Setting the stage for current conditions
In response to the financial crisis in 2008 the Federal Reserve loaded up its balance sheet as part of a stimulative liquidity program to increase bank reserves and stabilize the economy. Whether those accommodative monetary policies worked in the long run is still being debated and analyzed by economists.
The purpose of this article is not to evaluate the effectiveness of various monetary programs, but to succinctly identify meaningful trading patterns that emerged during this activity of unprecedented Fed monetary intervention. More importantly, I will focus on how certain patterns have changed significantly under the most recent monetary program and what opportunities may be available to traders into 2019 and 2020.
As a result of the QE programs, the Fed’s total assets rose from $882 billion in December 2007 to $4.473 trillion in May 2017 as illustrated in the Fed graphic below:
As the recovery from the 2008 financial crisis progressed, more assets were added to the Federal Reserve holdings until 2014. For the first time, substantial quantities of mortgage backed securities were also moved to the Federal Reserve balance sheet. As the chart below illustrates, the largest segment of the Fed assets are treasuries and of those more than half have maturities less than 5 years.
Correlated impact of Quantitative Easing on the S&P 500
There's little question that the program, known as quantitative easing or "money printing," boosted the stock market. The three iterations of QE between November 2008 and October 2014 each saw big boosts to the market, with a cumulative S&P 500 gain from beginning to end, including the various down periods between each leg, of about 140 percent. Source
The Fed chart above illustrates the substantial growth in the S&P 500 as each of three subsequent QE programs was introduced to the US economy in stages shown in blue. Following the second QE program the S&P 500 entered a period of remarkable stability starting near the end of 2011. At that time the Fed balance sheet holdings of treasuries were raised to approximately $1.8 trillion, up from less than $800 billion following the first QE program in 2010.
An important theory I will form in more detail is that a key liquidity threshold was crossed when Fed assets rose to nearly $3 trillion in total at the end of 2011, including the approximately $1.8 trillion in treasuries. To explain by way of illustration, I created a chart of the number of days per year with larger than 2% swings in the S&P 500 from 2008 through November 2018.
As the chart of 2%+ daily moves per year shows, there was a significant decline in S&P 500 volatility at the conclusion of QE 2 in 2011. In fact, four of the six 2%+ events for 2012 were in the first half of the year. In my opinion, the unprecedented added liquidity delivered a stabilizing impact following the financial crisis of 2008-2009. This impact became most evident from 2012 through 2017 by providing a great deal of low interest investment capital and buyback opportunities that significantly attenuated price swings in stocks and their related indexes.
From 2008 to 2011 the chart shows 184 of these 2%+ S&P 500 moves in a four year period. Following $3 trillion in additional QE through the 2011, the next four years to 2015 resulted in only 27 S&P 500 moves over 2%. If this QE-2 theory contributed to an 85% reduction from 184 events between 2008-2011 down to 27 such moves over the next 4 years, what did QE-3 contribute as $4.5 trillion was reached in 2015?
To answer that question, an excellent chart of low volatility reveals how well the S&P 500 performed following the completion of QE 3 post 2015. Not only were there zero 2%+ volatility events for the S&P 500 in 2017, but the S&P 500 stayed within 5% of a 52-week high consecutively for more than 370 days.
This period from 2016 to 2017 is a stretch of low volatility that we have not seen in about 25 years. It is one of only four such periods since 1963. As I detailed previously, this period of extreme low volatility contributed to non-random trading patterns on the CBOE VIX volatility index shown below for 2017:
The mid-month VIX spikes may have corresponded with Federal Reserve market operations, but they followed a recurring non-random pattern that supports my ongoing theory related to the most current activity through 2018. The other key factor to note here is how tightly range-bound the VIX remained throughout all of 2017 as we encountered such record low volatility.
Then on Sep 20, 2017 the Fed announced the incremental start of the balance sheet normalization program to reduce some of $4.5T in holdings mainly acquired during quantitative easing programs from 2008 onward. This blandly labeled normalization policy was detailed in a June 2017 addendum identifying the steps of the program. The first scheduled roll-off of treasuries was capped at $6 billion in October 2017.
Not only does the normalization policy addendum cover the steps and caps of the Federal Reserve asset unwind, but it calls for raises in the target range for the federal funds rate. Rate hikes that we have already seen steadily progress the last couple years to an expected ninth hike coming in December.
All of these normalization policy steps are inverse actions to the accommodative monetary policy of the prior QE programs. Simply put, raising rates and reducing the Fed balance sheet reduces liquidity and the flow of capital to the same markets that have set records in low volatility conditions over the last seven years.
What comes next with Quantitative Tightening?
As of October , the Fed will let billions of dollars of securities mature each month without reinvesting them. It will gradually increase the amount of maturing bonds each quarter over the next year. HSBC Chief Economist Kevin Logan said this process is new territory for the Fed. "This is a big experiment," Logan said. "It's something that's never been done before." Source
The big picture of this Federal Reserve Roll-Off schedule is shown below on the chart of treasury maturity spanning from September 2018 to February 2048. As shown in the bar chart at the top of this article, the majority of treasuries on the Fed balance sheet have maturities of less than 5 years. Approximately $420 billion in holdings reduction is planned for 2018 and up to $600 billion is planned through 2019.
The schedule is intended for public dissemination and analysis as part of an effort to mitigate concerns and allow the market to process the normalization policy as efficiently as possible. Part of that process has led me to try to anticipate what non-random trading patterns are likely to emerge during this unprecedented unwind of Federal Reserve assets.
The purple line on the chart depicts the Quantitative Tightening monthly cap of securities that will be rolled off the balance sheet. You may notice that the cap line increased at the left most edge of the chart to a current maximum level of $30 billion per month showing in red the amount of securities that will be reduced. This cap line began at $6 billion per month in 2017 and has been steadily increasing until it reached the maximum pre-established level this past October.
According to the most recent Quarterly Report on Federal Reserve Balance Sheet Developments November 2018, the total balance of the System Open Market Account (SOMA) is at $2.29 trillion as of October 24, 2018. It is not precisely known how low the Federal Reserve will reduce total assets under the current program over the next few years. But according to some projections into 2020, we are nearly $420 billion into this estimated schedule of quantitative tightening around the $2.2 trillion level.
If $620 billion of additional reduction is planned for 2019 the effects we have seen in 2018 may increase in meaningful ways for next year. What do I mean by this?
First, in the S&P 500 chart above of the 2%+ moves per year, we have already encountered 14 large moves of the S&P 500 in 2018, which is more than 50% higher than the number of events in 2016 and 2017 combined, and there is another month left in the year.
Second, if there is a threshold level for the QT where the reduction in liquidity falls below a level that triggers an outsized market decline response, it is more likely to occur next year with full effects of the Fed fund rate hikes and another $620 billion in asset reductions.
Third, let us more closely examine the volatility behaviors of the S&P 500 through 2018 to try to identify some strong correlations and unusual trading opportunities.
Volatility pattern analysis for trading opportunities
After record low market volatility from 2012 through 2017, conditions changed dramatically in 2018. As the QT program moved into 2018 jumping up to scheduled levels above $10 billion per month, the market encountered a VIX shock that sent the S&P 500 down over 4% in a single day. The small reduction in the Federal Reserve chart above may have correlated strongly with the enormous VIX spike shown in February below.
The unusual mid-month patterns continued on the VIX from 2016 and 2017. Highlighted by rectangles above are the recurring declines of the VIX in the first half of each month. The two critical exceptions to that non-random anomaly are shown with arrows at the February and October VIX spike events. Both events corresponded with very large market declines and significant increases in the Quantitative Tightening schedule.
One way to categorize the past year was to apply the combined roll-offs of the treasury and mortgage backed securities to a chart of the S&P 500 as shown below. From January to October this year the tightening capped at $30 billion per month and then jumped to $50 billion combined in October.
By correlating the VIX summary with Fed QT activity a noticeable change in the VIX was observed increasing to more elevated levels after October. Additionally two substantial spikes occurred in both October and November that reflected a change from the past several years of VIX behavior of a single mid-month event.
A closer look at the volatility index for November shows the two separate VIX events moving as much as 21% in a two-day period from November 19-20. Now I can't say with certainty that these volatility events are directly caused by Federal Reserve activity, but the correlations and expected increases follow closely with what me might expect from a substantial month-end and mid-month liquidity reduction event.
Another way I found to test the likelihood that these VIX spikes and large S&P 500 moves could be directly related to the Federal Reserve tightening was to overlay three years of data on the FRED database related to these variables.
The seismographic looking chart above compares the daily percentage changes of the S&P 500 (red) overlaid with the percentage change of treasury securities held by the Federal Reserve (blue). In the 3-year chart above you can clearly see that when there were no changes in the Fed holdings of treasuries, there was also very little change in the S&P 500. However, when the changes in the Fed balance sheet began to increase in January of 2018 the percentage changes of the S&P 500 correspondingly increased in magnitude.
Naturally, I wanted to examine this correlation across 2018 in more detail so I produced the following chart to analyze these effects:
The red and green rectangles correspond to time periods of declines and spikes in the S&P 500 that positively correlated with a similar change in direction of asset changes. More often than not as you can see, an increase in the S&P 500 occurred while Fed activity was at or near zero and a decrease in the S&P 500 occurred when changes (planned reductions) in treasury holdings were taking place.
Trading the patterns from the volatility event analysis
So I have provided some background on this unprecedented and yet planned quantitative tightening schedule over the next several years. I have shown some evidence that volatility and Federal Reserve asset reductions may have a direct causal relationship. The next step is to show a couple potential applications from the data that may contribute to profitable trading in the future.
1. Forecasting VIX event patterns. Now that the QT has capped at the maximum $50 billion per month and has begun showing bi-modal patterns in both October and November, there could be a strong basis for another bi-modal pattern in December.
Watch for the two breakout events per month to recur in December and into next year as it has happened to occur in both October and November. Watch for changes in elevation of the trend as well as the number of large moves per month and how well they correlate with the published roll-off schedule.
2. Trade the signals. Either using options on the VIX or the VelocityShares Daily 2x VIX ST ETN (TVIX), trade the signals for maximum profitability. The example below comes from the October 4th VIX breakout event shown as number 1 on the larger chart above.
The indicators on the intraday chart below signal the positive move prior to breakout conditions at the start of the month for TVIX gaining more than 16% in one day.
The larger scale daily chart shows the second day move on the 5th of October at around 31/share long before TVIX moved up over $50/share in the first half of the month and over $60/share by the end of October in two separate events.
These signals and applications could benefit you greatly through the end of the year and well into 2019. I continue to monitor and apply these trades on a regular basis as I have over the years. While we are already in some uncharted waters, I fully expect that the Fed's balance sheet normalization policy is likely to give us even more excitement and volatility than we have seen in many years. All the best,
JD Henning, PhD, MBA, CFE, CAMS
Disclosure: I/we have no positions in any stocks mentioned, but may initiate a long position in TVIX 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.