JPM's Quant Wizard Returns To The Dark Side, Warns Of Coming “Market Turmoil”
Having taken a jaunt with all the other recently converted bulls over the past month – ever since Trump’s presidential victory, which nobody anticipated and all the experts predicted would be bearish for stocks – JPM’s quant wizard, Marko Kolanovic has released an “add on” to his wildly bullish “Outlook for Equities” piece from last week, in which he makes a gracious return to the dark(ish) side, warning that “further withdrawal of monetary accommodation would likely lead to turmoil in financial markets” although, he adds, higher oil prices and lack of investor focus on China/CNY mean risks are lower than last January.
He notes that the initial market reaction to the election, with positive equities and the idea of reflation, have been accepted as base-case, however he then warns that this will revert similar to the first quarter of this year, and we will see a pullback in USD, and outperformance of EM assets, commodities, and value equities.
Kolanovic also says the “risk on” nature of the market reaction (bonds down, equities up) prevented a more rapid deleveraging among quant-driven funds going into the U.S. election, however should bond yields continue increasing, with 10Y beyond 2.75%, this will risk an equity selloff that usually triggers a broader deleveraging of var-based strategies. As a reminder, 2.75% on the 10Y is also the threshold which Goldman said last week would result in a stocks selloff.
“Absent another leg of the market rally, this overhang will weaken and at some point reverse, leading to an increase in realized volatility levels” Kolanovic notes.
Next, when looking at his bread and butter, volatility, Kolanovic says the VIX appears to be 3 points too cheap (1 standard deviation) relative to dozens of different macroeconomic variables, however instead of buying the bottom in VIX, Kolanovic is a seller of bounces and repeats his call that VIX in 2017 will likely trade in a similar range to 2016. To wit:
Periods of low volatility are likely to mask underlying fundamental risks and be followed by quick outbursts of volatility that may not last long enough (due to unwinding of hedges, opportunistic selling of volatility) to be captured by an average investor. Hedgers may buy volatility ahead of an event and sell shortly before the catalyst to capture volatility grinding higher. To gauge market risks, equity investors should watch for further increases in bond yields and strengthening of USD.
Still, he does warn that “an upside risk to our base case volatility view is if the US were to enter a recession (to which our Economists assign only a ~25% chance over the next 12M), and a downside risk for volatility would be a quick and effective US fiscal stimulus alongside continued monetary accommodation that causes a rally in risky asset classes.”
When looking at sector and factor dispersion, the JPM quant believes low volatility stocks and segments of tech and discretionary (large cap Internet) may continue to be under pressure due to rotations.
Furthermore, confirming something we have been saying since 2009, Kolanovic warns that the increased participation of machines and algorithms are quicker to adjust to significant events and can eliminate trading activity of slower investors.
Poking inside the vol surface, Kolanovic says that in the post-election rally, current option imbalances are heavily toward calls, which puts downward pressure on market realized volatility, something traders have surely noticed over the past month. Still, he remains troubled by the record divergence between market volatility, as measured by VIX, and political uncertainty, although if the recent political shocks have taught us anything, it is that a surge in political risk is merely an opportunity to BTFD and forget what people were concerned about in the first place.
In an amusing tangent on this topic, Kolanovic attributes swift market rebound following Brexit, U.S. election and Italian referendum to shortened time horizon of macro traders
Finally, the most amusing party of Kolanovic’s piece was his comment on the state of information and content distributors:
Given the failure of many traditional data sources to anticipate geopolitical developments this year, investors should use both traditional and non-traditional data sources (such as big data/internetbased sentiment indicators, independent and smaller media outlets, etc.).
Readers may want to take him up on that, especially if the law that criminalizes “fringe media” as Russian propaganda is enacted into law.
* * *
His full note below:
Outlook for Equity Risk
In our recently published Outlook for Equities, we outlined a cautiously optimistic view for 2017. While there is more upside to equities in 2017, we think that market risks will increase substantially. The main risk for equities will come from a strong USD and higher rates. A strong USD negatively impacts equities. The most direct impact is through international exposure of US equities, and we estimate that a 1% increase in the USD represents a ~0.5% drag on S&P 500 EPS. Indirect impacts are via the pressure on Emerging Markets and Commodity prices (Energy, Materials, Multinationals, EM Equities). Higher bond yields will put pressure on equity multiples (P/E), as a significant re-rating of equity valuation came from near-zero bond yields in recent years. The most directly impacted market segments are yield sensitive sectors (Utilities, Telecoms, Real Estate, Staples), and more generally ‘Low Volatility’ stocks still overweighed by quants and real money asset managers (e.g. as a substitute for bonds). A rapid rise of rates not only destabilizes equity multiple, but also puts the housing market at risk. Over longer periods of time, higher yields create a drag on EPS via interest expenditures, and our estimate is that a 100bp increase in bond yields results in a 1.5% drag on S&P 500 EPS. Finally, higher bond yields and a stronger USD will undercut the ability of the new US administration to revive US manufacturing jobs or aggressively expand the fiscal deficit. Even without further deterioration in yields and USD strength, we may be at increased risk of repeating the scenario from January 2016 where fundamental and systematic investors were selling assets in the aftermath of a Fed hike. We do note that risks are somewhat lower this time, given significantly higher Oil prices and a lack of investor focus on China/CNY.
We also expect an increased level of geopolitical risk and increased policy-related uncertainty in 2017. In Europe, significant risks remain as a result of Brexit, aftermath of the Italian referendum, elections in France and Germany, and continued tensions related to immigrants from the Middle East. The Middle East will likely see further geopolitical turmoil, and oil prices will stay volatile and continue to exert pressure on oil exporters’ budgets. While the US macroeconomic cycle may get a boost from the proposed fiscal stimulus, corporate tax reform and de-regulation, both the passage and efficacy of these measures are not certain at this moment. As various datasets show an increase in political uncertainty, market realized volatility stays contained. The record divergence between market volatility (e.g. as measured by VIX) and political uncertainty (e.g. as measures by EPU index) is shown below. The spread between the two measures indicates that the VIX has been underperforming political risk measures since 2011, likely due to aggressive monetary accommodation. Indeed one can recall the market’s intense focus on Greece in 2011 pre-ECB vs. the currently muted reaction to Brexit, the US Election, political uncertainty in Italy, tensions with Russia, etc.
The current strong USD, rapid increase in yields, and high geopolitical risk provide a fair amount of stress to financial markets. While market realized volatility is contained, this may all change if the monetary accommodation is removed. Bond buying programs that kept yields low across the globe, also put upward pressure on asset prices and downward pressure on volatility. Implied volatility was pressured by yield seeking investors selling options to generate yield. Options that were sold in turn put pressure on market realized volatility via gamma hedging effects. This was e.g. obvious the past summer when the market got pinned and volatility collapsed as a result of record option imbalances (Figure below right shows S&P 500 price action in July and August). Withdrawal of monetary accommodation would also accentuate the problem of a strong USD, uncover geopolitical complacency, and remove the lid from market volatility. In fact, our view is that further withdrawal of monetary accommodation would likely lead to turmoil in financial markets.
Source: JPM QDS Research, Bloomberg; “An Index of Global Economic Policy
Uncertainty” by Steven J. Davis at www.PolicyUncertainty.com;
We warned of the risk of rising USD and rates in December last year, and called for a Value, EM, commodity rally and adjustment of the Fed’s dots lower. Later last spring we argued how fiscal measures are likely needed to re-accelerate global growth. Our thesis was that combination of monetary and fiscal accommodation will result in asset reflation, lightening of debt burden and higher growth across the globe. Since the Trump election (here), positive equity reaction that we expected (here) and the idea of reflation have been broadly accepted as the market base case. However, this is not the reflation we were describing, as the price action resulted in a stronger USD and higher real rates, creating a drag on prices of various real assets (e.g. housing, gold, etc.). We believe this initial market reaction will revert, and similar to the first quarter of this year, we will see a pullback in USD, and outperformance of EM assets, commodities, and Value equities. The underperformance of EM equities relative to the S&P 500 during November amounted to a highly unlikely ~3 standard deviation move. Historically, these moves were virtually always followed by a sharp correction (outperformance of EM over DM of ~5-10%). Furthermore, it is not unreasonable to expect a longer lasting reversion of the move since 2013 (Figure 3).
A tail risk for the market could be realized if the USD continues to strengthen, yields rise and global central banks turn more hawkish (see here). This combination would increase the risk of recession (or stagflation, if accompanied by aggressive fiscal stimulus). In that scenario (which is not our base case, but rather a tail risk that may occur over the next 2-3 years), we may subsequently see monetary policy reverse, leading to a simultaneous application of aggressive monetary and fiscal easing. In these circumstances, it would not be unreasonable for safe haven, inflation sensitive assets such as Gold to reach all-time highs. Combining our risk embracing view of reflation, and to hedge the tail risk of stagflation, we continue to advocate the risk barbell portfolio.
We followed flows from systematic strategies for the past year. Notable episodes that we warned of include de-risking in December/January, sector and style rotation from equity quant strategies in February, and March, and the equity sell-off in September. While there are indications that systematic strategies and their impact are changing, they will continue to be an important factor for fundamental equity as well as macro investors. Going into US elections, macro systematic investors (such as trend followers and various Var-based strategies) were long bonds. While the performance of these strategies suffered, the “risk on” nature of the market reaction (bonds down, equities up) prevented a more rapid deleveraging. Yet this risk is not entirely eliminated, and should bond yields continue increasing (e.g. 10Y beyond 2.75%) this will risk an equity sell-off that usually triggers a broader deleveraging of var-based strategies.
On a number of occasions this year, we have witnessed quant funds rotating away from stocks that were characterized by low volatility, growth and momentum, into value (this description is not quite exact as stocks comprising these indices are changing in the process). We believe this trend will continue and will be supported by our reflation thesis. This rotation has benefited Energy, Materials, Industrials, Multinationals (earlier this year), and more recently Financials and parts of Health Care. Quant rotations are usually reinforced by de-risking from fundamental investors with stringent risk limits such as sector hedge fund ‘pods’. Low volatility stocks and segments of Tech and Discretionary (e.g. large cap Internet) may continue to be under pressure due to these rotations. These pressures tend to escalate at the turn of the month (e.g. in February, and March) and have contributed to the recent selloff in Technology (November month end). As the momentum and volatility scores of these stocks de-rate, it will take more than one monthly rebalance cycle for this selling pressure to abate. Recent political developments in the US may also contribute to these rotations. Over the past years, on a relative basis the regulatory environment was more favorable to industries of ‘bits’ (e.g. internet, tech, etc.) than to industries of ‘atoms’ (e.g. financials, energy, materials, etc.). There are indications that these trends may revert, which could prolong the rotation beyond those driven by short-term rebalancing of quantitative and fundamental investors.
Systematic trading related to the index option complex continues to impact the market price action and volatility. While last year we highlighted how record levels of short gamma exposure contributed to the August market crash (see here and here), this year it was record levels of long gamma exposure that resulted in market pinning and the collapse of volatility in July and August (Figure above right). Given the post-election market rally, current option imbalances are heavily towards the calls, which puts downward pressure on market realized volatility. Absent another leg of the market rally, this overhang will weaken and at some point reverse, leading to an increase in realized volatility levels.
In our 2016 Outlook for equity markets we forecasted an increase in both volatility and tail risk (our forecast was for the VIX to average 16-18 this year). The trailing 12M average level of the VIX peaked at 18 after Brexit and then declined to ~16. Tail risk was also pronounced, with volatility of volatility spiking on Brexit and shortly before the US election (volatility of VIX at ~125). However, what was surprising is that in the aftermath of 2 seismic events – Brexit and US Elections – market volatility quickly declined, pulling the average levels of volatility lower throughout 2H16. It took several weeks for volatility to subside in the aftermath of the August 2015 sell-off, but it took only few days after Brexit, and a few hours after the US election and Italian Referendum. (Figure 4 below shows times needed for S&P 500 to recover in the aftermath of significant catalysts).
It appears that the time horizon of macro traders has shortened, likely as a result of increased participation of machines and algorithms that are quicker to adjust to significant events and can eliminate trading activity of slower investors. Consider for example the US elections – traders in Japan registered a 5.4% Nikkei drop on the 9th, followed by a 6.7% rally on the 10th, while S&P 500 investors did not register a significant close-to-close move over the election (due to market hours difference). These two days were enough to shift the volatility regime (usually calculated from closing returns) for the whole of 2016 for the Japanese equity market, and leave it unchanged for S&P 500 (e.g. think of rebalancing needs of a hypothetical risk parity fund, or a short volatility strategy based on Nikkei vs. one based on S&P 500). We also noticed that for a number of significant catalysts this year (Brexit, US Election, Italy Referendum) broad expectations were wrong both on the outcome and the directionally forecasted impact. It is possible that the lack of market reaction (or a reaction that went against the accepted narrative) was in part driven by investors’ reluctance to transact (“two negatives equal a positive”). Elements of randomness described above warn us not to be complacent about market risks in 2017.
According to our macroeconomic model, the current VIX level appears to be ~3 points too cheap (1 standard deviation) relative to dozens of different macroeconomic variables. However, despite our view that market risks will be higher in 2017, we are not convinced that the average VIX level will capture these increased risks. The VIX in 2017 will likely trade in a similar range to 2016, in our view. Periods of low volatility will mask underlying fundamental risks and will be followed by quick outbursts of volatility that may not last long enough (due to unwinding of hedges, opportunistic selling of volatility) to be captured by an average investor. Hedgers may buy volatility ahead of an event and sell shortly before the catalyst to capture volatility grinding higher. To gauge market risks, equity investors should watch for further increases in bond yields and strengthening of USD. Given the failure of many traditional data sources to anticipate geopolitical developments this year, investors should use both traditional and non-traditional data sources (such as big data/internet based sentiment indicators, independent and smaller media outlets, etc.). In addition to the aforementioned risks, an upside risk to our base case volatility view is if the US were to enter a recession (to which our Economists assign only a ~25% chance over the next 12M), and a downside risk for volatility would be a quick and effective US fiscal stimulus alongside continued monetary accommodation that causes a rally in risky asset classes.