Once a month or so we externally call-out the key shifts in a number of the behavioral indicators we track internally for market color. In the last look on 10/25 (A Refresh of Expectations), consensus shifted in a big way from August and post-labor day bullishness on U.S. equity markets broadly to much more cautious into the election (probably in part to hedging event risk). At that point (10/25) the Treasury Bond Volatility index (MOVE) was pinned at a low not visited since 2014 (against high-yield and junk at a 2016 high), and the market’s price for insurance against an increase in equity volatility was at a large premium (relative skew and implied volatility against realized volatility over multiple durations).
The market’s reaction and breaching of quantitative signaling lines (partly to a coming political regime shift) shifted us away from some of the TREND biases we held for much of 2016. In a research note earlier this week we outlined our process for getting out of financials on the short-side and long-bonds, gold, and utilities on the long side (Process Trumps Politics: Using Data To Improve Our Positioning. Much of the shift was also driven by our GIP model, and the fact that #Quad2 positioning biases look a bit stickier than we originally anticipated with the advent of Q4 data which is continuing to take shape – Retail sales was a large contributor which we also outline in the note.
Below we outline some important takeaways with respect to changes in market expectations. The point of the note is to call-out the more extended or extreme moves – much of the time these conclusions can’t be made, but the takeaway here is that more extreme bearish biases in U.S equity markets have mean reverted and been flushed out with the largest apparent uncertainty from investors currently on the direction of currencies from here. Ping us back with thoughts or questions.
- A rising curve, a stronger USD, and compressing spreads raise the question as to whether or not the growth trade has legs. Looking at cyclical industries, high yield spreads in Materials and Energy are tighter than the overall index over the period from July 2014-Present. When considering the prices of most commodities currently relative to July 2014 before the USD broke out, spreads say something about the market’s perceived risk resource-sector credit, and possiblt growth expectations in general.
- The U.S. Dollar index is now at a level not breached since 2002, and net futures and options positioning shows investors have piled in on the long side of the USD against short positions in other currencies – positioning in the British Pound and Swiss Franc is particularly extended on the short-side.
- S&P net long positioning was cut from YTD August highs when both realized and implied volatility were near historic and dangerous lows, and U.S. short-interest, after being cut by 12% from Fed to Aug. ticked higher in most sectors into the election (latest data is from 10/31 so the next refresh will be an interesting read).
- Another notable call-out is in the extreme bullish positioning in base metals next to the USD. Copper net futures and options positioning is +3.4x and +3.9x on a TTM and 3Yr Z-score basis and sitting at its most bullish absolute contract positioning ever. This coupled with the price and volume that was behind the strength in metals and mining stocks coming out of the election again forces us to take a hard look at the direction of the growth trade in cyclicals of which we had a bearish bias on for two years.
- A longer-term look at the sensitivity of metals producers (steel, aluminum, etc.) and pure-play miners to prolonged moves in the U.S dollar index suggests the short-term positive correlations will not make a long-term trend, but again recent strength coupled with the easier comps from a GDP perspective and election-related demand and trade narratives, we're forced to at least ask these questions. Below we show quarterly relative correlations, Daily absolute correlations, and commodity-related performance during peak-to-trough moves of +/- 30% in the U.S. dollar. The relationship is hard to refute.
- FX Volatility: With FX Net Futures & Options Positioning among the most extended in Macro (Long USD and Short other Currencies), both realized and implied vol. is heightened in currency ETFs relative to other asset classes.
- Looking at an aggregated measure of historical vol. (30D/60D/3Mth/6Mth), the British Pound and the Yen rank in the top 3 (highest percentile reading) in a large universe of tickers across asset classes. The pound is in the 81st percentile and the Yen is in the 67th percentile. Even on a shorter duration (30D which is well past Brexit event observations), the Pound ranks near the top.
- Trailing multi-duration vol. in the Euro hasn’t been as stretched as the Pound and Yen, but insurance on the Euro relative to realized ranges is the most extended of any ticker in our universe. Again the implied premium is more or less a comparison of forward looking volatility expectations against what has happened in the past. In the Euro the market expectation is for a much more volatile currency into 2017. The expectation among European Equities is the same. 5 of the top 10 tickers with the largest premiums are European equity related.
- U.S. Equity Market Volatility: S&P skew has shifted much flatter coming out of the election which is largely what one would expect given all of the (now shifting) narratives around election implications. While sentiment has shifted from what we saw as increasingly bearish in our last update, the conclusion is more “mean-reversion” centric rather than a swing the opposite direction to full on bull.
- We show two versions of skew in the first two charts below to answer the question as to which direction for the market demands higher insurance costs. Both charts paint the picture that the difference in implied probabilities between calls and puts has narrowed dramatically since our most recent external update. More specifically, insurance against downside is much cheaper on a go forward basis. Pick your duration – the chart looks the same, although the downside in front month and second month contracts has come in more than 3 or 6 month insurance.
- Downside protection in many U.S. equity related tickers is much less expensive than it was on 10/25 when S&P and U.S. sector-specific skew was near its most negative point of the year (Utilities, Transports, Staples, Health Care). Also implied volatility premiums put many U.S. sectors in the top 10 with the biggest divergence between trailing and forward looking volatility expectations – Nasdaq, Dow, Consumer, Transports, Health Care were all extended and present in the top 10. Currently this short-bias has been flushed across sectors and U.S. equity beta, with the Materials sector experiencing perhaps the largest shift (no surprise with election-related growth narratives which we don’t refute as of right now).
- Most of the behavioral indicators we follow exhibit momentum herding tendencies which is why a trend following, mean reversion process relies on these sorts of indicators for market color (i.e. the longer a market moves one way, the lower the consensus-labeled probability that it reverses).
Overall the conclusion we communicate is that we’re warming up to another quarter of a Quad 2 environment which as mentioned through multiple airwaves this week. This results in our backing off of the long side in gold, utilities, and treasury bonds and the short-side in financials (so more or less getting back to beta). From a sentiment standpoint to view on beta has moved more neutral post-election (probably because of the uncertainty associated with the coming months). While the only extremes to call-out in this updated are FX-related, we question the continued positive relationship between the dollar and the growth related metals/materials trade.