A cardinal rule of investing is that higher risks accompany higher returns. By combining non-correlated strategies in our portfolio, we can mitigate the risk and keep the higher returns. This is a central tenet of Modern Portfolio Theory. As Workshop investors, we try to reach the highest return we can at our desired level of volatility.
 |
|
|||||
|
||||||
|
||||||
By
The protocols are technical details. As good mechanical investors, we are resolving these issues in advance. But, selecting the strategies is the most important decision.
Our goal, simply stated, is to get best return possible (highest Compound Annual Growth Rate, or "CAGR") with the lowest amount of volatility (lowest Geometric Standard Deviation, or "GSD"). That's easy to say, much like "I want to be rich," but hard to do. High returns with low volatility are the Holy Grail of every investor.
In general, the greater the expected return on an investment, the greater the risk. U.S. Treasuries have virtually no risk, but have returns that barely beat inflation. The S&P 500 has much better historical returns, but with greater volatility. Some of our stock strategies have had very high returns, and accompanying high volatility, over the last 15-30 years. This is one of the rules of investing: If you want higher returns, you have to take on more risk/volatility.
Is there any way can we get higher returns without all that volatility? Maybe. That's what we are attempting to do by selecting a blend of portfolios that might partially offset each other's volatility.
Last week we discussed correlation coefficients as a way to compare two strategies. These coefficients vary from +1 (when one strategy goes up, so does the other) to -1 (when one strategy goes up, the other goes down). Imagine investing in two strategies. Both have returns of about 25% per year, and both have a lot of volatility. Strategy A returns 60% in year 1, 25% in year 2, and -10% in year 3. Strategy B returns -10% in year 1, 25% in year 2, and 60% in year 3. If we put half our money in Strategy A and half in Strategy B, we would have returns of 25% in each of those three years. Voila! By taking two strategies that are negatively correlated, we can get high returns with low volatility (in this simple example, no volatility), even though each strategy on its own has high volatility!
Real-life investing never works out that neatly, but with the right blend of strategies we can get theoretical improvements over the return of any one strategy. All that remains is to see how the theory plays out in real life.
Let's look at our proposed model portfolio, listing the CAGR and GSD for each strategy. The "blend" column shows the combined CAGR/GSD as each strategy is added. For example, PEG-13 monthly has a CAGR of 67 and a GSD of 41. CAPRS quarterly has a CAGR of 58 and a GSD of 48. The blend of PEG-13 and CAPRS has a CAGR of 66, a GSD of 38, and a Sharpe Ratio of 1.30.
Blend
� � � � � �� CAGR GSD� Sharpe� CAGR/GSD� Sharpe
PEG-13(M)� �� 67�� 41�� 1.04� � �
CAPRS (Q)� �� 58�� 48�� 1.08� �� 66/38� � 1.30
Plowback(S)�� 34�� 40� �0.79� �� 57/36� � 1.33
KeyEPS� � � � 28�� 31� �0.75� �� 50/33� � 1.30
LowPB� � � �� 31�� 28� �0.96� �� 47/29� � 1.38� �
As you can see, no individual screen had a Sharpe Ratio over 1.08. But, by blending screens, we're able to gradually increase the Sharpe Ratio to an impressive 1.38!
Here's an interesting bit. Notice that when KeystoneEPS was added, the Sharpe went back down slightly. A drop that small might not be significant, but I decided to try the blend without KeyEPS, and guess what? The four-strategy blend had a 52% CAGR, a 31% GSD, and a Sharpe Ratio of (hold your breath) 1.42. So, KeyEPS brought the volatility down, but not enough to justify the lowered return.
Should we even use the KeystoneEPS strategy? On the one hand, it offers some increased diversification. Our backtests are not perfect indicators of future performance. Minor differences in CAGR or GSD could be random "noise" in the data.
On the other hand, running an extra strategy means extra transaction costs in commissions and spreads. Why should we buy more stocks if it is unlikely to improve our returns? Tell us what you think on the Foolish Workshop discussion board.
This concept of combining non-correlated strategies was first presented by Harry Markowitz, one of the many stellar economists produced by the University of Chicago. (Here in Chicago, we say that "Harvard is the UC of the East.") This concept is called Modern Portfolio Theory, and it earned Mr. Markowitz a Nobel prize in 1990.
One of the precepts of Modern Portfolio Theory is that we can combine assets (or in our case, Workshop strategies) to maximize return at any given level of volatility. For example, if we could tolerate a GSD of 30%, we would combine various Workshop strategies until we found a combination that had the highest return possible without exceeding a GSD of 30%.
Theoretically, we can do this for all levels of volatility, plot those results on a graph, and get a line showing the highest possible returns at all volatility levels. This "maximum return" line is often referred to as the "efficient frontier." Our mission at the Workshop is to boldly go to this efficient frontier and travel along it until we reach the level of volatility that matches our risk tolerance.
That's what we are attempting to do with the Workshop Portfolio, and I hope this is helpful to everyone in constructing your own portfolios. You can learn more about the efficient frontier on the Workshop and Mechanical Investing message boards. I want to especially thank BarryDTO for his work out on the efficient frontier concept.
So, why hasn't everyone started investing in these high-return, low-volatility portfolios? Next week we'll talk about something called the "multiple hypothesis" problem, and look at our strategies with a skeptical eye. I hope to see you then -- same Fool time, same Fool channel!
RSS Headlines
Fool UK
