The Easy VIX: Overblown News Added To Risk Perceptions; Next Week Is Pivotal

by: Michael Gettings

The VIX futures algorithm puts the news in perspective and provides navigational aids for risky times. The coming week will be a pivotal.

On Wednesday, Jerome Powell gave a presser and traders, trained by ten years of Fed largesse freaked out, but the VIX futures curve indicated “no big deal.”.

On Thursday, the morning rally confirmed the transient nature of the Powell sell off by reclaiming the prior-day losses.

Then the tariff tweet happened, and despite having been telegraphed for months, the metrics suddenly indicated the risk is to be taken very seriously. Psychology trumps substance.

No sell signal yet, but a material downside opening on Monday will probably trigger a sell;this week will be pivotal.

On Thursday, right after lunch, I posted this comment to an article by The Balance of Trade regarding the market’s response to Powell’s presser:

“Yesterday seemed like spoiled brats having a temper tantrum. I guess it's what happens when too many investors have lived off Fed largesse for ten years. Powell cut rates by 1/4 point and announced he'd end balance sheet reductions next month, but he had the audacity to retain his options for later decisions. What nerve! Thankfully adults seem to be back in charge today. It was interesting to me that despite the transient rout, my risk metrics, keyed off the VIX futures curve, barely flinched. I took comfort in that and today seems to confirm the point.”

I had looked at my positions a short time before writing that, and they had reclaimed all of Wednesday’s losses plus about 50% more (fixed-income positions included both days). I felt good about it, not just because of the gains, but because I never felt that sickening urge to get out when the market was collapsing. I had checked the VIX metrics during the Wednesday free fall, and all indications were that this was a temporary overreaction, likely algorithm-driven selling because Powell’s words triggered some panic in cyberspace.

Then I checked the tape again and chaos had broken loose. Despite Trump’s months-long threats to do so, his announcement of coming new China tariffs sent my ETFs (SPY, DIA, QQQ, IWM, SSO) reeling. So, I updated the Easy VIX metrics, and found that this time the risk was real; i.e., no sell signal yet but getting close.

When I consider the fundamental issues, I don’t see what all the excitement is about on either issue. The Fed actions were pretty much as expected; traders seemed to take issue with Powell’s word selection. And the prospective impact of a 10% tariff on $300 billion of Chinese imports also seems overblown. The $30 billion tariff impact amounts to about .1% of US GDP, and surely the effect of that will be mitigated by supply chain adjustments. If the .1% had been reflected outright in GDP, a Q2 GDP of 2.0% rather than 2.1%, would the market have reacted as badly? I doubt it.

I’m sure some reader could make a case as to why I’m wrong on the fundamental impacts, but in the end it doesn’t matter. The market psychology has made a material move toward a risk-off attitude, and markets move on psychology.

To put a picture to the recent change in attitude, here is the progression of VIX forward curves from July 30 before the Powell press conference to the 31st (after it), to August 1 following the tariff tweet, and finally Friday’s close.

The Progression from Risk On to Risk Off

Source: Michael Gettings Data Source:

Notice that while the VIX level rose following Powell’s press conference, the more-important contango shape of the forward curve remained largely unchanged. That changed on August 1st with the tariff news, and by midday on August 2nd the forward curve was clearly backwardated. The midday values came within a whisper of triggering a sell signal, but by the market’s close on Friday August 2nd the curve had flattened, indicating hold. The progression reflects a rapid change from risk-on to risk-off psychology, and then a pause. Any material deterioration early next week will trigger a sell signal.

The Easy VIX dashboard shows the progression to a potentially imminent signal as of Friday’s close; the SHAPE is no longer in a preemptive “safe” condition, and the Confirming Slope is what stands between the current “hold” and a potential sell signal:

Easy VIX Dashboard, August 2, 2019 Close

Source: Michael Gettings Data Source:

Readers might notice a change at the bottom of the Dashboard. By postulating changes in the SHAPE and Slope metrics I can estimate the amount of change necessary to trigger a sell signal. The note – “Proximity to Sell Signal – Imminent” is the result of that analysis. It was becoming clear Thursday night, and especially Friday midday, that it would take a modest deterioration to trigger a sell signal.

If the sell is triggered this week, there is always the possibility that it will be a bear trap; I could get whipsawed. But history says, on a probability weighted basis, selling on a trigger is the smart move. Here is a graph I’ve extracted from a prior article “The Easy VIX: The Cost Of Drawdown Insurance.” Gains and losses for each interval are on the horizontal axis, and to appropriately reflect the impact as well as the probability of sell intervals, the height of each bar reflects the probability of occurrence times the magnitude of the gain or loss.

Distribution of Sell Interval Frequency Weighted by The Magnitude of Gains And Losses

Source: Michael Gettings Data Source: & Fidelity

So given the odds, I will abide by whatever signal arises.

Now I’d like to take a moment to describe my efforts at continuous improvement in the algorithm’s design and performance. In the May article when I described the eleven-year modeling results, the algorithm returned 15.5% average annually with a worst 12-month performance of (6.4%) loss. That article, The Easy VIX: VIX Futures Curve As A Risk Mitigation And Market-Timing Tool - An 11-Year Back Test was published on May 29, 2019. Since then I’ve made some changes at the margins which have resulted in some net improvement. Currently the algorithm produces a 17.0% average return over the eleven-plus years, a 1.5% improvement, and the trade-off came in the worst 12-month loss being slightly higher at (7.2%).

The small model changes included changing the weightings for the objective function in the artificial intelligence algorithm, adding a small interactive price-trend component to the Confirming Slope, and making the Primary Slope calculation interactive with the prevailing Shape of the futures curve. I mention this, not to make anyone’s hair hurt, but to maintain transparency. For example, the last graph showing the weighted distribution of sell interval results, reprinted from a July 3rd article would look slightly different today. The difference is so small that showing it would just add confusion, but from time to time, if differences become material, I will report the reasons why.

So in closing, the market action on Monday or Tuesday will probably determine the answer to the sell-or-hold question. This article, subject to editorial review, might be published before or after that time. As I’ve started to do recently, if a signal does occur I will enter a timely comment at the bottom of my last article or this one in order to keep you posted. Thanks for reading and especially for following. If you're not yet a follower, consider clicking the orange "Follow" button at the top to receive more articles and market signals.

Disclosure: I am/we are long SPY. 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.