by Wayne A. Thorp, CFA

The market collapse of 20082009 has led some investors to question the merits of buy-and-hold investing. This is not surprising, seeing that this was the second time in this decade the market has fallen by more than 45%. It has also led to renewed interest in ways of sidestepping such market meltdownsnamely, market timing. For proponents of Burton Malkiels random walk theory, market timing is a fools errand. They argue that you cannot use past market activity to predict future movements.

. To continue reading this article you must be a Computerized Investing Subscriber.

Discussion

Kurt Rutz from Ohio posted over 4 years ago:

I like this system. The only problem for me would be picking a stock that oscillates around its EMA after I bought it.

drb from NY posted over 4 years ago:

A dollar late and a day short (or a decade short)—this research shows that trend following does exactly what it is supposed to do and what diversified buy-and-hold has not done, that is, reduce risk and give better risk adjusted returns, preserve capital, etc, etc. Stovall cherry-picked data after the fact in his March 2010 AAII article—no one beforehand would have said that 60% Large Cap US stocks and 40% Long US Treasury Bonds is a well-diversified portfolio. A similar split between a broad stock market index and a broad bond index did nowhere near as well either in 2008 or in the decade. We all simply gave up too much following AAII’s holy grail of risk management through diversification. My moderately aggressive well-diversified portfolio that Risk Grades rated as a balanced portfolio with less than market risk in mid-2007 was rated as risky as a single small cap stock by mid-2008, and a portfolio allocated 25% each to US stocks/International Stocks/US Bonds/and International Bonds wasn’t doing much better. More follow-up on trend-following approaches would be helpful.

d from NY posted over 4 years ago:

Of course, it is a day late and a dollar short, not the reverse as I wrote in haste. Another plus of the trend following approaches noted in Wong’s research is the reduction in volatility. Wouldn’t it be nice to sleep better while doing better in the market with less draw down.

Bill from TX posted over 4 years ago:

www.aaii.com/boards/messageview.cfm?catid=74&threadid=780&enterthread=y

Ramesh from OH posted over 3 years ago:

Seems buy-and-hold is based on little more than hope, faith and misplaced patriotism. It is also induced and propagated by the brokerage houses financially boxed into their clients always buying and never selling, for the fear of broker being dumped by a company from its secondary offering. It is documented that brokers make lot more money from secondary offerings than from client commissions. So, when brokers push buy-and-hold to unsuspecting Archie Bunkers, it is obvious what they are doing: concealing and deceiving to pursue their own interest. They even wrap it in the flag. It is a shame that Wong and other truth seekers have to dig so deep and exert so much just to unearth the truth buried under the rubble of buy-and-hold propaganda.

Donald from MD posted over 3 years ago:

What’s never accounted for in these long term studies is demographics; the baby boom. During the 90’s the baby boomers reached their peak earning and investing years and now they are reaching their peak retirement years.

They will be drawing down their investments and investing fewer dollars. The only thing that can change this dynamic is if they continue to work past their retirement age, which they seem to be doing. The question is how many, what percentage of them will do this? Will they continue to invest or just draw down fewer dollars?

James from CO posted over 3 years ago:

Mr. Wong had an excellent research piece about 1-2 years ago in which he found that the 6-period EMA was the optimal time frame for a MAC system. Having used a similar system real time in my own risk management process, it was instrumental in helping get my clients out of the market in late 2000, back in in early 2003, out again in January 2008, back in in July 2009 and out at the end of this past July.

As Mr. Wong mentioned in his article, there is no Holy Grail in investing. After 10 years of investing through sound economics and fundamental analysis, 1987 happened. That changed my outlook and I began looking at charts and technical analysis — even though everyone scoffed at even the thought of placing any credibility in technical analysis, especially people like John Bogle of Vanguard who have a vested interest.

I turned to trend following in the early 1990s and have never looked back. Mr. Wong’s research should be considered very seriously by anyone who wishes to improve their risk management capability. While it may not be the Holy Grail, my real life experience over the past twenty years validates all of Mr. Wong’s research.

Milo from TX posted over 3 years ago:

Are you sure that you computed the ema correctly? I can not reproduce your results by following your formula for the exponential moving average.

Thanks.

Doug from CA posted over 2 years ago:

Could you please be more specific? MAC means moving average crossover. Wong gives summary results for 10 mo EMA. Is this 10 month EMA computed on a monthly, weekly or daily series of index readings? Is the crossover signaled when the daily, weekly or monthly reading crossed the EMA? Many studies have shown that after accounting for taxes and dividends and commissions, that MAC systems do not beat the buy and hold strategy. Wong is claiming dramatically different results Is there someplace we can obtain the detailed formulae or spreadsheets that prove this?

PG from NY posted over 2 years ago:

Warning bells should go off when unrealistic assumptions are made such as:

There are no transaction costs (commissions, bid-ask spreads, etc.); and There are no tax implications.

This is a bit like announcing you’ve come up with a blackjack strategy to beat the casinos, but adding the caveats: It is assumed that all of the other players at the table are also using this strategy; and The pit boss isn’t watching you employ my card-counting tactics.

Sure, you can add all the caveats you want, but caveats render a study useless when they don’t replicate real-world experience. Because there ARE transaction costs and tax implications of buying and selling stocks. Factoring those costs out of the equation produces meaningless results.

Hamilton from NJ posted over 2 years ago:

While transaction costs and tax implications must be kept in mind, they’re not significant today in a tax-deferred IRA or 401k holding mutual funds or heavily-traded ETFs. The only issue in a mutual-fund IRA is minimum holding time, but even it is not significant. The long-term annual yield penalty of a 3-month holding time is only a few tenths of a percent. One also has the option of holding a widely diversified bond fund instead of cash, which adds nearly one percent compounded annual yield. The minimization of drawdowns can be very significant during the payout period of a retirement account. All in all, this technique has significant long-term benefit — its biggest potential undoing could be its own popularity.

Robert from TX posted over 2 years ago:

My question is how Wong calculated the gain when you are out of the market. If you get out at a point when the market is high and can invest in something else for a reasonable return you would be better off. But what if you get out when the market is close to its lowest point. He makes a big deal about how it increases your yield but how does he compute it? It sounds a lot like Will Rogers method of beating the market, buy low, sell high.

R N Lea

John from WI posted over 2 years ago:

Paul from New York argues “There ARE transaction costs and tax implications of buying and selling stocks. Factoring those costs out of the equation produces meaningless results”

It is my belief f you try to factor IN tax implications, the results would be meaningless, simply because of the complexity of the changes in capital gains taxation over the study period.

Taxes apply to the individual buying and selling stocks. The Wikipedia article on the history of taxation in the United State shows from 1913 to 1921, income from capital gains was taxed at ordinary rates, initially up to a maximum rate of 7 percent. Congress began to distinguish the taxation of capital gains from the taxation of ordinary income according to the holding period of the asset in 1921, allowing a tax rate of 12.5 percent gain for assets held at least two years.

In addition to different tax rates depending on holding period, Congress began excluding certain percentages of capital gains depending on holding period. From 1934 to 1941, taxpayers could exclude percentages of gains that varied with the holding period: 20, 40, 60, and 70 percent of gains were excluded on assets held 1, 2, 5, and 10 years, respectively. Beginning in 1942, taxpayers could exclude 50 percent of capital gains from income on assets held at least six months or elect a 25 percent alternative tax rate if their ordinary tax rate exceeded 50 percent.

The 1970s and 1980s saw a period of oscillating capital gains tax rates. In 1978, Congress reduced capital gains tax rates by eliminating the minimum tax on excluded gains and increasing the exclusion to 60 percent, thereby reducing the maximum rate to 28 percent. The 1981 tax rate reductions further reduced capital gains rates to a maximum of 20 percent.

Later in the 1980s Congress began increasing the capital gains tax rate and repealing the exclusion of capital gains. In 1986 it repealed the exclusion from income that provided for tax-exemption of long term capital gains, raising the maximum rate to 28 percent (33 percent for taxpayers subject to phase outs). When the top ordinary tax rates were increased by the 1990 and 1993 budget acts, an alternative tax rate of 28 percent was provided. Effective tax rates exceeded 28 percent for many high-income taxpayers, however, because of interactions with other tax provisions.

The end of the 1990s and the beginning of the present century heralded major reductions in taxing the income from gains on capital assets. Lower rates for 18-month and five-year assets were adopted in 1997. In 2001, some individuals were taxed at an 8% rate.

Now factor in the filing status of the individual taxpayer, other income, deductions and personal exemptions.

I can take the results of Mr. Wong’s study and apply it to my individual situation. But how useful will it be to anyone else?

Hamilton from NJ posted over 2 years ago:

To answer Robert’s question, as the article says, All proceeds from selling are held in non-interest-bearing cash. So the yield whey you’re out of the market is zero. I’ve done backtests back to the 1980s that show that investing in a broad bond index instead of cash generates nearly one percent additional average annual return, at very little additional downside risk. The technique can also be applied to international equity index funds, REIT funds, and commodity funds, as shown in Mebane Faber’s work. Check out his site at www.mebanefaber.com.

Scott from CA posted over 2 years ago:

Are we supposed to assume that ALL stocks were sold and then re-purchased? I would think that a real-world version of this would be to trim your holdings 25%-50% once you see in the MAC signal that a decline is happening. One of my biggest concerns is missing out on the dividend payments, which are significant in my case.

A simpler way to obtain this benefit may be holding your positions, but buying/selling(to close) Puts at the point where you expect a decline/increase. If you are wrong, you lose the premium but nothing else. If the market falls as expected, you are compensated dollar for dollar by the value of the put going up as your positions fall — you could leverage this if you want.

I’ve been seriously considering this as a strategy. you would not incur significant transaction fees either, because the intent would NOT be to excercise the put, but to use it as a hedge. The hardest part would be choosing the time-frame, but you could link that to the MAC time-frame and then use the cross-over as the time to sell(to close) the put and collect your profit. Meanwhile, you are still fully invested.

My brokerage charges too much commision to buy and sell stock — typically 2% on each side, so I would need a swing/decline greater than 10% to really make it worth my while to sell and re-purchase. Meanwhile, the loss of dividend payments would outweigh any concern that my portfolio has temporarily declined 20%.

I will assume that Wong sold everything and then repurchased everything according to the MAC signals. I agree that it is frustrating to stand by and watch your gains melt-away, but how do you apply this analysis to a real portfolio?

I don’t have a stock index, I have a portfolio of high-quality dividend paying stocks — I would be very surprised if the results on the overall index, with a yield of

L Toll from CT posted over 2 years ago:

Convincing and yet confusing article. Is there analysis on this considering tax/fee charges? Using Puts? Thank for more clarity.

Scott Sneddon from CA posted over 2 years ago:

I got excited seeing the title to the article, only to realize I’d not only already read it, but commented on it.

This is an important topic. sure would be great to get a knowledgable discussion going.

I find it frustrating to simply hold but I still think you need to really define what buy and hold really is, because I bet almost nobody buys a portfolio of stocks and then lets it sit there untouched for 50 years. So what is Buy and Hold?

I’ve been daytrading for the past 4 months — in a specific trading account, not mixed with my investment account — and I think everyone should try it for at least a month. There is MAC, MACD, Stochs, Mac n Cheese. so many studies and methods to watch to help you time your trades. They work GREAT except when they don’t and you never know when that time is that the signal is false. The only time you really know if the market is going down is after the red candles form and it has GONE down. then you can be sure.

I find it interesting that for this test he waited until it had crossed over the moving average, since in my experience if you wait that long you’ve missed at least half of the move. I would love to have a system or set of rules in place — soon, with the fiscal cliff, etc. looming — to help me preserve my gains (yes I do think a bigger pull-back is coming.

Any ideas?

Michael Leigh from PA posted over 2 years ago:

Wayne, you and I have discussed several times in prior years my suggestion to include some form of trend following parameter within the AAII Stock Pro sorting fields; you have responded that those fields are fundametally driven, and not designed for technical trend folloing or market timing. The article you now highlight clearly points out that long term trend analysis can provide an important risk management tool, potentially protecting an asset base by simply being out of the stock market during major declines, and reentering markets as the trend changes. My key point is that deratives (including ETF’s)have changed forever the volatility of the stock market. Because of investment pain thresholds, critical market timing is now an important investment consideration, not as a trading tool, but as a longer term capital preservation tool.