No. of Recommendations: 266
Over the course of the last 8 months, the MI Board community has come together to collect the data to look at the 1969-1985 time period and how TMF mechanical investing strategies would perform in a non-bull market. The 1969-1985 period includes a time when 3 distinct aspects of investing were present that are not present in the 1986-1999 time frame for which the MI Board has currently looked at. The 3 distinct aspects are:

1) A prolonged bear market.
2) A period of time where Small Cap investing worked as well or better than Large Cap.
3) A period of time when 'Value' investing did as about as well than 'Growth' investing.

One of the primary concerns about the MI strategies are how they will do in the above investing climates. In looking at the Keystone family of screens, two main conclusions appeared:

1. Investing in the Keystone strategy provided a small advantage over investing in the general market regardless of holding period or start date.
2. Investing in a Keystone strategy during a significant bear market virtually guarantees a one year drop of 40% during that time period.

Keystone introduced the concept of investing like you are driving towards a cliff while blindfolded and without any breaks. With Keystone, regardless of holding period, you will almost assuredly incur a very large loss during a bear market.

The RS strategies paint a slightly different picture, both with the annual holds and the shorter holding periods. The RS strategies as a whole had CAGRs in the 20% range (RS-26, RS-13, RS-O, RSW334) indicating the ability to significantly outperform the market in all types of conditions. There was great variation within the hold times and start periods, ranging from a low 12% CAGR (RS Semi Jan Start, RS Annual Jan, Feb, Mar Start, averaged over all 4 variations, 4 stock to 10 stock screens), to a high of 24% CAGR(RS Quarterly Feb Start, RS Semi Feb Start, RS Annual Aug Start, averaged over all 4 variations, 4 stock to 10 stock screens. The highest return, apart from a 37% CAGR for the RS26 Annual 4, August Start was a 30% CAGR. The lowest return, apart from a 6% CAGR for the RS26 Annual 4, February start was 10%.

The first thing to notice about the annual holding period is that both the returns and the maximum loss vary substantially both within a single RS Variant and across multiple RS Variants. While the least harmful Keystone Annual screen (4-6 stocks held, 1-3 year period) presented a loss in portfolio value of 32%, the least harmful RS Annual screen (4-6 stocks held, 1-3 year period), presented a loss in portfolio value of 23%. The most harmful Keystone Annual was the Keystone 100 Annual, July 1972 rebalance at -62%, almost identical to the RS-26 Annual, June 1972 rebalance at -58% (4 stock to 10 stock screen average). The -38% average maximum loss in the Keystone Annual Family was similar to the average -36% maximum loss in the RS Annual Family (minimum 4 stock screen).

However, the similarities within the Annual holds end there. The defining statistic of the RS-Family Annual hold is the 1973 Returns for many of the mid-year rebalance periods. The April 1973 to September 1974 period was firmly planted inside the Bear Market. Take a look at the following table:

RS Rebalances, 1973, 12 month hold, 4 stock to 10 stock screen average:

Month S&P 500: RS 13 RS-26 RS-O
April -15.73% 63% 18% 38%
May -15.57% 35% 43% 39%
June -16.84% 28% 26% 25%
July -17.51% 21% 33% 23%
August -26.71% 4% 61% 20%
September -30.79% -6% 23% -8%
For many of these strategies, either the 1972 rebalance or the 1974 rebalance was the maximum loss. However, it is very interesting to see that over a six month period, an annual hold in this strategy provided extremely strong outperforming years, usually complementing the extremely negative years either prior or following the positive year. This type of behaviour does not exist within the Keystone family.

Moving to the Semi-Annual holding period, the type of hot and cold performance seen with the annual version of the RS Screen continued. The most interesting minutia is that the RS-13 Semi started in March and April has a maximum yearly loss of only 10% for the 4, 5 and 6 stock screens. In general, the maximum losses here averaged 30%. The variants were segregated: average maximum losses (4 stock to 10 stock screens, averaged over Jan-Jun start times) were 33% and 34% for RS-26 and RSW334 but only 26.5% for each of RS-13 and RS-Overlap. The most striking thing about the Semi Annual holding period is that almost without fail, a poor year (e.g. a 30% loss) was both proceed by and followed by positive years and that either the earlier or the later year was a gain of more than 30%. Most striking is that a bear market seemed to be a slingshot for this screen. Following a bear market, the RS screens with a semi annual hold had rocket-like returns. Returns in 1982 were routinely in the 100% range. The lowest RS26 Semi return for the 12 month period starting June 1982 was 240%(4 stock to 10 stock hold). The lowest return among all variants for this time period was 164%(4 stock to 10 stock hold, all 4 RS variants). Also, without fail, a longer or stronger RS metric (RS-26, RS-O, RSW334 vs. RS-13) did better coming out of a bear market.

Shorter RS periods do the best coming into and do the worst coming out of a bear market.

All RS variants with a semi-annual holding period had a 20% CAGR during the 1969-1985 time frame, using the average of 4 stock to 10 stock screens and the average of all 6 Jan-Jun starts.

By far, the RS Quarterly screens had the lowest maximum loss of any family of screens for any holding period. The average maximum loss for Jan, Feb and Mar starts for all variants was 32%, 18% and 22%. Both RS-Overlap and RSW334 had an average maximum losses of 22% while RS-26 and RS-13 had an average maximum loss of 26%. It is clear that within the RS-Quarterly family, going with a RS metric that combines more than one holding period will help you avoid large losses in a bear market.

The RS Quarterly strategies had an 22% CAGR, averaged over all start months, all variants and all 4 stock to 10 stock screens. RS26 has the low CAGR at 20% while the remaining strategies were within 1% of the average. The difference in CAGRs is not significant.

The final, and perhaps most anticipated category are the RS Monthly strategies. The RS13 Monthly strategy had the lowest maximum annual loss of 27%. The RS-Overlap and RSW334 strategies were next at 33% and the RS-26 strategy had the highest maximum annual loss at 37%. This once again indicates that coming into a bear market, shorter-term RS strategies are preferable to longer-term RS strategies and multiple-period RS strategies fit in the middle of these two extremes.

However, in the single 30 day periods of Nov 1973, Oct 1978 and Mar 1980, 95% of all RS Variants, regardless of whether you were holding as low as 4 stocks or as high as 10 stocks, lost between 28% and 30%. The minimum loss was 26% and the maximum loss was 36%.

A 30% loss in a single month happened three times within a 7 year span.

Two more interesting statistics can be pulled from the RS strategies:

The RS-13, RS-26 and RS-Overlap strategies all have a CAGR of 15%, averaged across 4 stock to 10 stock screens. However, the CAGR of the RSW334 strategy is 19%. Perhaps an RS Metric with a 52 week component in it provides additional value in a bear market that is not apparent when looking only at a bull market such as 1986-1999.

The last statistic is the most curious statistic of all:
RS Monthly CAGRs, 1969-1985, holding X number of stocks:

4 5 6 7 8 9 10
RS-13 14% 13% 13% 15% 16% 16% 17%
RS-26 12% 14% 16% 15% 16% 17% 19%
RS-O 10% 10% 14% 15% 16% 16% 19%
RSW334 17% 17% 20% 20% 19% 19% 18%
With the exception of RSW334, all the monthly RS variants show an upward trend when adding more stocks. This is completely contradictory with the data we see with the 1986-1998 time period. A brief look at the position by position returns confirmed this chart: The #10 position in the RS-13, RS-26 and RS-O screens all had between a 32% and 34% CAGR, more than double the average of the first 9 positions. The only explanation I can give is that perhaps the best RS stocks get hit the hardest in a bear market and thus going deeper actually helps the RS family of screens in a downturn.

What does this say about bullet strategies? A lot. The CAGR of the 1 and 2 stock bullet strategies within the RS family of stocks ranged from -1% to 11% for monthly hold and -9% to 33% for all holds combined. The March RS-26 Quarterly bullet strategy has a CAGR of 1% and a regular Standard Deviation of 85%.

"Go ahead. Make my day."

I have looked at a lot of data for this post. I'd like to summarize some of the lessons learned:

1. RS Strategies strongly outperform the market in both bull and bear markets alike
2. There is a lot of variation with regard to start time on the semi annual and annual hold times.
3. During a bear market, the RS strategies may have a bad year. However, they virtually always have a strong rebound year, preceding, following or both preceding and following this bad year.
4. RS Metrics with a shorter 'longest RS' component will have less worse maximum losses going into and throughout a bear market than RS Metrics with longer 'longest RS' components.
5. RS Metrics with a longer and stronger (i.e. multiple) RS components will do better in the recovery period following a bear market.
6. A quarterly rebalancing period for an RS strategy is the optimal holding period for mitigating the negative effect of loss of capital in a bear market.
7. (Questionable) Holding more stocks within an RS strategy will improve your returns in a bear market environment and diminish your returns in a bull market environment.


Print the post  


What was Your Dumbest Investment?
Share it with us -- and learn from others' stories of flubs.
When Life Gives You Lemons
We all have had hardships and made poor decisions. The important thing is how we respond and grow. Read the story of a Fool who started from nothing, and looks to gain everything.
Contact Us
Contact Customer Service and other Fool departments here.
Work for Fools?
Winner of the Washingtonian great places to work, and Glassdoor #1 Company to Work For 2015! Have access to all of TMF's online and email products for FREE, and be paid for your contributions to TMF! Click the link and start your Fool career.