Several years ago, the Logical Invest whitepapers first introduced me to rotation investment strategies. After some research, I subscribed to the MYRS (the “Maximum Yield Rotation Strategy“), with good results. It was the first paid investment subscription I had ever used. I moved on to another service to learn options trading, and did well for two years, until a number of sharp setbacks made me come to terms with my investment goals.
After further reflection, here is what really matters to me:
- Consistency. One must have clear goals and consistent methods for selecting, entering, and exiting positions. Emotions make for very bad decisions.
- Efficiency. I work two jobs and love spending time with my family. I can’t spend hours each day actively managing investments and assessing risk.
- Risk management. I need a method that consistently redistributes risk, and that doesn’t rely on my own knowledge or intuition, which are so often flawed. It is too easy to chase gains and too difficult to recoup the inevitable losses.
- Objectivity. I am not in finance, and I don’t have the knowledge required to manage investments in individual stocks, where risk is difficult to predict at best. Programmatic investing substitutes objective rules for guesswork.
- Simplicity. I don’t want to manage too many investments at the same time. I like the aesthetic appeal of simplicity, which has the added benefit of limiting transaction costs.
- Results. I hate drawdowns, especially ones that take a long time to recover. I don’t mind volatility, as long as it slopes up. I love gains. Big gains. Go figure.
So, that’s it. I need a fast, precise, powerful, scientific method shown to perform through thick and thin. Only an automated investment strategy can hope to fulfill these requirements.
In the time since I last used Logical Invest services, they have expanded their offerings to include several different and interesting approaches, and for some time, have been discussing a “strategy of strategies”, which I found very intriguing. After a long wait, they finally released QuantTrader (QT) rather than publishing an SoS solution. This is a far more powerful tool, since it not only provides access to all of their offerings, but also allows for full customization and testing. For example, I’ve been able to build a solution for my company’s 401(k) offerings, which are extremely limited, with acceptable results.
Starting with QuantTrader
I started using QT during a free trial period, during which I learned how to use the software. There isn’t a big learning curve – there are just a few screens used to configure settings, manage symbols, build portfolios, and backtest strategies. It isn’t pretty, and it can be unstable, but that’s typical for new software. The focus isn’t on UI/UX or QA yet. Rather, it is designed with a singular purpose: to build and evaluate rotation strategies, and that it accomplishes well. It’s promising that in 2 months, I’ve already seen 3 releases with bug fixes, and the development team is openly soliciting feedback.
While QT supplies a number of built-in strategies, I immediately set about building my own. Using the existing strategies as a benchmark, I’ve back-tested dozens of combinations in an effort to improve upon them. Most of the time spent with the software is testing various combinations, adjusting parameters, and re-tuning. I’ve invested time over the past 2 months off-and-on with it, and have developed a solution with outstanding results. Even so, I’m still unsure of the meaning of certain settings and parameters. The help documentation is limited, but it is enough to get things started. The dozens of failed experiments help develop a sense for how various combinations and settings will work together. I mostly resorted to brute force, repeatedly backtesting combinations that hold my interest.
Guidelines for Strategy Development
A few general rules are important when developing strategies. You must have a basic understanding of the built-in strategies and ranking algorithms, and use creative combinations. Typically, strategies with dissimilar approaches but similar volatility and returns pair well. Keep the tuning tweaks to a minimum, and only change one at a time. Backtest repeatedly. Look for solutions in the optimization window that are stable, meaning there aren’t large step-changes if you look a few cells in any direction. I tend to avoid overly short or long lookback periods. Sometimes, changing the rebalancing period or volatility limit can have positive results, but usually it’s best to leave those settings alone. Let the optimization choose the lookback period and volatility attenuator for you, but don’t be afraid to choose options with slightly lower performance that are more stable. Change up the ranking method and number of ETFs to test any available combination, as well as the maximum and minimum allocation percent. More recently, I’ve also found it useful to adjust the maximum allocation percent for an individual investment in the Portfolio Manager, or to adjust the volatility multiplier in certain circumstances.
To be honest, none of this needs to be done to achieve an acceptable result. For example, I have very few investment options in my 401(k). Essentially, I can choose SPY, AGG, or cash. By adding the two to a portfolio, and backtesting using the SRRP ranking with no adjustments whatsoever, a 12% CAGR results from 2007 to present, versus 7.2% for SPY, and maximum drawdown is reduced from -55.19% to -15.02%.
The Morpheus Strategy
My current champion strategy, a simple but powerful variant of what is now being referred to as The Beast in the QT forum, rotates monthly between two distinctly different sub-strategies, each of which is a meta-strategy composed of 2-3 different Logical Invest core strategies. Both sub-strategies rotate monthly using the “SR” ranking method, with maximum allocations of 60-70%, so they are always each invested in either 2 or 3 core strategies.
The first sub-strategy, The Blue Pill, offers a lower-volatility baseline. It combines the BRS and GLD-USD core strategies, ranking them by the SR top 2 method, with 60% maximum and 0% minimum allocations. The strategy uses monthly rebalancing and a short lookback period of 24 days. Each of the underlying strategies has also been modified. The BRS strategy uses monthly rebalancing with a 78 day lookback period, SR top 2 ranking with maximum 60% and minimum 40% allocations. The GLD-USD strategy uses monthly rebalancing with a 48 day lookback period, DR top 1 ranking with maximum 100% and minimum 0% allocations. It also uses a -150 mean reversion weight with 4 day mean reversion period. The Blue Pill achieves a 5 year CAGR of 12.4%, Sharpe Ratio of 2.06, volatility 6%, and maximum drawdown of -5%.
The second sub-strategy, The Red Pill, combines 3 aggressive market-risk core strategies, Nasdaq 100 Hedged, MYRS, and UIS SPXL-TMF 3x Leveraged. It ranks them using the SR top 3 method, with 70% maximum and 0% minimum allocation. Each of the underlying strategies has also been modified. The Nasdaq 100 Hedged strategy uses monthly rebalancing with a 26 day lookback period, and ranks the top 2 ETFs using the SR method. The strategy allows 100% maximum allocation and 0% minimum. The MYRS strategy uses semi-monthly rebalancing with 42 day lookback period. It uses the SR top 2 ranking method with 100% maximum and 0% minimum allocation, and uses a -100 mean reversion weight with 3 day mean reversion period. The UIS SPXL-TMF 3x Leveraged strategy uses monthly rebalancing with 72 day lookback period, and uses the SR top 2 ranking method. It allows 100% maximum and 0% minimum allocation, and uses a -100 mean reversion weight and 4 day mean reversion period. The strategy uses monthly rebalancing and a lookback period of 122 days. The Red Pill achieves a 5 year CAGR of 53%, Sharpe Ratio of 2.7, volatility 19.6%, and maximum drawdown of -12.7%.
Morpheus, the top-level strategy, allocates monthly between The Blue Pill and The Red Pill, ranking them by the SR top 2 method, using a 144 day lookback period. It uses a maximum allocation of 60%, and minimum of 0%. The level of risk exposure can be reduced by choosing a lower maximum allocation percentage, or by applying a volatility limit. At 60% maximum allocation, the strategy achieves a 5 year CAGR of 35.8%, Sharpe Ratio of 2.7, volatility 13.1%, and maximum drawdown of -7%. Its 1 year CAGR is 52.2%, Sharpe Ratio of 4, volatility 13.1%, and maximum drawdown of -4.3%.
I call this strategy Morpheus, because he is the Greek God of Dreams, and because he found Neo in the Matrix, and let him choose the blue pill or the red.
“After this, there is no turning back. You take the blue pill—the story ends, you wake up in your bed and believe whatever you want to believe. You take the red pill—you stay in Wonderland, and I show you how deep the rabbit hole goes. Remember: all I’m offering is the truth.”
The Blue Pill combines two very different core strategies, the BRS (Bond Rotation Strategy) and GLD-USD, a clever method to invest in gold. The optimization results are extraordinarily stable. It outperforms BRS in CAGR, volatility, and drawdown, and dramatically outperforms USD-GLD in volatility and drawdown. The strategy is simple, and uses at most 4 concurrent investment symbols, though the symbols do frequently rotate.
The Red Pill combines 3 aggressive market-risk strategies with very dissimilar approaches. The Nasdaq 100 Hedged strategy is the main driver of growth. It invests in the top 4 symbols from the Nasdaq 100 in equal portions, and hedges that investment against TMF, a triple-leveraged 20+ year US Treasuries ETF. The strategy returns astounding results, but has worrisome levels of volatility and drawdown that would shake out all but the most stoic of investors. The second strategy, the MYRS, invests in an inverse volatility ETF that was the subject of Frank Grossman’s first Seeking Alpha article, ZIV. That ETF is also hedged against TMF for results almost as impressive as the Nasdaq strategy. The third strategy, UIS-SPXL-TMF 3x Leveraged, invests in SPXL, a triple-leveraged S&P 500 index ETF, which is also hedged using TMF. The combination of these three different approaches retains most of the CAGR of the Nasdaq 100 strategy, while reducing volatility and drawdown below any of the strategies alone. The Red Pill is slightly more complex than The Blue Pill – it is always invested in 5-7 different symbols. The optimization results are less stable, but are still extremely good, given the underlying volatility.
I have since gone on to develop several improvements to Morpheus that outperform these results, with the most complex implementation called (of course) The Matrix, which achieves a 5-year Sharpe ratio of 3.1. That strategy, however, is for another day.
Morpheus Current Allocations
The Morpheus strategy, then, invests in 9 to 11 symbols that will rotate fairly regularly on a monthly basis. One of the underlying strategies (MYRS) rotates on a semi-monthly basis, and so must be tended to twice as often. The monthly trading costs should be reasonable, but when added to the subscription cost, it sets a floor on the reasonable investment amount you might expect to manage.
The current allocations for this strategy are:
- The Red Pill: 60%
- The Blue Pill: 40%
Notice that at each level, the allocations sum to 100%. To find the actual allocation to each market symbol shown in bold, all of the superseding percent allocations must be multiplied together. For example, CSX would be 25% of 70% of 20% of 60%, or 2.1% of the portfolio, while the net TMF investment would be (30% of 20% of 60%) + (20% of 70% of 60%) + (30% of 10% of 60%), or 13.8% of the portfolio. I’ve built a spreadsheet to help me calculate these percentages and determine the buy/sell signals for each symbol, which is proving to be a very useful tool when the time comes to execute the trades. QuantTrader does have integration with Interactive Brokers, and it looks like the trades could be executed through the IB API from QT directly, but I have not yet moved my funds to IB or experimented with those features.
Summing It Up
Overall, I am quite happy with QuantTrader. It allows the rapid prototyping and evaluation of simple investment strategies that can achieve results that would be impossible even for some hedge funds. I look forward to continued experimentation, and to possibly helping shape some of the feature set and improving its stability. It is a democratization of investing, providing a simple, powerful, and inexpensive tool for achieving investing results we might have once only dreamed of.