Several years ago I began to tinker with the typical portfolio construction models offered to clients of the banks I worked at. In training, we were told that the best models were those that allocated money to active, established money managers such as Franklin Templeton, MFS, etc. Time taught me that even the Bill Miller’s of the world run into walls, causing investor distress due to the blind trust that often exists in the asset allocation world.
That frustration led me to become skeptical of the magic-elixir-peddling mutual funds of the world. It wasn’t just the inconsistency, it was also the cost and the time lost in transaction execution, not to mention the excessive fees that go to the pockets of those that provide little alpha to investors.
I guess my focus has developed into an unending search for alpha. For the uninitiated, alpha is the excess return provided by your investments over an index. For example, if your portfolio gained 10%, while the S&P 500 gained 5%, the resulting alpha is 5%.
Since I live in the Republic of Panama, over time I have been exposed to investors with benchmark indexes such as MSCI World, S&P 500, local deposit rates, and pretty much everything under the sun. What becomes abundantly clear is that the ideal investment return is different for everyone, but more recently is typically higher than the returns provided by the S&P 500.
Investors seem at times to have short memories and more recently have become somewhat accustomed to lower returns. This is the result of the low rate environment that we live in and the dis-inflationary pressures that have taken hold in many parts of the world. Rampant inflationary pressures exist, but mostly in countries such as Venezuela and Brazil, where the FX rate is crumbling versus the US Dollar and the state apparatus is disease-ridden.
Still, recent investment returns in the US and Europe are either negative or flat, depending on the time frame you use, and there is little conviction that substantial economic growth will come back any time soon, despite the heights at which the US equities find themselves. Since June of 2009, the S&P 500 and Nasdaq Composite are up 117.7% and 155.09% respectively, while the MSCI World is up 59.32%. This is great, but US equity returns since the beginning of 2015 have gone absolutely nowhere and the MSCI World is down -4.25%.
Risk On or Risk Off
So, how does one create value for investors in a bull market that many believe has run its course? This is the question that led me on a long analytical journey, studying correlations between asset classes and exchange traded funds, in addition to reading white papers produced by people such as Charlie Bilello and Michael Gayed, of Pension Partners. If you have the time, I strongly recommend you read some of their award winning research, it is well worth it.
During the process of research, I realized that complexity leads to the creation of successful strategies, but also can be burdensome for average investors. The information processing power is not available to everyone and good returns can be destroyed by factors unknown or unaccounted for in the original research. Therefore it became clear to me that simplicity was the key, especially if the strategy is to be understood by less experienced investors.
Answering the next question became my focal point: What assets or indicators are effective signals of “Risk On” and “Risk Off”? At the end of the day, investors know that markets are cyclical, but given the cacophany of opinions provided by all sorts of media, whether it be on-line, printed or cable, it is difficult to trust any “expert”.
Personally, I get more from the viewpoints of “independents” such as Hedgeye, Danielle DiMartino Booth, Mark Yusko, Jawad Mian, Jesse Livermore, Emad Mostaque and Ben Hunt. I make a conscious effort to listen to people that have had important roles in banks or government and have translated their previous experiences into leadership as independent thinkers and money managers. The latest financial crisis and the term “Too big to fail” was confirmation that big banks and the political class in the US are too intertwined, the small investor being the victim of existing conflicts of interest.
To arrive at an answer regarding what assets best represent “Risk On” and “Risk Off”, one must compare asset returns over different time frames. A”Risk On” asset is one that provides exposure to the best growth opportunities, while a “Risk Off” asset is one that provides shelter from a downdraft in “Risk On” assets. After identifying these assets, one must have an indicator that signals when to be invested in risky assets and when to take shelter. Ideally, this is a binary decision, which should be easily executed, simply swapping one for the other at a given point in time.
First let’s compare the price returns of the most readily recognized indexes and assets:
What becomes abundantly clear is the out-performance provided by the Nasdaq in bullish periods and the relative tranquility provided by longer dated US Treasury bonds in periods of market stress. In fact, global stocks rarely, if ever, provide attractive returns relative to other well known US indices. Admittedly, there are many exceptions, but your typical investor does not have time or access to reliable information that would allow them to do appropriate due diligence, which is necessary to know when to buy or sell a particular asset. The sole stated purpose of my research is to identify which assets provide acceptable returns over the long run and shelter in times of stress, therefore it becomes readily apparent that the Nasdaq Composite and longer dated US Treasury bonds fit those roles quite well. After much analysis of correlations and returns, these became the two pillars of “Absolute Return”.
Description of Absolute Return
This investment/trading strategy uses two exchange traded funds only: PowerShares QQQ Trust (QQQ) and iShares Barclays 20+ Yr Treasury Bond (TLT). In terms of daily liquidity, the QQQ has an average daily volume over 20 million shares traded and the TLT averages over 6 million shares a day, which ensures more than adequate liquidity for the expression of this strategy.
The QQQ is an index ETF, which is comprised of 100 of the largest domestic and international non-financial companies in the Nasdaq Stock Market. As you may be aware, this list is heavily weighted in technology, consumer discretionary and healthcare, but also has exposure to consumer staples, industrials and telecom services. The TLT is also an index ETF, but its sole focus is the purchase of US Treasury bonds with a weighted average maturity of over 26 years.
Many ask why I prefer long dated paper in moments that I have to reduce portfolio risk, especially when the US Federal Reserve is bent on raising rates. The simple answer is that , in comparison, the other “safe” assets provide little or no return: cash yields close to nothing and shorter dated treasuries are more sensitive to short term changes in interest rates. There are moments in history where stocks and short dated bond prices are closely correlated, which would defeat the purpose of this exercise. In the following chart, it is clear that the 30-yr Treasury bond yield (pink) moves less proportionally than 1-yr treasuries (blue) when Fed Funds rise sharply (green).
Add to this my view that the Federal Reserve will not raise rates very much in the future, since the long term trend remains downward. Needless to say, I have a long argument for this case, but that is a discussion for another time. So far, I have been correct.
As far as the mechanics of the strategy, every Friday prior to market close, I run a proprietary algorithm that indicates which of the two assets I should be invested in for the following week. There are weeks that require no change from the existing position, but at times I will have to sell one and buy the other, effectively swapping out of one and into the other. On average, there are 10-11 swaps per year. In 2015, the strategy required 10 swaps, while in 2016 the algorithm has signaled 4. At the time of this writing, the strategy is 100% “invested” in TLT for the third consecutive week.
Results of Absolute Return
In creating the strategy, I ran into some limitations when back-testing the data. In order to present valid comparisons it would be ideal for the historical data to go as far back as possible. Unfortunately, the MSCI All Cap World Index began on the 14th of July of 2003. The iShares Barclays 20+ Yr Treasury bond ETF began trading on the 26th of July of 2002, whereas the QQQ began trading in 1998.
Presented with this challenge, I analyzed the price returns provided by the two most common indices represented by QQQ and TLT, the Nasdaq 100 and the CME CBOT 30-year US Treasury Bond Price. Using the 11th of November of 1985 as a starting point, the algorithm returned a total of 1398.28%, or 9.33% annualized. Again, this compares favorably to the 30 yr long bond, which returned 3.25% annually, while the S&P 500 returned 7.97% and the Nasdaq Composite returned 9.42%.
At first blush, it seems as though an investor would do just fine leaving his/her money invested in the Nasdaq and never bother with any rotation of assets. Still, investors are easily spooked by increased volatility in stock markets, and justifiably so. The financial crisis truly put the global financial system on the edge of a cliff. I would argue that policy makers simply kicked a major can of worms down the road and the US will eventually have to deal with that. Recessions generally occur every 7 years and falling stock prices create so much uncertainty, it is wise to have a strategy that one can trust, instead of making hurried decisions in periods of stress.
The main purpose of the Absolute Return is to provide a return that supersedes the principal benchmark of the investing world, the S&P 500, creating Alpha.
The model uses an initial investment of $100,000. It is important to remember that the return calculations are based on price only. The TLT has a 2.65% distribution yield (dividend) and the QQQ has a distribution yield of 1.2%. If your brokerage account costs over 1% in annual fees, the strategy will still provide you with a return that exceeds what most money managers or hedge funds can provide on a consistent basis, certainly much better than the individual indices by themselves.
What is obvious in the graph is that the equity indices and Absolute Return may be both positively and negatively correlated at times. For example, in 2008, Absolute Return gained 11.6%, while the S&P 500 lost -38.8%. In 2009, Absolute Return and the S&P 500 both gained over 21%. The algorithm is primarily momentum based and there have been periods where equities provided better returns, but similar to the true purpose of hedge funds, Absolute Return is designed to provide positive returns when equity markets are in free fall, shelter from the storm.
That is when the model generates the most Alpha, as evidenced in the returns provided in 2015. Absolute Return provided 19.8%, while the S&P 500 was flat, the Nasdaq rose 7% and the MSCI World returned a negative -3.4%. The best calendar year for Absolute Return was 2012, with a return of 27.9%. Its best trailing 12-month return was +44.18% (52wks ending 9/9/11), while the worst trailing 12-month return was -18.69% (52wks ending 11/14/08). Over time, the strategy has a Sharpe Ratio of 7.97 and the 36 month Average Deviation of the Mean is 1.31%, which I believe is more than acceptable in this low rate environment.
For Further Info
Weekly results of the algorithm will be displayed every Friday before the market close as a pinned tweet on my Twitter profile, @Charlie_Ledezma. If you want more information regarding the construction of this algorithm, feel free to contact me directly via DM on Twitter.