"I don't like to think that maybe I'm just getting old. I'm not too excited about watching a huge explosion. I'm more interested in people and characters." - Norman Jewison, Canadian director
A "Sympathetic detonation" (SD, or SYDET), also called flash over, is a detonation, usually unintended, of an explosive charge by a nearby explosion (in our case Emerging Markets). Sympathetic detonation is caused by a shock wave (think about the Taper Tantrum and the surge in Mack the Knife aka US dollar + US positive real interest rates), or impact of primary or secondary blast fragments. The initiating explosive is called donor explosive (commodities rout leading to Emerging Markets woes), the initiated one is known as receptor explosive. In case of a chain detonation, a receptor explosive can become a donor one, which is exactly what we are seeing in Emerging Markets and High Yield credit, particularly in the Energy sector which will no doubt lead to defaults and restructuring, rest assured.
We find our title analogy appropriate given that in a "Sympathetic detonation", detonators with primary explosives are used (surge in the US dollar and collapse in oil prices), the shock wave of the initiating blast (Taper Tantrum) may set off the detonator and the attached charge (Emerging Markets). However even relatively insensitive explosives can be set off if their shock sensitivity is sufficient. Depending on the location, the shock wave can be transported by air, ground, or water. The process is probabilistic, a radius with 50% probability of sympathetic detonation often being used for quantifying the distances involved. Often "Sympathetic detonations" occur in munitions stored in e.g. vehicles (mutual funds), ships (hedge funds), gun mounts, or storage depots (dwindling EM FX reserves), by a sufficiently close explosion of a projectile or a bomb. Such detonations after receiving a hit can cause many catastrophic losses (such as the ones received recently by investors in various asset classes except cash, still being "king" in a deflationary bust scenario). What we are seeing playing out, is of course what we warned about since 2011, namely a currency crisis, which is the wave number 3 we have discussed in numerous conversations.
Just remember this, before we dive into our conversation, from our credit perspective and in relation to our chosen analogy even cheap money has to be paid back sometimes. And that is exactly what is becoming increasingly unlikely.
Major equity / Credit divergences should always be taken very seriously to quote the Guest note from November 2014 from our good friends at Rcube Global Asset Management entitled "US Equity / Credit Divergence: A Warning".
Whereas our friends were probably a little bit too early in their call, we think that the recent significant deterioration witnessed in Credit and in particular US High Yield (a subject we have discussed at length during this summer), is for us, and some others, showing that the credit cycle is indeed turning and taking, we think, a turn for the worse.
In this conversation, we will again point out to a worrying instability sign we have indicated, namely the impact of positive correlations leading to "Bayesian" price movements. On that subject we would like to refer to our August post "Positive correlations and large Standard Deviations move".
While typing this very note, we mused on twitter with our estimate for Nonfarm payrolls at 155K 30mns before the number came out. Given our recent blunder in hoping the Fed would hike by 25 bps in September thanks to us succumbing to "Overconfidence effect", we decided to opt for the "contrarian effect" which this time around came to fruition as NFP came out with a miserable 142K, putting us firmly back in the "deflation" saddle thanks to our US long duration exposure (partly via our ZROZ ETF exposure) as well as our long gold miners exposure which suddenly sprung back to life.
- Increasing correlation between asset classes is leading to instability and weighting on diversification - are inflation linked-bonds a solution?
- In US High Yield this time is not different.
- Final chart - Correlation to Oil very positive and unusually high
- Increasing correlation between asset classes is leading to instability and weighting on diversification - are inflation linked-bonds a solution?
"The greater the volatility, the greater the disadvantage of owing negative convexity bonds like you find in the High Yield space. In the current low yield environment, both duration and convexity are higher, therefore the price movement lower can be larger..."
"The past two quarters have been characterised by an increasing correlation between asset classes and between equity and bonds. Diversifying a portfolio is becoming more difficult. On the other hand, in a context where inflation should gradually normalise in the next quarters, inflation-linked bonds bring some diversification benefit.
In the broader picture, on a historical basis, inflation-linked bonds have become quite correlated with other asset classes lately, in line with increasing average cross-asset
The US TIPS market is the one for which, on a historical basis, the correlation with other asset classes is least extreme, bringing more diversification benefit (see chart in margin).
"The correlation between inflation-linked bonds and nominal bonds has increased significantly in recent years, particularly as themes such as secular stagnation, deflation and even Japanification of the economy are still part of the economic newsflow. The current low growth and low inflation environment has made inflation-linked bonds and nominal bonds move in tandem since 2013. As discussed earlier, we have decided to introduce inflation-linked bonds into our portfolio by giving a maximum weight to US TIPS. US TIPS look attractive not only on a valuation basis, but they also provide lower correlation to other developed markets nominal bonds compared to UK peers." - source Société GénéraleWhereas in their long and interesting report Société Générale make the case for using inflation-linked bond as another layer of diversification and in particular US TIPS, we have to agree that given secular stagnation, and "Japanification" of the economy (which has long been our scenario, Europe wise), indeed US TIPS are more "compelling" than UK linkers and still are less positively correlated to nominal bonds for a very simple reason: their embedded "deflation floor".
Unlike TIPS, gilt linkers do not have a "deflation floor" protecting the cash flows from deflation, hence the current favorable attractiveness of US TIPS. But, in similar fashion, French government issues inflation-linked Obligations Assimilables au Trésor (OAT) bonds, which reference either a pan-European inflation index (the Euro-Zone Harmonized Index of Consumer Prices excluding tobacco (HICP)) or a French inflation index (French CPI ex-tobacco) also feature a "deflation floor" in that the principal face value and coupon payment can never decline below the value at issuance. Swedish linkers as well provide a similar deflationary protection feature as per the table below from Pacific Alternative Asset Management presentation from the 3rd Quarter 2011:
"US Inflation-Linked BondsIndeed, in a prolonged period of deflation, such as the one we are currently experiencing, the embedded "deflation floor" both applied to the stated coupon as well as to the face value of the bonds. These are indeed very interesting features that should not be neglected when selecting an exposure to the index-linked sovereign bond markets.
The largest market for inflation-linkers is in the United States, where the US government issues Treasury Inflation Protected Securities (TIPS). TIPS feature protection of principal and interest payments from inflation. The principal, which is payable at maturity, is linked to changes in the non-seasonally adjusted Consumer Price Index for Urban Consumers (CPI-U). The semi-annual coupons on the bond are also inflation-adjusted since they are based on the underlying inflation-adjusted principal amount. Notably, the bonds also feature a ‘deflation floor,’ as the amount owed on the notes can never fall below the original face value of the bonds. This floor also applies to the coupon, which can never decrease below the stated coupon rate at issuance. The floor is an important component that can protect investors against prolonged periods of deflation." - source PAAMCO
When it comes to gauging the risk in rising deflationary trends, the US Deflation hedge premium has been rising significantly recently as displayed in the below Bloomberg chart via @boes_ on twitter:
Furthermore, the global deflationary forces have been gathering steam as well as displayed in the below Gavekal Graph:
- source Matthew B - @boes_ on twitter
World CPI is falling, not a good indicator and no matter what central bankers are telling us, they are inept at re-igniting the "inflation" genie so far.
Furthermore, the US 10 year government bond market is pricing more and more a deflationary/quasi-recessionary picture:
- graph source Bloomberg
Post NFP data, we have broken again the lows on the US 10 year.
How can there be a recovery and a rise in "inflation expectations" when Average Hourly Earnings MoM are flat and Average Hourly Earnings YoY are growing at 2.2% when the market was expecting 2.4%?
This is proof that the US much vaunted recovery is weaker than expected.
Also the Atlanta Fed has revised its GDP outlook, On October 1, the GDPNow model forecast for real GDP growth in Q3 2015 is 0.9%. Many sell-side economists who also got their NFP call wrong (99 of economists polled by Bloomberg in fact) still suffer from "Overconfidence effect" when it comes to their US GDP growth expectations:
- source Atlanta FedCould the Fed really hike when US breakevens are falling rapidly, which is clearly indicative of deflationary forces at play? This also another reason why the embedded "deflation floor" in US TIPS is so enticing - graph source Bloomberg - Robin Wigglesworth on Twitter
-graph source Robin Wigglesworth on Twitter
In September as well, Europe has clearly shown it was moving back into the dreaded "D" territory with the latest Eurostat Eurozone MUICP coming at -0.1% - graph source Bloomberg - Holger Zschaeptiz on twitter:
- source Bloomberg - Holger ZschaeptizOr, maybe you call this a recovery?
Bloomberg graph displaying the Unemployment-to-Population Ratio vs Unemployment rate - graph source Bloomberg - Michael McDonough
- graph source Bloomberg - Michael McDonough
Back in September 2012, in our conversation "Zemblanity",(Zemblanity being defined as the inexorable discovery of what we don't want to know), we discussed the relationship between credit growth and domestic demand and why ultimately our central bankers will fail in their useless reflationary attempts:
"credit growth is a stock variable and domestic demand is a flow variable"We even asked ourselves at the time the following question:
"Does the end (lowering unemployment levels) justify the means (increasing M) or do the means justify the end (deflationary bust)?"What we have long posited is that while wanting to induce inflation, QE induces deflation.
This is what we discussed in March this year in our conversation "The China Syndrome". At the time, we quoted CITI's Matt King's 27th of February note entitled "Is QE Deflationary":
"It’s that linkage between investment (or the lack of it) and all the stimulus which we find so disturbing. If the first $5tn of global QE, which saw corporate bond yields in both $ and € fall to all-time lows, didn’t prompt a wave of investment, what do we think a sixth trillion is going to do?No wonder Société Générale thinks US TIPS are enticing from a"diversification" perspective. They are not enticing because they are "cheap", they are enticing because of the "deflation floor" embedded!
Another client put it more strongly still. “By lowering the cost of borrowing, QE has lowered the risk of default. This has led to overcapacity (see highly leveraged shale companies). Overcapacity leads to deflation. With QE, are central banks manufacturing what they are trying to defeat?”
Clearly this is not what’s supposed to happen. QE, and stimulus generally, is supposed to create new demand, improving capacity utilization, not reducing it. But as we pointed out in our liquidity wars conference call this week, it feels ever moreas though central bank easing is just shifting demand from one place to another, not augmenting it.
The same goes for the drop in oil prices. In principle, this ought to be hugely stimulative, at least for net oil consumers. And the argument that it stems solely from the surge in US supply, not from any dearth of global demand, seems persuasive as far as it goes.
But in practice, the wave of capex cuts and associated job losses in anything even vaguely energy-related feels much more immediate than the promise of future job gains following higher consumption. The drop in oil prices, while abrupt, in fact follows a three-year decline in commodity prices more broadly. It’s not just oil where we seem to have built up excess capacity: it’s the entire commodities complex." - source CITI
The lack of evidence that consumers are spending what they are saving on gasoline should also start to concern investors.
Stronger USD leads to higher deflationary risk leading to lower long-term bond yields (that's what we are playing right now). Even though exports are only 13% of the US economy, 40% of S&P 500’s earnings now come from outside the US.
- In US High Yield this time is not different.
"If we take CCC Default Rate Cyclicality as an early indicative of a shorter credit cycle, then it is the rating bucket to watch going forward
Why the CCC bucket? Because there has been this time around a very high percentage of CCC rated issuers accessing the primary market in High Yield.
A rise in defaults would likely be the consequences of a deterioration in credit availability. Credit ratings are in fact a lagging indicator." - source Macronomics
At the time of our July conversation we did read with interest UBS's take in their Global Credit Comment from the 27th of July entitled "The scary reality":
"The problem, however, is some of the tourists underappreciate the exponential loss and mark-to-market functions for low quality high yield assets. As we have noted previously when the credit cycle turns annual triple C default rates can surge from 5% to 30% while average triple C prices can fall into the $40 - $50 range (versus $83 currently) - especially given expected recoveries average in the $20 - $25 context. The scary reality is those investors in triple Cs are seeking high single digit returns when they are likely to end up with negative total returns over the next several years (if our view of the credit cycle proves correct)." - source UBS
"Are we there yet?
Let’s take stack of how far we’ve unwound in September. US HY has backed up a 100bps to 670bps, yields have broken into the 8 handle, and retail has pulled out at least another $1bn from mutual funds. At current levels, HY spreads are pricing in a 6.5% default rate for the overall market, and 4% on an ex-energy basis over the next 12 months. So, is it time to buy yet? To tackle that question, we don’t have to look too far. In 2011 when there was fear of a global contagion, yields reached 9.8%, and ex-commodity spreads 900bps. Yes, the source of contagion was different (Greece and US downgrade then vs commodities and EM now) but the transmission mechanism is still the same- investor risk aversion. And where we are today is arguably worse than where we were then from a credit standpoint because HY earnings then were better, leverage was lower and USD was weaker. Low quality issuers were still able to access the market and we had the Fed easing. Today, not only is the Fed in normalization mode, but the proportion of CCC issuers that have managed to access the market on an LTM basis stands at <20 font="" nbsp="">20>a far cry from the peak of 52% two years ago. Looking back, the periods when access to liquidity for lower quality issuers was this poor and going in the wrong direction, was in 2008, 1999 and 1989, just ahead of each default cycle (Chart above)." - source Bank of America Merrill Lynch.
"Altice on Friday sold $4.8 billion of junk bonds to fund its $10 billion purchase of Cablevision Systems Corp., according to S&P Capital IQ LCD. When the deal was shopped earlier this month, Altice expected to sell $6.3 billion of debt, investors said.
A 10-year bond was priced to yield 10.875%, compared with yields as low as 9.75% that were suggested by bankers initially, according to S&P Capital IQ.
Olin on Friday sold $1.2 billion of bonds to pay for its pending acquisition of Dow Chemical Co.’s chlorine-products unit. Earlier in the month, Olin was expected to sell $1.5 billion of bonds, fund managers and analysts said.
The annual interest rate on Olin’s 10-year bonds sold Friday was 10%, up from 7% expected earlier in the month, according to S&P Capital IQ. The steep increase in yield reflects growing concerns that slowing demand from China could hit sales of chemical makers." - source Wall Street JournalALTICE SA's total debt amounts €45 billion (including Cablevision). It represents nearly twice the turnover and 5.7 times gross results for the group whereas the average for the sector is around 2 to 3 times.
Back in June 2015 in our conversation "Eternal Return", we indicated that for us M&A activity was the last sign we were getting late in the credit cycle à la 2007. This is as well confirmed by the same Wall Street Journal article quoted above:
"Companies have announced $3.2 trillion of M&A this year, according to Dealogic, emboldened to merge by cheap debt and the long stock rally that began after the financial crisis. That puts 2015 on pace to rival 2007 as the biggest year ever for takeovers. Issuance of junk bonds backing M&A deals hit a year-to-date record of $77 billion through Friday, according to data from Dealogic."
A souring of investors on junk bonds could limit the availability of financing for deals that require a lot of borrowing. Banks have been under pressure from federal regulators to reduce their loans to such companies, and a pinch in the bond market could leave those deals struggling for financing." - source Wall Street JournalYes, dear readers, time to adjust to reality, this time it is not different and cost of capital is going up. Trade accordingly.
"US HY: This selloff is differentThe speed and magnitude of the recent US HY selloff has caught credit investors offguard and captured the attention of the broader market. Spread widening of 46bps over the last 2 days1 was the worst seen since October 2011, leaving the overall market at 689 bps, 216bps wider than June 1st. Naturally, clients are asking what has changed and what is causing the selling pressure. Many would assume that credit markets are being hit with substantial retail outflows, leading to forced mutual fund selling. However, in context, the recent outflows are not commensurate with the spread widening experienced. From June 1st to yesterday, HY spreads widened 216 bps on $11.4bn of retail mutual fund outflows. Interestingly, from Sept 1st, HY spreads widened 97bps on effectively zero mutual fund outflows. This lies in stark contrast to prior recent selloffs during Taper Tantrum (+57bps on $12.8bn of MF outflows, Jun- 2013) and last summer’s swoon (+87bps on $21.6bn of MF outflows, Jul-Sept 2014). This begs the obvious question: What accounts for the greater severity of today’s price action vs. the prior two years?
At the core, we believe fund managers are aggressively raising cash in anticipation of a structurally more volatile and dangerous environment. Investors are no longer only experiencing the pain of mark-to-market pricing on their portfolios, but they are becoming acutely aware of downgrades and default risks emanating from falling commodity prices, weakening EM economies, and rising idiosyncratic risks. This fundamentally-driven selloff is also laying bare the problems of illiquidity, through the default-driven liquidity channel as we wrote about in “A Primer on Corporate Bond Liquidity”, Oct 2014.
In short, investors are internalizing that we are in the late stages of the credit cycle. We have seen aggressive cash-raising in the past when investor sentiment has shifted so rapidly, using the VIX as a proxy for overall risk appetite. There is a reasonably tight correlation (0.62) between HY fund manager cash balances (%AUM) and the average monthly VIX level through time (Figure 1).
We indeed have also seen an increase in cash balances already, from 4.25% in May to 4.75% in August, bringing cash balances to the highest levels since February. We think future cash balances in September and beyond will easily surpass that seen earlier this year. Using a linear regression based on the relationship between cash balances and the VIX, we expect HY cash balances to increase to 6.5% of AUM in September. This would be consistent with cash levels seen in 2010/2011, during the depths of the Eurozone crisis.
Hence, we believe today’s selloff is fundamentally different than that experienced during Taper Tantrum and last summer’s swoon. Both of the latter selloffs, despite having larger retail outflows, saw little cash raising by funds (Figure 2).
In fact, during last summer, cash balances were actually drawn down, which limited the impact of these outflows. This is consistent with the fact that overall risk sentiment did not shift enough for fund managers to change their behavior; the VIX during both episodes remained subdued. This is not the case today.
Another factor we may be missing is institutional outflows, which may be moving first, with retail outflows to follow later. Anecdotal comments abound that the bid for lower quality paper is not what everyone had envisioned. While we will not get preliminary Q3 data on institutional third-party flows until later in October, there is evidence that the institutional bid, once very solid during Taper Tantrum, is starting to waver, as we saw last summer (Figure 3).
Even in Q2 ’15, there were more institutional outflows than during Taper. We would expect to see further evidence of institutional outflows based on this trend. What is our prognosis going forward? This is a re-pricing of fundamental risk as the market transitions to a more volatile environment. As our equity colleagues detail , this risky environment is likely to persist due to a structural shift in monetary policy (from easing to tightening) and rising credit risks and losses. The result: Higher HY fund cash balances should be the norm for the foreseeable future, making the prospect of a significant tightening in spreads unlikely. On the downside, there is a clear risk that a further reduction in risk appetite would trigger significant retail outflows on top of funds raising cash, which would greatly exacerbate spread widening and illiquidity. However, if volatility stabilizes, even at these high levels, incremental cash will be put to work as investors tread carefully, leading to a slow grind lower in spreads from today’s elevated levels. In the shortrun, absent a further leg down in commodities, we would expect this grind lower in spreads to occur as the market is moderately cheap by our model estimates. Over the longer-run though, the downside risk of an outflow-driven selloff looms in the background." - source UBS
"Significant outflows from high yield and loansBut High Yield and Emerging Markets have led as well to some other detonations as reported also by Bank of America Merrill Lynch in their Follow The Flow note from the 2nd of October entitled "Equities show outflows symptoms":
Amid the slowly worsening global macro backdrop, we saw a flight from risk assets into relatively safer areas this week. US high yield reported $1.5bn in outflows (-0.74%) for the week ended September 30th, which is the largest weekly outflow from the asset class since July 1st. The brunt of it was borne by ETFs, which saw -$809mn leave, while the remaining -$700mn came from HY non-ETFs. Loans reported $605mn in outflows (-0.61%), making it the worst performing asset class YTD, having lost 10% in AUM. Second worst YTD are EM bond funds, which realized outflows for the 10th week in a row, -$1.35bn (-0.52%), and now stand at -5.3% in terms of AUM lost YTD. High grade funds had a more modest $1.84bn outflow (-0.2%). The only fixed income classes recording inflows were munis, with $260mn (+0.1%), and money markets with $2.16bn (+0.1%). In aggregate, fixed income funds lost $3.57bn (-0.18%) to outflows on the week. The last week in which fixed income had more than $3bn in outflows was June 17th (-$4.17bn).
Other funds reporting outflows were equities (-$452mn) and non-US high yield (-1.49bn). Commodities and money markets recognized inflows of $268mn and $2.16bn, respectively." - source Bank of America Merrill Lynch
"Credit and equity fund flows – all negativeIf this is not indeed a "Sympathetic detonation", then we wonder what all of this is! In terms of initial blast (collapse in oil prices and surge in the US dollar), the shock wave has indeed set off the detonator in the credit market and the attached charge (leverage). This leads us to our final chart which once again shows how central banks meddling in asset prices leads to "unusual high correlations" with oil prices.
Market stress has been reflected in European fund flow data. Following a week of rising idiosyncratic risk outflows hit both credit and equity funds.
Outflows from high grade funds more than tripled from the previous week; and are now at four weeks high. The withdrawal from high yield funds quadrupled w-o-w as the asset class recorded their second biggest outflow of the year. Last week’s cumulative outflow from credit funds was the seventh in a row and the highest in four weeks. Credit ETF fund flows have also dipped into negative territory.
Equity funds inflows seem to have run out of steam. While the recorded outflow was only marginal, it remains indicative of a change in the trend. Note that this is the fifth week of the year where the asset class displays a negative flow number. Global EM debt funds continued to see more outflows; recording their tenth consecutive week of outflows. The asset class has suffered the most year-to-date with close to $15bn of outflows.
Commodities fund flows remained negative even though the outflow was not sizable it was the fifth week of continuous outflows.
The exception this week was European government bond funds, which posted a positive inflow, the first in six weeks. The asset class endured close to $7bn of outflows since the bundshock back in April." - source Bank of America Merrill Lynch
- Final chart - Correlation to Oil very positive and unusually high
"US IG Credit Spreads:
Since we hit the double-digit tights of June 2014, US IG spreads have widened to 163bps over the past five quarters. This 60% jump in spreads stems from several factors at the management level which include increased buybacks, higher than expected issuance, and a heavy M&A pipeline (see Credit Continuum: The Buyback Bonanza, April 28, 2015 and Credit Continuum: M&A Makes Its Mark, June 10, 2015). However, the macro environment has also played a large role. Market volatility remains elevated as a result of macro growth concerns, and plunging oil prices have had a major impact on the Energy sector and to a lesser extent, the Basics sector.
Oil vs. Major Asset Classes:
Earlier this year, our Cross Asset and Commodity teams noted that recent correlations between oil and US High Yield have been unusually elevated, and have only been this high once before, during the peak of the 2008/09 crisis (see Cross-Asset Dispatches: Are You Just Trading Oil?, August 16, 2015). Certainly for IG, the story has been similar.
- source Morgan StanleyYou can expect additional "Blue Mondays" going forward, more "Bayesian" price movements (+/- 4 standard deviations) and even higher positive correlations courtesy of your "local/global" central banker. This of course will lead to yet a bigger "Sympathetic detonation" to 2008, but that's another story and as any good story teller tells you: "To be continued..."
"Our duty is to believe that for which we have sufficient evidence, and to suspend our judgment when we have not." - John Lubbock, British statesmanStay tuned!