point out reasons against market timing and other hot tips all the time, so why not help me refute this as well? :)

  9. I must say I agree with Madame X & Jay. Why buy high and sell low? If you believe a stock certificate is an actual business that has intrinsic value, you would want to get that value at an awesomely cheap price and sell when you think people are overpaying for it.

    I guess technical analysis is trying to predict the market movement, whereas fundamental analysis predicts intrinsic value. I think I’ll stick to fundamental analysis, but if there are any compelling articles about technical analysis, I’d like to read them.

  10. NBR has had a market timing portfolio for many years (since 95) that has outperformed the S&P 500 index, check it out:

    http://www.pbs.org/nbr/site/re.....ach/drach/

    Drach does daily analysis, meaning it’s not an “auto-pilot” portfolio.

  11. I agree with Bucky… market timing does not work, by defintion. If there was a way to predict the future by analyzing the past (which there isn’t), “the market” would take this informatoin into account and the price of stocks would immediately compensate and adjust to their true instrinsic value. A lot of people think there are methods of market timing that work. Some work when “back tested”. But it’s the future that counts. At any given time, about half the experts turn out to be right and about half turn out to be wrong.

    That being said, Jonathan, why would you look for a methodolgy that told you to be OUT of the stock market in late 2008 / early 2009? The market is up big time since then. With the benefit of hindsight, that’s when you should have been loading up!

  12. Giving examples of a market timing methods that have outperformed S&P does not prove that MT works. (Similarly, giving examples of MT underperforming S&P does not prove that MT doesn’t work either).

    You have to look at MT methods as a whole and compare it to S&P. I’ve seen articles stating that collectively, actively managed equity funds underperform index by about 2%. Even if you ignore the expense ratio, they’re still losing.
    http://www.fool.com/School/Mut.....Record.htm

    I’ll give you one big reason why MT loses (collectively and over time). In order to MT, you can’t be investing 100% of your funds. Assuming that the market is overall growing, you’ll be missing out.

  13. I have studied both technical and fundamental and they each have their own merits. to the technical nay-sayer, I totally agree with it is unreasonable to have to WAIT for an undervalued stock to go above the moving average before you can buy it. but technical analysis does help you avoid huge losses like what happened back in Oct 08.
    so currently this is what I do to try to take advantage of both technical and fundamental analysis.
    1) find a stock that’s intrinsically cheap (using fundamental analysis)
    2) buy it and put a stop loss. but if it the stock does go above moving average, make sure you then FOLLOW technical analysis and sell the stock when it goes below the moving average later in the future. this will make sure you have profit from the trade.

  14. I think it is important to note the distinction between what Shiller is talking about and the moving average people… As one other person pointed out, astronomy vs. astrology – fundamentals vs. emotion.

    One of the few tried and true principles of investing is reversion to the mean which is what Shiller is talking about and is a sound investing principle. It is essentially saying if something is out of whack, it will likely revert back to it’s long term average. It is the same principle that rebalancing works on and target date funds take advantage of.

    Interestingly, Shiller’s approach is exactly the opposite of what the other people propose. With Shiller, when the market is higher than the historical average, you sell. When it is below, you buy.

    As Swensen at Yale once said, he doesn’t time the market, but he does “price” the market. If something is under or overpriced, they adjust the portfolio accordingly. Put simply, they rebalance and sell things that have had an above average run.

    Shiller nailed not only the tech crash, but the housing crash. He was probably a little bit early on both, but he was right all the same.

    The problem with being right though, is the market can remain irrational longer than some people can remain solvent. Still, reversion to the mean continues to be one unfailing principle of investing… sometimes it just takes a bit to revert, but it always does…

  15. Joshua @ Accountable Living says:

    Well, I use market timing. It goes like this: Buy mutual fund. Wait 5- 10 years. Sell mutual fund. My timing method just takes longer than some others !! :)

  16. Shiller’s PE10 is certainly different, but it is still market timing based on mechanical formulas using moving averages. So does mean that supporters of PE10 as “astronomy” say that some market timing methods do work?

  17. @Market Timing Naysayers – It’s good that Jonathan is looking at MT. It doesn’t mean his blog is “going down hill.” There is a crowd out there that doesn’t believe in buy and hold. They may be outperforming buy and hold right now. When people like me are looking at stock market returns in their lifetime and seeing that they would have been better putting money in a mattress over 10 years, then our eyes start to wonder. Now, I’m not straying from the flock, but I have looked at these MT examples and they do look enticing.

    Some of you treat buy and hold as a religion as though you would be down to one dollar from a high of one hundred thousand and you would still say, “Buy and hold is the answer!”

    Again, I’m not straying from the flock, but my faith is not what it used to be when I load up my 5 vanguard funds and see red over the long term which is exactly the opposite of what buy and hold predicts. Knowing that Jonothan is not a zealot makes me happy. He is at least open to presenting the alternatives while still maintaining that buy and hold is his “religion” of choice. It’s not like these market timing examples are selling some “system”, it’s free information. It’s not like he is touting some black box insurance annuity (you would think he is by some of the comments).

  18. Jonathan, if you haven’t yet, I highly recommend that you read Faber’s book The Ivy Portfolio. It will give you a lot more data about this kind of strategy than the speculation on this thread seems to so far.

  19. Market timing is easy to bash, but “market timing” tends to be a catch-all for a lot of different techniques. As was pointed out, Moving Averages are an indicator of momentum, and even Burton Malkiel (A Random Walk Down Wall Street) admits that momentum is the one area where markets are not random. This form of technical analysis is less about predicting the future as it is discerning what’s happening now – is there currently more buying pressure or selling pressure. Other forms of TA (head and shoulders, Fibonacci retracements, etc, etc) are a different category. Since every active mutual fund uses some form of market timing, the only true buy-and-holders would need to be total market indexers with no downside limit.

  20. Right now seems like a good time to buy regardless of your stance on MT… and you cant tell me that you don’t see potential for major upside gains over the next 5 years.

    So, if you buy this, and find yourself getting a little greedy these days, then you are a market timer. And if you are scared and are buying a little less these days, then you are a timer as well. I would wager that 95% of us are impacted by the market trends to some degree. I would love to hear from anyone who has consistently invested the exact same way day in and day out over the past 2 years.

    On the way down I know I built up cash as fast as I could, and only recently started buying in again at the same level I had been. I did not stop buying in completely at any time, I just cut the rate at which I invested in half – I am now playing catch-up, and I think I am ahead for doing it.

  21. A couple of comments:

    I agree with Jerry Verseput in that Moving Average analysis is not necessarily market timing. I know Faber on his blog sometimes refers to his system as a timing system but it is in actuality a momentum strategy, as are all moving average techniques. And Malkiel (Random Walk Down Wall Street) does indeed point to momentum trading (or growth trading) as a major driver of stock prices during certain periods since the Great Depression and also the method that Keynes used to make his money in the market. Market timing in contrast to momentum trading attempts to predict where the market will go. Traders who try to time the market will look at a chart and say,”This is a bottom, buy everything,” or conversely,”Sell it all, we’ve reached the top.” This type of trading is true market timing in that the practitioner believes they can predict market movements. Momentum trading makes no such claim. Traders who use a momentum strategy think that a market may continue to trend in a certain direction and if it does so they could make a lot of money. Just in case it doesn’t, and no one knows whether it will or not, they set a tight stop to limit losses. Therefore Faber’s paper is more momentum trading as it attempts to ride uptrends and sit out downtrends but not guess when those trends start and stop.

    In response to those readers who talk about intrinsic value of companies and buying stock as if buying a company, I think this kind of valuation is flawed for readers of this blog who are presumably small scale investors. This method works for Buffet and Swensen as they can buy 10%, 20%, or even an entire company. In contrast if you own a few hundred shares you are essentially buying current management. If you bought 10%, or even 5%, of a company then you could help them unlock the hidden value that you see in it. If you own a few hundred shares you essentially hope management is able to unlock this value without running the company into the ground first. Small scale investors should stay in indexes and bond funds.

  22. Why does everyone strongly link “market timing” with finding the absolute (really “local”) peak or trough of the market? Anyone who got out between 2004 and 2007 would be ahead even if they missed the peak over 14000 in 2007 (you could even argue about having gotten out in ’98-’01…) The market and business CYCLE is inherently cyclical. It is only prudent to be alert to when the cycle is going down. Saving ones self from the black swan -30-50% market dives is the only way to stay sane while investing. While I dont think it makes sense to follow the 200day SMA in/out on a month by month basis, it certainly makes sense to be AWARE of it (along with the Leading Economic Indicators) to see what business phase we are in. “The return OF my money is more important then the return ON my money…”

  23. Justin wrote: “Some of you treat buy and hold as a religion as though you would be down to one dollar from a high of one hundred thousand and you would still say, “Buy and hold is the answer!””

    Just to clarify, I am not necessarily advocating buy and hold of a single stock. I’m talking about a whole market index. You would not have the whole market index down to $1 from $100,000. If it did, it probably would be a good idea to hold (or buy) since it must be severely undervalued.

    I do believe that there is underlying value to the stock market, so maintaining asset allocation can help adjust for that a little bit. But to use market timing is dangerous because 1) you will be out of the market for some periods, or 2) the stock market can be overvalued for 10+ years at a time.

  24. Author Andrew Tobias has written some great analyses of ‘systems’.
    F’rinstance “The Only OTHER Investment Guide You’ll Ever Need”, 1987,
    detailing many systems that have looked great on paper, but whose
    originators are somehow not rich. He’s quite hard on technicians, but
    pure value and other schemes all have faults….one idea that has held up
    fairly well over time, and over the past year is based on Libertarian
    Harry Browne’s “Permanent Portfolio”, of a quarter of your assets in
    stocks, a quarter in cash, a quarter in gold, and a quarter in long
    treasuries. No, alas, I didn’t have the ‘foresight to invest a year and a
    half ago…..buy and hold from $100,000.00 down to $1? well, I DO
    have some GM stock, LoL.

  25. Hey Jonathan -

    If you’ll forgive a long comment, Faber has a really interesting analysis in The Ivy Portfolio:

    “Many pudits also show how missing the ten best days of the market can decimate your returns, and use that as proof that market timing doesn’t work. What they don’t understand is that the vast majority (roughly 70%) of both the best and worst days occur when the market is below the ten-month simple moving average. The simple reason is because the market is more volatile. … from 1984 to 1998, the buy-and-hold return was 17.89% for the Dow. Missing the ten best days resulted in returns of 14.24% per year, while missing the worst ten days resulted in annual returns of 24.17%. However, missing both the ten worst and ten best periods resulted in retunrs of 20.31% – higher than the buy and hold return (and likely with less volatility).”

    This is from p. 167-169 of my copy of the book. Again, it is a very good read, and I recommend it highly.

  26. Jonathan

    You should also look at a method called “Value Averaging”. The book is available on Amazon and it could be another alternative to the question that you pose. It’s been a while since I read it, but I seem to remember that the underlying theme was something similar…

  27. Timing can be done successfully. Trend following has been proven to be a successful long term strategy.

    As for fundamental v. technical analysis….are not the fundamentals and all recent news for a stock reflected in the price of a stock? If so, technical analysis should reflect all fundamentals.

    As for the 10 best days, over the past 10 years, if you had been invested 30 days prior to the best days, AFTER catching the 10 best days, you would still be down 7.55% on average. This is due to the fact that the 10 best days typically occur after severe or protracted downturns.

    Your odds of being able to miss the 10 best days….and no other day over the course of 10 years?

    One in 14.06 Quintillion.

    Just a few thoughts.

Speak Your Mind

*