This page has been archived and commenting is disabled.
Closing Context Update: Surface Calm Hides Some Tension
From Capital Context
TSY complex back to post-NFP high yield prints with no looking back this afternoon even as S&P futures rolled over into the close.
Stocks managed to rise haphazardly, thanks to some super aggressive bids around 230ET (lifting us out of a narrow range) among other things but closed off their highs. TSYs and secondary bonds though saw net selling pressure once again following yesterday's notably larger than normal net selling in secondary corporate bond land.
The slow drift lower in the USD pretty much all day accelerated around 1ET just as selling of TSYs began - makes you wonder if the Chinese wanted to make a point after the 'negotiations' this morning eh?
The point being that TSYs and corporate bonds never bounced back even as stocks rolled over in the last hour and DXY managed to cling above the seemingly magical 74.5 level. HY and IG spreads in CDS land did manage a rally - tracking S&P futures most of the day - handily outperforming intrinsics but low beta outperformed high beta as the up-in-quality theme contonued to play out (in cash and synthetic).
Stocks outperformed credit (beta-adjusted) and for the first time in a while, HY (index) managed to outperform IG (index) on a beta-adjusted basis but we heard chatter of some small arbs (index arbitrage selling index protection versus buying underlying single-name porotection to close the 'skew') which also fits with the fact that high beta names underperformed and the widest credits underperformed (were actually wider).
IG and HY closed pretty much at recent tights, same index closing levels as 5/2, but we note that IG fair value has widened 1.7bps and HY 7bps in that period - index overlays rotated into single-name protection? index arb? Breadth has been very negative at around 3 wideners to every tightener and front-end curves have flattened notably among IG names as financials , industrials, and TMT have underperformed over that period. S&P and VIX are flatish from the close (so top-down all looks similar) but 10Y TSYs are 7bps lower, Oil $10 lower and Gold $30 lower (even as DXY is 2.2% higher).
The reason we note all of this is that we want to encourage contextual analysis. There is a great deal going on beneath the surface that is hidden by indices such as S&P, VIX, IG, and HY. As one final note, since 5/2 secondary corporate bonds (as per TRACE) have seen net selling pressure but somewhat fascinatingly (at least to us), IG has seen net selling and HY net buying with investment grade net selling dominated by AAAs and BBBs - we suspect a focus up into core IG and selling financials has caused this optical while HY flows have relatively supported HY (even if CDS has been decompressing). The relative buying in HY could also be justfied on the basis of a burst of issuance recently as concessions were snapped up.
Anyway, back to today, positive breadth in credit and some positivity clearly inspired by some positive earnings and whatever spin you shoose on last night's China trade balance (which seems like it would/should mean a drag on US GDP via net exports?). Plenty of chatter today about how low high yield yields are - according to some indices at record low yields but we note that this is not a signal of an all-clear (as some might think) as HY spreads remain well off their cycle lows (of this cycle, let alone the previous cycle).
Off topic again but worth it we think -
we want to briefly highlight a couple of supposed 'knowns' wiith regard to rates and spreads in corporate bond land
. While there is a negative correlation between the two, it is highly unstable, very much time-varying, and has a notable cognitive bias (i.e. it varies given where you are starting from). Furthermore, when managers seek exposure to HY credit (most specifically bonds) they will typically hedge the underlying rate exposure in line with their bogey (or adjust based on view) and then over-/under-weight their specific credit exposure. The critical piece here is that the duration hedges they choose (i.e. selling TSYs/IR swaps to manage away corporate bond rate risk) is different depending on where spreads trade.
TSY duration hedges adjust very quickly in a rate compressing and falling rate environment - perhaps explaining some of the recent TSY underperformance.
While so many use duration (in all its wonderful variety of calculations), HY bond durations need to be adjusted for their actual spread (or credit risk). Extreme examples illustrate the point clearly - A Govt bond has zero credit risk (well supposedly) and its 'real' duration is equal to its 'calculated' duration, however a bond in severe distress has almost no interest rate risk at all (trades on price to recovery) despite calculated durations telling you otherwise.
The point of this, as we have shown managers and traders over the years is that a real empirical duration hedge must be adjusted for credit risk and at times of low rates and high relative durations, this hedge difference from calculated can be very significant as spreads decompress quickly and rates drop fast.
The chart above shows the relative adjustment (apologies for slightly outdated chart - time constraints on our side) and how it can escalate quite quickly as rates and spreads compress. We bring this up as recent rate compression and spread compression may be putting some additional weakness into TSYs and may help explain some of the steepening as the longer-dated the bond, the larger the differential will be as duration rises.
Significant drops in the Citi Economic Surprise Index have historically portended reasonable turning points in the S&P500.
A
quick bite on the macro apple is also warranted
. We published this quick chart last week and had some very helpful comments (hattip John L!) on the potential problems associated with the underlying index's mean reversion tendencies but we thought it woirth at least sharing for discussion. The chart shows the two-month change in the Citi Econ Surprise Index poltted against the S&P. There has been significant divergence between the Citi Econ Surprise Index and the S&P for a few months now but it seems that the efforts of QE1, Lite, and 2 are responsible for adjusting the relationship over time (i.e. when the Citi index drops typically S&P follows with a decent correlation (though some lag) except when the Fed undertakes QE (of whwtever form).
We use the change here as we suspect it is more about the turning points than the actual tick by tick shifts and given the imminent end of QE2, we thought it may be useful as a check on risk appetite to consider the relatiev success of this rate of change indicator in calling intermediate tops in equities. The level of change in the index currently is dramatic to say the least and worthy of some reduction in equity weightings in our opinion.
We bring this up as we suggested today a re-entry into our ETF Arb position (which is a hasty re-entry following
last week's successful trade
) as equities have once again run away from credit quite dramatically in the last few days. We entered at just over $18 credit with a target of $15 and stop around $20 - we will track the relative shifts and evaluate exits based on both ethe ETFs and the chifts in spreads versus equity. As we mentioned above, this is a rapid re-entry for us, and that may make us a little less aggressive on the sizing of this trade but nevertheless, the indications are for some reversion of equity relative to credit in the short-term.
Contextually
, 32% of CDS were wider today (in our capital structure universe) while only 15% of equities were lower. 57% of vols rose though (which admittedly is lower than in recent weeks). Utilities and Services saw notable underperformance in credit relative to equity expectations (this is based on our capital structure model's relative expectations for equity, vol, and spreads) and Basic Materials was the only sector ion which spreads relatively outperformed today (equities weaker than spreads would have expected).
Low beta names dominated action with spreads tighter on average (relative to unch or slightly wider on average in high beta names) and a modest oputperformance by low beta names also in stocks today. Vol moves were small but low beta saw vol rise more than high beta. We did point out the drastic drop in implied correlation
this morning
which appears to have been single-name vol bid relative to index vol quite notably which given the net rise in single-name vol (moneyness adjuasted) among our capital structure universe (with VIX dropping) is confirmed somewhat. We remain of the opinion that implied correlation is more a carrry strategy for now than a big positional strategy even though the tri-party m,arket of Financials, Energy/Basic Materials, and Others would suggest a drop in index implied correlation is likely.
CMBX tranches and ABX tranches continued their shift from senior to junior and down suggestingg that the double-dip in housing is being played out here more so than in the homebuilders themselves who seem able to hide behind mountains of tax carry forwards.
Bottom-line
is that while the indices show apparent strength, we see some less than resilient behavior under the surface and in the short-term at least prefer credit to equity and remain long up-in-quality positions.
Europe
More chatter over peripheral bonds, bailouts, and the incredible finding of the WSJ's editing of the True Finns statement was food for thought today. SovX rallied quite notably (though intrinsics not so much) as financials outperformed non-financials. Much was amde about the positive steps and so on and on but we want to point out something very interesting.
We talked about the ban on naked shorts in sovereign bonds (fair enough) and the fact that we suspected that there was actually very little naked shorts and most was either covering longs at a loss. Furthermore, we warned that an extension of this to banning CDS was worrisome - today saw a great example of what we said last week. Greek CDS widened 40bps or so (breaking 1400bps closing for the first time) while GGBs managed to rally 19bps or so in 5Y. Compression in Greek bonds (as well as other PIIGS) was noted as positive by headline makers BUT ever the water-pourers we want to highlight that this was a major basis compression contonuing and that we striongly suspect that the only investors actually buying GGBs are basis traders.
After reaching some rather epic levels of basis (that we highlighted as very juicy last week), the spread between GGb 5Y and 5Y CDS for Greece (adjusted for Germany/Bund relative performance) has compressed 125bps! from over 152bps to around 26bps - much more 'normal'. We suspect Portugal will be next for the basis compression as its basis is very wide indeed so expect bond yields to drop and CDS to widen in the short-to-medium term - at least in the belly of the curve.
Asia
Very modest compression in Australia, Asia Ex-Japan, and Japan spreads overnight. Our AXJ-Japan spread compression trade has stalled at around 15bps here, well south of our entry at around 32bps (wides of 37bps) but we remain in position looking for close to pairty to re-appear - especially now TEPCO is being nationalized.
Index/Intrinsics Changes
CDX16 IG
-0.76bps to 88.49 ($0.03 to $100.43) (FV -0.34bps to 88.75) (38 wider - 75 tighter <> 72 steeper - 52 flatter) - No Trend.
CDX16 HVOL
-3.6bps to 148.2 (FV -0.15bps to 148.13) (14 wider - 15 tighter <> 16 steeper - 14 flatter) - No Trend.
CDX16 ExHVOL
+0.14bps to 69.63 (FV -0.39bps to 70.72) (25 wider - 71 tighter <> 39 steeper - 57 flatter).
CDX16 HY
(30% recovery) Px $+0.2 to $103.205 / -4.9bps to 421.9 (FV -3.22bps to 413.02) (21 wider - 74 tighter <> 60 steeper - 37 flatter) - Trend Tighter.
LCDX15
(70% recovery) Px $+0.06 to $101.375 / -1.62bps to 237.13 - No Trend.
MCDX15
+1.75bps to 118.75bps. - Trend Tighter.
ITRX15 Main
-1.53bps to 96.97bps (FV-0.92bps to 100.06bps).
ITRX15 HiVol
-1.48bps to 134.02bps (FV-1.19bps to 133.93bps).
ITRX15 Xover
-4.51bps to 355.49bps (FV-3.84bps to 344.85bps).
ITRX15 FINLs
-3.93bps to 135.32bps (FV-2.41bps to 135.39bps).
DXY
weakened 0.25% to 74.55.
Oil
rose $0.78 to $103.33.
Gold
rose $2.38 to $1516.13.
VIX
fell 1.25pts to 15.91%.
10Y US Treasury yields
rose 5.1bps to 3.21%.
S&P500 Futures
gained 0.88% to 1354.5.
Spreads were mixed in the US with IG tighter, HVOL improving, ExHVOL weaker, and HY rallying. IG trades 3.6bps tight (rich) to its 50d moving average, which is a Z-Score of -1.4s.d.. At 88.49bps, IG has closed tighter on only 28 days in the last 606 trading days (JAN09). The last five days have seen IG flat to its 50d moving average. HY trades 12bps wide (cheap) to its 50d moving average, which is a Z-Score of -0.9s.d. and at 421.86bps, HY has closed tighter on only 26 days in the last 606 trading days (JAN09). Indices generally outperformed intrinsics with skews mostly narrower.
Comparing the relative HY and IG moves to their 50-day rolling beta, we see that HY outperformed by around 1.8bps. Interestingly, based on short-run empirical betas between IG, HY, and the S&P, stocks outperformed HY by an equivalent 9.5bps, and stocks outperformed IG by an equivalent 2.4bps - (implying IG underperformed HY (on an equity-adjusted basis)).
Among the IG16 names in the US, the worst performing names (on a DV01-adjusted basis) were GATX Corporation (+3.07bps) [+0.02bps], Metlife, Inc. (+3bps) [+0.02bps], and Progress Energy Inc (+1.45bps) [+0.01bps], and the best performing names were SLM Corp (-4.03bps) [-0.03bps], Expedia, Inc. (-3.17bps) [-0.02bps], and Conagra Foods Inc (-2.5bps) [-0.02bps] // (absolute spread chg) [HY index impact].
Among the HY16 names in the US, the worst performing names (on a DV01-adjusted basis) were Eastman Kodak Co. (+71.14bps) [+0.53bps], K Hovnanian Enterprises, Inc. (+68.57bps) [+0.51bps], and KB Home (+10.62bps) [+0.1bps], and the best performing names were MBIA Insurance Corporation (-85.64bps) [-0.59bps], Dean Foods Co. (-53.26bps) [-0.5bps], and PMI Group Inc/The (-30.33bps) [-0.24bps] // (absolute spread chg) [HY index impact].
»
- advertisements -
Nice post. This market is suspicious, but I would not be surprised to see it take off to the upside till sometime in the summer and then the reality of the economy should win over. Probably best to wait in cash; however, I am short and will most definitely come to regret it. I doubled down on my short position today like the fool that I am. If there is not at least a 20% correction in 2011 I am going back to being a Buffett/Graham clone, holding a huge cash position for the occasional fat pitch, and making fun of technical analysis.
CC - very well done. Your post has reconfirmed my recent long Usd/Jpy position. Will look to add to short 30 yr as well. Cheers.
The vix is trading like a utility stock ... unbelievable!
good work. what lookback period do you recommend for calculating empirical durations?
Tnx for the question. Actually it is not so much a lookback period as regressing cohorts of spreads (bucketed by say 50bps groups) against ALL empirical moves in excess of what a 'common' duration measure would have expected - what you end up with is a convex curve sloping from upper left at 1 to lower right at 0ish (y-axis is multiple of calculated duration to use and x-axis is spread). Once you have that - created across ALL bonds and ALL time (or as much decent history as you have) you can create a functional form of that relationship that is closed form to use for the adjustment. Note - the same relationship BUT inverse shape also enables this to map out an 'equity duration' - in so much as you end up recreating an empirical relationmshipo that mimics what you would expect from a structural model (Merton etc.). Drop us a line if you want more details. cheers.
Gotcha, thanks for the detail
As an update, IG has traded down sub 88bps after-hours and while HY is tighter, nothing like as much as IG. The shift was a little gappy and reminds us of our recent comments on the market being short gamma and exaggerating moves in the index (as swaption vol traders follow their equity brethren).