By Carlos M. Sera and Carl E. Sera, CMT
The purpose of this paper is to send one message and one message only; long term simple moving averages (LTSMA) will approximately Cut Your Risk by Half. We measure risk by using the metric Maximum Draw Down (MDD). This paper confirms the research of others which shows that a LTSMA provides the investor with approximately equal rates of return, in most cases better, than using a buy and hold strategy with about half of the risk. This is not new information. This paper simply shows it in a simple to understand fashion.
We designed a simple experiment to demonstrate that you get equal to slightly better returns with half the risk. Anyone that wants to replicate the experiment and is proficient in any programming language can get the same results as us in less than an hour. We took the period from 1928 through 2008 and used monthly data obtained from Ibbotson on the US large company stock market to compare a Buy and Hold approach to a 12 Month SMA approach. When an investor uses a 12 Month SMA approach on the stock market, they are utilizing a mechanical crossover trading system that tells them exactly when to own stocks and when not to own stocks. You own stocks when the price is above the 12 Month SMA and you don’t when it is below. It can be used to trade or predict anything that has a continuous price stream. A LTSMA is a low pass filter. Which means it isn’t very sensitive but it captures the general trend.
The following table shows the 10 periods over the last 80 years that the stock market dropped in excess of 20% from its peak after at least a 50% increase from its previous trough. It also shows what an investor that utilized a 12 Month SMA would have lost, as measured by MDD, during the same time periods. In every significant market decline over the last 80 years, the investor that utilized a 12 Month SMA system would have lost less money than the buy and hold investor. A LTSMA is an effective risk control technique.
It is even more noteworthy to examine the 5 largest losses over the last 80 years. For the period ending in May of 1932, the Buy and Hold investor lost 83.47% while the 12 Month SMA investor only lost 23.86%. For the period ending in February of 2009 the Buy and Hold investor lost 50.78% while the 12 Month SMA investor only lost 10.55%. For the period ending in March of 1938 the Buy and Hold investor lost 50.41% while the 12 Month SMA investor only lost 13.77%. For the period ending in September of 2002 the Buy and Hold investor lost 44.71% while the 12 Month SMA investor only lost 11.40%. Finally, for the period ending in September of 1974 the Buy and Hold investor lost 42.64% while the 12 Month SMA investor only lost 10.95%. These are significant deviations and make the difference between portfolios that can quickly recover from bad periods and those that will take years longer. In portfolios that are distributing money it is the difference between having money to distribute and in many cases running out of money or severely curtailing distributions.
Please note, a 12 Month SMA during the late 1930s through the year 1941 produced a MDD of 43.03%. If one were to compare the worst case of using a 12 Month SMA to the worst case of using the Buy and Hold, this 43.03% MDD is comparable to the 83.47% MDD. This number is roughly half. We did not show this number in the table because we wanted to show how a 12 Month SMA does in comparison to the actual peak to trough cycle witnessed over the last 80 years. Later in this paper we show how the Buy and Hold does over various time periods vs. the 12 Month SMA.
Does using a 12 Month SMA always outperform a Buy and Hold strategy. The answer is no. We know of nothing that outperforms something else at all times. So it’s important to understand the periods where a 12 Month SMA will and won’t outperform. The following table compares the compounded annual growth rate of a Buy and Hold strategy to a 12 Month SMA strategy. During periods where the 12 Month SMA model said to be out of stocks we invested the money in US Intermediate Government Bonds. Bond data was also obtained from Ibbotson.
We can see that there are long periods of time where the Buy and Hold strategy outperforms the 12 Month SMA strategy. We can see that there are periods where the reverse is true. It is worth noting however, that the 12 month SMA outperformed the Buy and Hold regardless of what year you started in every time frame through the year 2009. However, it isn’t always the case. For example, from 1975 through 2000, the Buy and Hold had a 25 year compounded growth rate of 17.3% while the 12 Month SMA only had a 16.2%. Why is it that during this period there was an outperformance? The answer has to do with bull and bear market cycles. The period that started in 1975 was coming off of a severe bear market that ended in 1974. The period that ended in 2000 was the end of a major bull market that ended in 2000. These are ideal conditions for a Buy and Hold approach and why it did better for this period. A good way to think of it is that Buy and Hold will more than likely outperform the 12 Month SMA if you measure from the end of a bear market to the end of a bull market.
What can we gather from the use of any LTSMA approach? One of the advantages of a LTSMA approach is that it will always be fully invested at market tops. It will not miss trends. One of the disadvantages of a LTSMA approach is that it will always be fully invested at market tops. It will lose the maximum possible from a market top before the mechanical crossover signal tells the investor to get out. One of the advantages of a LTSMA approach is that it will never be invested at market bottoms. It will avoid periods of rapid market declines and preserve capital. One of the disadvantages of a LTSMA approach is that it will never be invested at market bottoms. It will always underperform in the early stages of a bull market. Our paper entitled Market and Investor Behavior provides some answers to the paradoxes of any LTSMA approach.
The following table we call Buy and Hold vs. Buy and Fold. The Buy and Fold refers to the implementation of a long term moving average where the investors folds their hand when they get a signal and sells their stock in order to preserve their capital like any good poker player. We tested a 30 year period to illustrate two things. The first is our claim that you get an equal return for half the risk. The second is to show that there is nothing magic about a 12 Month SMA. A 9, 12, 15 and 18 month SMA work approximately as well and in fact, over the 30 years tested in this table the 15 month SMA is best.
As we can see, investing $10,000 in the S&P 500 grew to $228,229 by the end of 2008. Investing in the 9,12, 15 and 18 month SMA grew capital to in one case less and in the other three cases more. However, when you examine the compounded annual growth rates they are very close. The top performing LTSMA of 15 months is only 0.72% better than the Buy and Hold. However, what is significantly different about the Buy and Hold compared to any of the 4 LTSMA cases shown is that once again, we see that using a LTSMA reduces MDD. It does so by a factor of approximately 2.
Our favorite way of evaluating the merits of any investment approach or manager is to calculate the compounded annual growth rate and divide it by the maximum drawdown. This ratio is known as the MAR ratio and in our opinion is the most relevant measure of how people should balance return with risk. So for example, we know that large company stocks made 8.92% compounded annually over the 80 year period with an MDD of 83.47%. So we divide 8.92% by 83.47% to get a MAR Ratio of .1068. We also know that using a 12 Month SMA for the same 80 years made 11.26% compounded annually with a maximum drawdown of 43.03%. This gives the 12 Month SMA a MAR Ratio of .2617.
It is well understood that investors are more likely to maintain an investment strategy if they can reduce the volatility of their portfolio. Our paper entitled Market and Investor Behavior explains some of the reasons why we are certain that a LTSMA strategy has a higher likelihood of success than a Buy and Hold strategy. This likelihood of success is exacerbated when one takes into consideration a portfolio that is taking periodic withdrawals. Who then should utilize a LTSMA strategy? In our opinion, retirees, charities, foundations, endowments and retirement plans would benefit greatly. These 5 types of portfolios have one thing in common. In most cases they are taking money out on a periodic basis and during bad times or down markets they are selling investments that will not have a chance to recover during the subsequent good times or up markets. Reducing the magnitude of losses during these bad periods increases the probability that a portfolio will experience sustainable distributions.
For those that want further insight on sustainable distributions, please refer to the February 2010 Journal of Financial Planning article entitled A Simple Dynamic Strategy for Portfolios Taking Withdrawals: Using a 12-Month Simple Moving Average. In this article we examine scenarios investors encounter when making decisions about their portfolio in the distribution phase or withdrawal phase.
Many people that read this paper will want to understand how LTSMAs with their focus on minimizing maximum drawdown equate to the basic building blocks of modern portfolio theory or MPT as it relates to the inclusion of both stocks and bonds in a portfolio. The answer is LTSMAs are in direct conflict with the MPT notion that portfolios should always have some amount invested in stocks as well as bonds. There is no way to reconcile the two. The two are in stark contradiction in this regard. The LTSMA investor believes there are times to be in stocks and times to be in bonds but no time to be in both. The closest we can come to reconciling the two is to give an example that highlights the differences.
Assume the MPT practitioner sets a stock allocation of 70% and a bond allocation of 30%. Over time this 70% target is rebalanced as stocks and bonds diverge from each other. Our research shows that stocks spend approximately 70% of their time above most LTSMAs. So the difference lies in this approximate 30% of the time where the MPT practitioner owns stocks while the LTSMA practitioner owns bonds. Our paper entitled Market and Investor Behavior shows which one is the better option during this 30% period of dispute. Investors can also benefit from our paper entitled The Natural Asymmetry of the Stock Market to better understand the actual annual distribution of stock market returns. It is this asymmetry that LTSMAs recognize while MPT does not that make LTSMAs a superior stock to bond rebalancing approach.