- I’ve recently learned that VIX® is trademarked and CBOE has an issue with the name of my Marketplace service. The name will change; the metric will still drive algorithms.
- So effective now, The Easy VIX will be called “EZV Algorithms.” The name conveys the original legacy as well as the quantitative underpinnings of the portfolio-management tools offered.
- With the name change, a reintroduction describing services is appropriate; it’s been a busy year so the opportunity to revisit offerings is probably a blessing.
- I’ll provide a view of how members can choose from simple to complex applications of the various algorithms that drive the service. Emotion-free, quantitative rules.
At its core, EZV Algorithms is built around risk-mitigation protocols where avoiding the big drawdowns lowers risk and creates exceptional returns by harvesting the rebounds, starting from an undamaged investment base. That's more realistic than hoping to identify each up and down in the market.
The core algorithm is built around tracking various metrics derived from the VIX futures curve, also called the VIX term structure. The algorithm looks at the shape of that curve, the rate of change over various look backs, and a few subordinate factors like VVIX and the SKEW index. The service also offers algorithms for allocating holdings among ETFs and other supplemental protocols to further enhance returns.
By avoiding about 1.5 material downturns per year, returns can be 2.5 to 3 times normal, and drawdowns constrained to low-single-digit percentages in the worst of all rolling 252-day periods. But be patient; sometimes months go by with nothing to show from the algorithm other than providing the confidence to ride the upward momentum. Think about how often you’ve sold when driven by an emotional reaction, only to miss out on a continuing rally. And consider that the strongest rallies follow the most fearful times. A confident ‘hold’ signal in the face of an emotional roller coaster is a powerful thing.
There are some false signals, small, quick sell intervals that average out to nothing gained or lost. That’s the small burden that enables avoidance of big losses and the tripling of gains.
Results have been replicable since the service’s October 2019 inception. That’s because EZV Algorithms uses sequential trailing optimizations to continuously calibrate parameters. Think of it as an artificial intelligence (“AI”) approach. The reason optimizations are "trailing" is because decision criteria for today is based on data through yesterday. So, the model has performed about 3,150 optimizations over nearly 13 years without fitting anything using contemporaneous data. If it worked 3100+ times, there's a good chance it will work next week.
Some modelers build a test data set and then utilize the parameters for the future. In doing so they have a process that relies on a single sample set, so the parameters are calibrated once but the process is not. In contrast, the trailing optimization process is organic, always building on updated trailing data. So, parameters are calibrated, and the process is as well. Overfitting is a big concern when your process is a massive test data set because whatever parameters you build into the structure could produce an accidental result for the single calibration. If a process is replicable 3100 times, it's unlikely to be an accident.
EZV Algorithms – Simple To Complex
At EZV we’ve got a diverse group of members. Some are seasoned traders willing to put in material effort, and some just want an easy safety-first approach for managing their retirement investments. Some look for a thorough understanding of the EZV methodology and metrics with all their nuance. Others look for simple buy-sell signals and have no interest in further complexity.
This post will reconcile those different perspectives in a way that fulfills the needs of each investor type. To do so, I’ll categorize the service offerings and the different levels of effort necessary to implement them from basic to complex; I’ll also describe the component algorithms.
First a brief roadmap in table form, where the degree of effort is color-coded from requiring little maintenance in green shades to more active approaches in yellow and gold.
Source: Michael Gettings
The service relies on trading buy-sell signals from a core algorithm, and that is enhanced by supplemental algorithms for those who care to put in some extra effort for greater rewards. The discussion below has been grouped into three categories, ranging from low maintenance, the green tiers, to those who are looking for the full monte and care to put in the extra effort.
Tiers 1 to 3: Core EZV Algorithm (Various Green Shaded Functions)
The core service consists of binary signals – buy and sell. By trading SPY and taking shelter in IEF on any sell signal, the algorithm produces these results from May 2008 through November 2020:
For other S&P-correlated ETFs, results improve. The two tables below show the modeled results for QQQ/SPY/IWM starting in May 2008 and for 3X-leveraged variants, TQQQ/SPXL/TNA, since trading began for TQQQ on Feb 11, 2010.
Source: Michael Gettings
Source: Michael Gettings
Note 1: Leveraged positions are typically deployed as 25% to 30% of holdings to cap exposure.
Note 2: The later start date missed the large returns in 2009 but leveraged returns have exceeded 3X unleveraged returns as volatility ‘decay’ turns positive when avoiding drawdowns.
Following the model since the October 2019 EZV inception, results have far outperformed the 13-year numbers because COVID, for all its disruptions, provided a great opportunity for the algorithm to excel. The numbers are almost too good to publish, but here they are:
Source: Michael Gettings
I hope we won’t have a COVID-type disruption often, so I would expect the 13-year modeled numbers to be a better benchmark for the future although there are never any guarantees.
At the Category 1 level, the only requirement is to follow the buy-sell signals which are communicated at irregular intervals about 12 times annually by text alert or email. It’s a simple approach, but if you get cold feet on a buy signal, like the buy signal that went out on that fearful November 3, 2020 election day, you’ll need a bit more understanding. That requires some knowledge of what constitutes a signal. So, I’ll take a slight detour and explain what I would do if trying to get back on track.
I publish a Daily Dashboard for members which is updated automatically at various intervals from about 6:00 AM to 5:30 PM eastern time. In that dashboard one can see each metric that could contribute to a buy or sell signal and a brief description of how a signal might be generated. Sell signals are generated by some combination of breached metrics, i.e., metrics that worsen to the point of falling below an AI-calculated trigger.
The status of each metric is reported in the dashboard; that status is expressed as the percentage of each metric’s historical range that, if traversed, would trigger a breach. My personal rule is this: if any single metric is within 2 percentage points of breaching and that breach could trigger a sell signal, I won’t add positions. I’ll wait for an improvement or I’ll wait for the next sell-buy cycle to play out. Otherwise, it’s likely that adding a new position will work out. As always, there are no guarantees, but it seems to be an effective rule of thumb.
Having said that, each sell interval is akin to a coiled spring of market energy and buy signals are often followed by very rapid price increases. So, it’s better to buy immediately on a signal.
Tier 4: VaR-Attenuated Sell Signals (Yellow Shaded)
Risk exposures can be assessed by measuring transient price volatility and the related Value at Risk (“VaR”). It’s a well-established methodology published by a JP Morgan affiliate about 30 years ago and it has become a standard risk management tool.
I’m constantly searching for a bigger advantage. False signals are a small annoyance since the sum of advantaged and disadvantaged results come to zero, but I’ve tested various means of further validating signals to improve performance. One successful attempt identified how VaR metrics could be used to hold partial positions on a sell signal, allowing another day to validate the signal. By following a simple protocol on each sell signal, selling a portion of holdings consistent with a threshold risk tolerance, returns can be improved using SPY or the other ETF baskets, leveraged or unleveraged.
The attenuation protocol requires the partial exchange of equity ETFs for IEF on the day of the sell signal, and then a day later, either exchanging the rest of equity ETFs or reestablishing original holdings depending on how metrics have changed. It’s only a small increment of trading effort for slightly more active investors. Members have the details on performance but suffice it to say the additional effort results in a commensurate increase in performance for those who want to undertake it.
Tier 5: Weekly Rotation of ETF Allocations
This requires weekly effort because it requires reallocating equity allocations at the margin.
The allocation model uses daily trailing optimizations comparing recent price performance and risk factors to decide on relative holdings. Price trends are measured over two look-back periods. The longer period establishes smoothed allocation ratios which are then fine-tuned using the short-period price performance and risk factors. Risk conditions are the major factor; they are quantified using an exponentially weighted average of the spot values for VXN/VIX/RVX to gauge risk for QQQ/SPY/IWM respectively.
Once again, the performance increment can be material for those willing to expend the weekly effort required.
That is a summary of how the service has evolved since its October 2019 inception. It started as a simple binary approach, alternating equity ETFs and IEF, but thanks to the dynamic chat room and EZV members’ appetite for pushing the envelope, innovations have been constant.
On a personal note, I use the leveraged ETFs as 25% to 30% of holdings. That was a big change because using those in minority allocations caps losses and allow most holdings to be deployed in bonds, preferred stocks and yield-oriented covered calls which tend to dampen volatility. The empirical results also indicate that the EZV algorithm accelerates returns beyond the 3X leverage ratio. Research indicates that when you can screen out the bad times and ride the good times it turns volatility decay on its head.
So, that describes the evolution of the service in the first year. It will continue to evolve. I expect to look at trading UVXY and SVXY in coming weeks and we've already spent time with various options strategies. The members' chat room is a lively place for those who enjoy robust discussions of intraday developments and anything else trading related. If you'd like to consider signing up, the all-5-star reviews might provide some flavor from some of the members.
Happy trading! -- Mike Gettings
Analyst's Disclosure: I am/we are long TNA, SPXL, TQQQ.
I trade a basket of ETFs and any tickers mentioned using the algorithm described as well as other analyses. The algorithm monitors daily performance and periodically recalibrates parameters and triggers in a stepwise sequence. New calibrations are applied prospectively only, and not to the historical period from which they were derived. The algorithm described and the discussions herein are intended to provide a perspective on the probability of outcomes based on historical modeled performance. While I track one or more reference portfolio(s), I make no recommendations as to specific investments. I reserve the right to make changes to the algorithm as I deem appropriate. Neither modeled performance nor past performance are any guarantee of future results.
Seeking Alpha's Disclosure: Past performance is no guarantee of future results. No recommendation or advice is being given as to whether any investment is suitable for a particular investor. Any views or opinions expressed above may not reflect those of Seeking Alpha as a whole. Seeking Alpha is not a licensed securities dealer, broker or US investment adviser or investment bank. Our analysts are third party authors that include both professional investors and individual investors who may not be licensed or certified by any institute or regulatory body.