- Investment Management must be elevated to a Science.
- Quant methods and AI will replace most human PMs.
- Combining strategies boosts performance.
- Elegantly simple math can beat more complex alternatives.
In my new book, I explain a strategy which integrates two strategies my firm has invented: Structural Arbitrage and Hedged Convexity Capture.
My macro view is that unique intellectual property, based upon science, rather than superstition, will drive investment performance in the future. In addition, high quality rules-based strategy indices, marketed through strategy-based ETFs and mutual funds, will eclipse hedge funds due to their transparency, liquidity, and low costs.
I believe that transparent, rules-based strategies will allow investors to take far more control of what risks they decide to take on since they will have the opportunity to evaluate a strategy's reasonableness for themselves, by removing the black boxes that previously shielded algorithms from public view, consideration, or censure.
In the final calculus, the investment industry will move to a more pure form of competition which resembles the marketplace of ideas. And the best ideas can be considered, evaluated, monitored, and chosen or discarded once they are made transparent to all.
Our continual focus on innovation is driven by an obsession to push the performance envelope. And innovation is the product - not any individual algorithm. "Making Money with Math" is not just our mission. It is a value system which chooses evidence-based scientific methods over the superstition and intellectual laziness which threaten to destroy not only personal wealth, but also the modern financial system itself.
Correctly practiced, investment research and management is a scientific endeavor with clear, objective rules requiring zero interpretation which can be proven or disproven using data. That's science. Testing ideas scientifically is not only real work, but is also an ethical responsibility.
The idea behind Hedged Convexity Capture is to benefit from the path dependent negative convexity that has been widely observed in leveraged ETP products in an efficient risk/reward structure which dramatically lowers the potentially ruinous drawdowns of shorting leveraged inverse ETP products. The idea behind Structural Arbitrage is that profits are possible by acting as a synthetic insurance company which sells expensive insurance in the volatility market, then synthetically reinsures that market risk with long duration government bonds.
Combining the benefits of Hedged Convexity Capture with Structural Arbitrage is really combining the advantages of exploiting two different major market inefficiencies.
There are a variety of well known factors which are old, well-worked over, tired, and commoditized which are favored by most investment industry pseudo-intellectuals. Therefore, these factors no longer earn outsized returns. By targeting less well known and understood factors (inefficiencies) which are purely mathematical in nature as opposed to intuitive factors which are well-known to equity fundamental practitioners, or would-be economists, we can achieve outsized returns.
Here are the integrated strategy's rules for testing purposes:
I. Short TVIX (NASDAQ:TVIX) with 25% of the dollar value of the portfolio.
II. Short TMV (NYSEARCA:TMV) with 75% of the dollar value of the portfolio.
III. Rebalance weekly to maintain the 25%/ 75% dollar value split between the positions.
Here are our testing results in a linear scale:
The CAGR is 73.8% and the Max Drawdown is 31.05%, yielding a MAR of 2.37, which is outstanding. In addition, I believe this is the first time that a strategy has been published which has a CAGR exceeding 70% with a MAR exceeding 2.
For another look at this equity curve, let's examine the same strategy's performance in log scale.
Even though the strategy tests well, I never rely on theory alone. The advanced non-public version of this strategy for our clients does not use shorting in order to circumvent the tail risks associated with shorting. Readers who do not enjoy the risks associated with shorting would be well served to only use Structural Arbitrage.
Even though the performance of basic Structural Arbitrage is much lower, it eliminates the open-ended tail risks associated with shorting.
However, examining the strategy's performance in 2011 is especially instructive for those who believe that capturing negative convexity is always riskier than simply going long the S&P 500 (NYSEARCA:SPY). During certain periods of market drops, the strategy actually vastly outperforms passive indexing alternatives. Indeed, the strategy is actually slightly inversely correlated to the S&P 500 during 2011 and outperforms the market by over 80 percentage points!
(click to enlarge)During the choppy market of 2011, the strategy had a dramatically higher Sharpe ratio and a dramatically higher MAR ratio than the S&P 500. This does not mean that this massive outperformance will continue, but it does mean that the strategy is well worth further exploration. The sweet spot for investors is non-correlation combined with a superior CAGR/Max Drawdown ratio, or MAR ratio. The data points to the strong possibility that dramatic outperformance could continue.
But remember, there are risks such as hyperinflation or other events in which both volatility could skyrocket and government bonds could simultaneously tank. This or other risks could make both short positions skyrocket uncontrollably. Because of this, I believe this strategy is interesting as an explication and exploitation of two massive market inefficiencies, but in my personal trading I do not like the open-ended tail risk introduced by shorting. The gentle reader would be well served by implementing extremely strict risk control which turns the strategy on and off to reduce the potential for ruinous losses which could exceed the trader's original investment. Strict risk control could include daily stop loss or volatility limits, hedging with options, or other means the reader deems fit.
My personal choice is to use strategies which do not have open-ended shorting risks at all, even if they have a lower performance. I enjoy making money with math because it allows me to sleep soundly.
Remember that the ETN instrument TVIX used above was for testing purposes because it gives us far more data to work with than (NYSEARCA:UVXY), its ETF counterpart which has not had some of the past tracking errors of TVIX. For actual trading (if one does not mind shorting as I do), the reader might consider using UVXY instead of TVIX. Below is a graph of the same strategy using UVXY. It has a shorter history than TVIX, but the performance is just as impressive.
Again, almost identical performance for common periods with the TVIX, but with a potentially safer instrument. Not bad.
Investors are constantly bombarded with highly complex, but intuitively appealing strategies. And complex intuitive strategies may be excellent marketing vehicles, but they are often far worse performers than non-intuitive strategies. The investment world does not reward complexity. Investors need to realize that effective strategies can be both elegantly simple and mathematically powerful.
Strategies which rely upon an understanding of pure mathematics have a higher probability of sustained outperformance because most people are not wired to feel emotionally comfortable with mathematical strategies. It is this emotional discomfort which not only hinders the popular adoptions of such strategies, but also creates the potential for sustained outperformance by those who do appreciate their logic.
We have examined a strategy which powerfully combines Structural Arbitrage and Hedged Convexity Capture. The takeaway for regular investors is that strategies which are simple to execute often dramatically exceed the performance of more complex, intuitive alternatives.