📄 Extracted Text (604 words)
4 September 2015
US Fixed Income Weekly
removed. 8%. This puts the current episode's maximum decline of 11% in
good company with historical episodes. But other aspects of the historical
record suggest that we may not out of the woods yet. For one, previous
episodes of shocks to the volatility risk premium tend to last substantially
longer than two weeks; they take an average of 50 days, or 38 days if the
crisis-era episodes are excluded. (While there are some previous examples of
"short" vol shocks in early 2007 and October 2014, these might be more
appropriately viewed as early warning indicators for more extended periods of
high volatility, and less as standalone examples.) Also, the equity markets have
tended to hit rock bottom during these episodes an average of 30 days or so
after implied volatility rises. The S&P 500 hit its low just four days after the vol
shock, which seems out of line with historical patterns.
Figure 4: Equity implied volatility relative to realized vol IFigure 5: S&P level during elevated volatility episodes
2.25 2.300
2.100
2.00
1.900
1.75 1)00 1-- — — —
1.500
1.50 1.300 r
1.100
1.25
9C0
too 700
2000 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2004 2005 2006 2037 2008 2039 2010 2011 2012 2013 2014 2015
—3m Implied/12m Wind Vol Ratio. SPX SPX . ......iinplird/RitaR zed a 1.Sx
50iO4/ bunt" pant 2 komp. bold" an
The shock to the VIX index can be attributed to three inter-related measures of Figure 6: Seasonality trends in HY/IG
equity volatility: the general level of at-the-money volatility relative to the trend
of realized volatility, the premium for options expiring in the near-term (1m) Awsra96, M OAS ovAwlPy ch9eve
relative to somewhat longer-expiry (3m) options, as well as the premium for
it . • • •
out-of-the-money strikes over at-the-money strikes. An additional way to to
measure the magnitude of last week's equity market shock is to consider the
elevated level of volatility risk premium, measured here as the degree to which
option-implied volatility exceeds realized volatility. Finally, it's worth observing
that credit markets are also participating in these developments. The implied-
to-realized ratio on the CDX indices is also elevated, sitting in the 90th" 20 0
percentile over the last 3.5 years for 1m options on IG CDX, while the equity Mf ha MV Oen In 17 Aut Sp On 00. Dee
• 07070*07177) Stu fircrod7Tantninsi
risk premium is in the 99th percentile over the same period.
ConG fusions
Overall, we find market moves over the past week were in line with our
expectations, directionally, although their speed, volatility, and reversals were
certainly as much a surprise to us as they were to most other investors. We
thought equities and other risk asset classes were much closer to what we
perceive to be fair value at their bottom last Tuesday than they were following
a retracement. Developments in China could have significant repercussions for
broader EM universe, and we don't find EM credit spreads to be properly 0
reflecting those consequences. We see main risks associated with EM credit Ian tea My An. MM An Si hag .5•0 Oci Nov On
•CPAnCr 14
/ ) aftteAren0074007****
assets being forced to re-price more substantially and having second-order
effects on US credit markets. Additionally, historical evidence suggests that San/ Dana ant
periods of extreme volatility similar to those witnessed over the past few
sessions tend to exhibit propensity for aftershocks, usually lasting for weeks if
not months.
Page 26 Deutsche Sank Securities Inc.
CONFIDENTIAL - PURSUANT TO FED. R. CRIM. P. 6(e) DB-SDNY-0051327
CONFIDENTIAL SDNY_GM_00197511
EFTA01362030
ℹ️ Document Details
SHA-256
32e37d58e06a4dfd98a17d8b5bde9026ef528bdf498b2af2cc2bd0b5eee31347
Bates Number
EFTA01362030
Dataset
DataSet-10
Document Type
document
Pages
1
Comments 0