This is the seventh article in the series that aims to develop portfolio investment approach that 'beats the market'. The goal is to equip the readers both with 'the knowledge about the path' and 'the confidence to stay on the path'.
In the previous articles we have reviewed three theories of investment and how they can help in developing superior portfolios:
- In the first article ' Efficient Market Hypothesis And Random Walk Theory: Buy 'David Swensen's Portfolio', the author recommended using 'Swensen portfolio'.
- In the second article ' Modern Portfolio Theory: Introduce Allocation To 'Alternatives', we suggested ways of enhancing 'Swensen portfolio' through adding alternative assets to the portfolio.
- In the third article ' Tilt Your Portfolio to Achieve Superior Returns', we discussed Noisy Market Hypothesis and reviewed historical performance of small cap and value stocks. We highlighted a list of ETFs that could be utilized to 'beat the market'.
- In the fourth article ' Different Take On A Dividend Growth Investing', we reviewed how dividend growth investing, Dogs of Dow and Core 10 can be used to 'beat the market'.
- In the fifth article ' Will Low Volatility ETFs Save You During Market Turbulence?', we reviewed how low volatility stocks outperform during various market cycles.
- In the sixth article ' Are Momentum Strategies The Panacea?', we reviewed how momentum strategies could add some extra return to your portfolio.
Initial two articles serve as a practical guide to structuring the core portfolio. In the later articles, we started discussing strategies for the model satellite portfolio. Discussion of satellite portfolio is not finished yet. There are at least two more topics that I plan to cover before concluding the series - dual momentum and multi-factor investing (will be covered in the future articles).
In this article, I plan to share with you one of the few strategies / portfolios that could provide pure alpha. To ensure that we are all on the same page, I will be referring to alpha as identified using classic CAPM. One can, however, see that under some alternative models (e.g. Fama-French's 3 or 4 or 5-factor models) the strategy that I propose might still demonstrate some alpha.
60/40 portfolio, isn't it so cliché? What else could be said about 60/40 equity/bond allocation? Perhaps, nothing much. However, the concept of combining a couple of asset classes that exhibit a negative correlation to lower overall portfolio risk is so flexible. So let's take that concept and apply it to less conventional asset class - i.e. volatility.
Some would argue that 'volatility' is not an asset class in its own right. Some can argue that 'volatility' is merely derivative of beta and, therefore, should be viewed as a subset of 'equity' exposure.
I have no desire to argue one way or another. What is important for me is the fact that shorting volatility gives you nice high beta position. Shorting volatility could be achieved through holding a long position in inverse volatility products, such as XIV and ZIV. In this article, we will focus on ZIV (I will discuss XIV in the future articles as part of my other investment strategy).
What is good about VelocityShares Daily Inverse VIX Medium-Term ETN (NASDAQ:ZIV) is beta of ~2 and significant negative correlation with bonds -0.6:
Why beta of 2 is good? It means that you are getting levered exposure to the equity-like asset class. This leverage comes at very attractive implicit all-in funding cost. Compared to other ways of levered exposure to equities (e.g. ITM LEAPs, futures, etc.), single ETF/ETN solution is less operationally burdensome. Why negative correlation with bonds important? Well, you know the answer. Diversification benefits are highest when you combine negatively correlated assets.
So, let's get to it. The strategy is to combine ZIV and TLT as a replacement to SPY. Please, note that the goal is not to replicate SPY, the goal is to take some active risk that pays back nicely (i.e. attractive risk-return profile). To keep things relatively simple, we will focus on standard deviation and maximum historical drawdown to compare risk profiles of ZIV/TLT portfolio to SPY.
In order to keep, the standard deviation of ZIV/TLT portfolio similar to SPY, we would combine ZIV and TLT at 40/60 ratio. Result:
Slightly better risk profile: standard deviation of the portfolio (11.8%) is similar to SPY's (11.6%), while maximum drawdown is better (-11.9% vs. -16.3%). Significantly better return: 17% vs 12.4%
Please refer to the table and graph below.
Are we talking about 4.6% historical alpha?! Of course, we are. And, of course, there is a catch. Haven't we seen enough cases of some strategy working in the past and then miserably failing when we start using the same strategy?
Why do I think that this investing strategy will continue generating alpha going forward?
First of all, I don't think that this strategy actually provided real alpha (or if you want to call it "pure alpha"). Using Fama's argument for creating 3-factor model, then 4-factor model, and I guess the latest one was 5-factor model - if you get alpha, it is likely that your model is not sophisticated enough to capture all the risks. Who told you that beta alone captures all the risks? That's the reason for adding "value", "size", and other factors to the model.
Following the same thought process, there might be inherent risks in investing in volatility products that beta is unable to catch. Perhaps, the market is paying just enough to compensate volatility-buyer for incremental idiosyncratic risk. Such idiosyncratic risk could be seen in the cases when market overall might be going up while increased market volatility might be still pushing down the price of ZIV.
Secondly, by engaging in a long position in ZIV, investor/speculator should understand that ZIV/TLT mix might provide better historical risk-return statistics, however, that is not an indication of the future performance.
The last point, and, perhaps, the most important one - instead of focusing on historical performance to make inference about future, I would suggest:
Investors should consider asking themselves whether inclusion of additional asset classes with less than perfect positive correlation is useful (hint: of course, it is). For more on diversification benefit, please refer to the graph below. Clearly, the inclusion of ZIV was helpful in shifting efficient frontier out! Speculators should consider asking if barbell strategy might be a better option for them: i.e. instead of exposing yourself to average "equity" risk, it might make sense to have a sizable portion in safe investment (i.e. TLT) and the rest in very volatile ZIV (pun intended). Mixing safe heaven investments with risky ones should provide a better estimate of maximum potential loss. This barbell idea of investing was well explained in Nasim Taleb's books, which I strongly recommend .
To put is simply, I recommend you consider TLT/ZIV portfolio only if you either think that volatility is a very useful asset class for further diversification of your portfolio (see bullet point above addressed to investors) or if you like gambling (see bullet point above addressed to speculators).
This, also, means that I strongly discourage you from reading the rest of this article if you are contemplating the use of TLT/ZIV portfolio only because of its historical performance.
As alluded to above, historical analysis covered the only period since 2011 - ZIV was around only for the last 5 years. This means that it was not possible to capture period covering 2008 recession. There is a significant likelihood that ZIV might be a disastrous choice of investment during market turbulence.
Additionally, as with any other highly volatile investment, the market gyrations might test your resolve and can have other side effects (e.g. loss of sleep, problems with stomach, permanent loss of principal).
And, of course, as with any investment/speculation, it always pays to have a clear game plan and limit your total exposure. It is not clear whether short volatility position is a suitable investment for the long-term investor. Perhaps, we would need to see how ZIV performs during next upcoming recession before being able to make any preliminary suggestions.
At this point, all we can do is ask following questions:
- What type of risks I'm exposed if I put together ZIV/TLT portfolio?
- Am I being compensated fairly for those risks? I will leave it up to you to answer those questions. At the end of the day, when ZIV/TLT portfolio will start going through turbulence - it will be your resolve and conviction that will be tested.
My personal view: volatility is an asset class in its own right and I would do a disservice to my net worth by not having some exposure to it in my investment portfolio.
Mix 4 parts ZIV and 6 parts TLT:
- ZIV (expense ratio 1.35%, ETN issued by Credit Suisse, AUM $70M, inception November 2010).
- TLT(expense ratio 0.15%, ETF by iShares, AUM $8B, inception July 2002)
Few takeaways: ETN structure means that you are exposed to Credit Suisse's credit risk in addition to numerous other risks. The expense ratio of 1.35% is quite large in nominal terms: one needs to consider both the price paid and also the value received in exchange.
Increasing rebalancing frequency did not materially improve the performance of the portfolio. If anything, rebalancing too frequently resulted in lower historical returns. It seems that quarterly or annual rebalancing should work fine.
Don't forget to "follow me" so that you don't miss the future articles. Also, don't forget to check out previous articles:
Article #3: Tilt Your Portfolio to Achieve Superior Returns
Article #4: Different Take on A Dividend Growth Investing
Article #6, Are Momentum Strategies The Panacea?
Note: I don't claim to be the first person who has thought of using inverse ETN products in their portfolio. Here is just a few of the well-known Seeking Alpha contributors who discussed such products: Chris DeMuth Jr., Goombarh and Bill Luby (published as early as 2011). In a recent article David Easter discusses mechanics of actual ETNs, debunks some misconceptions and explains why inverse ETNs provide very promising potential. Additionally, there are other authors (e.g. Harry Long, Fred Piard) who explored trading and portfolio strategies involving such products.
Disclaimer: Nothing in this article should be viewed as investment advice. Reviewed strategy is likely to be unsuitable to most of the investment community. The author does not assume any responsibility for the actions of the reader.
Disclosure: I/we have no positions in any stocks mentioned, but may initiate a long position in ZIV, TLT over the next 72 hours.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.