Second Quarter 2017 (data as of June 30)
A review of financial market themes and developments
The Volatility Paradox: Tranquil Markets May Harbor Hidden Risks Financial markets were exceptionally calm in the second quarter by most measures. Only three times in the past 90 years has volatility been so low: twice during bull markets in the 1960s and 1990s, and once in the lead-up to the financial crisis of 2007-09 (see Figure 1). Is today’s low volatility a sign of calm or a threat to financial stability — or both? This edition of the OFR’s Financial Markets Monitor investigates the volatility paradox: the possibility that low volatility leads investors to behave in ways that make the financial system more fragile and prone to crisis. We analyze three channels through which a prolonged period of low market volatility may introduce financial stability risks: increased leverage, reduced hedging, and institutional investors’ use of risk-management models. We find some supportive evidence of these channels at work, but better data are needed to make definitive conclusions. Volatility alone is not a good indicator of impending financial stress.
Figure 1: S&P 500 Index 90-day Realized Volatility (percent) Volatility in U.S. equity markets is the lowest in decades 75 6.47 on May 14, 2017 7.12 on Jan. 2, 2007
50
6.45 on May 2, 1995 4.01 on April 23, 1964
25
0 1928 1938 1948 1958 1968 1978 1988 1998 2008 2018 Note: Volatility is the standard deviation of daily returns over 90 days expressed as annualized percent change. Source: Bloomberg Finance L.P.
Key findings •
Volatility for most asset classes across the world fell below historical averages during the second quarter. In some cases, volatility is near all-time lows. Drivers of low volatility may include expectations that the long U.S. economic expansion and still-easy funding conditions will persist.
•
Some institutional investors have adapted by increasing leverage and the use of yield-enhancing strategies.
•
Shocks could produce procyclical responses if market participants use measures of realized volatility to manage the risk of their portfolios.
This monitor reflects the best interpretation of financial market developments and views of the staff of the Office of Financial Research (OFR). It does not necessarily reflect a consensus of market participants or official positions or policy of the OFR or the U.S. Department of the Treasury. Contributors: Meraj `Allahrakha, Viktoria Baklanova, Danny Barth, Ted Berg, Jill Cetina, Dagmar Chiella, Arthur Fliegelman, Dasol Kim, Francis Martinez, John McDonough, Philip Monin, Mark Paddrik, Eric Parolin, and Daniel Stemp.
Volatility alone is a weak risk indicator. Volatility measures for most asset classes across global financial markets fell below their historical averages during the second quarter (see Figure 2). Some measures approached all-time lows (for example, see Figures 1 and 3), which may have been driven by expectations that the long U.S. economic expansion and still-easy funding conditions will persist. There are two types of volatility: realized and implied. Realized volatility reflects the historical price fluctuations of an asset. Implied volatility is forwardlooking. It captures the market’s expectation of future price fluctuations of an asset, derived from the options markets.
Figure 2: Realized Volatility by Asset Class (z-score) Volatility has declined across major asset classes and markets 9
U.S. equities U.S. interest rates Global currencies Global equities (ex-U.S.)
6 3 0 -3 2007
2010
2013
2016
Note: Realized volatility is the standard deviation of daily returns over 30 days, expressed as annualized percent change. U.S. equities are represented by the S&P 500 index. U.S. interest rates are the weighted average of the Treasury yield curve. Global currencies are based on weights from JPMVXY index. Global equities are MSCI All Countries World Excluding U.S. Index. Standardization uses data since Jan. 1, 1993. Sources: Bloomberg Finance L.P., OFR analysis
When implied volatility exceeds realized volatility, the Figure 3: Chicago Board Options Exchange Volatility Index (VIX) difference reflects the extra return investors demand (percent) volatility on equities has fallen to near all-time lows to hold a security solely because it is volatile. This Implied 100 difference is known as the volatility risk premium. 80
9.75 on June 2, 2017
9.31 on Dec. 22, 1993 One of the most widely cited measures of implied volatility is the Chicago Board Options Exchange 60 Volatility Index (VIX). The VIX is the 30-day implied volatility of options on the benchmark S&P 40 500 equity index. A low VIX doesn’t necessarily signal that severe financial stress is unlikely. For 20 instance, the VIX provided no advance warning of 0 extreme volatility in the months leading up to the 1995 2000 2005 2010 2015 financial crisis. Realized volatility of the S&P 500 Note:1990 Implied volatility is derived from options markets and is the expected standard index was often substantially higher than the VIX had deviation of daily returns over the next 30 days, expressed as annualized percent predicted 30 days earlier (represented by the blue dots change. Source: Bloomberg Finance L.P. over the 45-degree line in Figure 4). The relationship between realized and implied volatility for other asset Figure 4: VIX and Realized Volatility of S&P 500 Index (percent) The VIX did not predict the global financial crisis classes followed a similar pattern during the crisis. 100
Market risks may seem low when volatility is low. However, low volatility may also serve as a catalyst for market participants to take more risk, thereby making the financial system more fragile. This phenomenon is known as the volatility paradox. Low volatility directly incentivizes risk-taking. Lower volatility may contribute to greater leveraging and risk-taking through at least three channels. The first channel is through changing asset-return correlations, which tend to increase when markets are volatile. Low correlations could entice investors to
OFR MARKETS MONITOR
Realized volatility
75
50 Global financial crisis (8/1/2008 to 11/1/2008) Other dates (3/1/1993 to 5/31/2017)
25
VIX
0 0
25
50
75
100
Note: Realized volatility is the standard deviation of daily returns over 30 days, expressed as annualized percent change. Sources: Bloomberg Finance L.P., OFR analysis
Second Quarter 2017 | 2
accumulate risky exposures, believing they are Figure 5: 3-Month Moving Average of S&P 500 Sector diversified. Prolonged periods of low volatility may Correlations, VIX Index (correlation, percent) further decrease correlations, encouraging further Correlations between sectors have fallen amid low volatility risk-taking. This procyclical behavior increases 1.00 Sector correlation (left) investors’ risk of loss from a systematic shock, when VIX index (right) volatility spikes and asset-return correlations revert to 0.75 historical levels.
75
Some evidence exists that this channel may be at work in equity markets. Sector correlations have declined significantly during the past two years, while volatility has remained low (see Figure 5).
25
50
0.50 0.25
0 Second, low volatility could encourage the use of 0.00 Jan Jan Jan Jan Jan Jan Jan other yield-enhancing strategies, such as selling deep 2005 2007 2009 2011 2013 2015 2017 out-of-the-money put options (those with a strike Note: S&P 500 sector pairwise monthly correlation; 3-month moving average. price substantially below current prices). Investors Sources: Bloomberg L.P., OFR analysis collect a premium from selling these options, but can Figure 6: Margin Debt Balance over Market Capitalization and be obligated to purchase the underlying assets if the S&P 500 Index 30-day Realized Volatility (percent) price drops below the strike price. Investors who Realized volatility has fallen as investors increased margin debt accumulate these risky exposures could be more 3 100 Margin debt / market capitalization (left) likely to experience financial stress if prices sharply S&P 500 index realized volatility (right) decline. Available data on investor portfolios are not 75 sufficient to assess this channel adequately.
Third, low volatility can directly incentivize leveraging by lulling investors into underestimating the odds of a volatility spike. One measure of marketwide leverage is the ratio of margin debt to market capitalization. This measure is imperfect because it doesn’t account for other positions on investor balance sheets, including derivatives positions. Figure 6 uses margin debt balances and market capitalization data from the New York Stock Exchange. The ratio increased from 2002 to 2007 amid low volatility, declined after the crisis, and has been climbing since as volatility again reached longterm lows.
2
50 25
1 Jan 2005
0 Jan 2007
Jan 2009
Jan 2011
Jan 2013
Jan 2015
Jan 2017
Note: Values are the New York Stock Exchange (NYSE) market capitalization and margin debt balances of its members. Dealer margin debt balances may reflect positions on securities not listed on the NYSE. Realized volatility is the standard deviation of daily returns over 30 days expressed as annualized percent change. Sources: Haver Analytics, OFR analysis
Evidence also exists that some large investors are highly leveraged and, for that reason, may be susceptible to volatility events. For example, the top decile of macro and relative-value hedge funds has been leveraged about 15 times in recent quarters. These funds combined account for more than $800 billion in gross assets, about one-sixth of all hedge fund assets. Low volatility could also disincentivize investor hedging.
OFR MARKETS MONITOR
Second Quarter 2017 | 3
Another way investors may adapt to low volatility is Figure 7: SPY Options Held for Hedging Purposes (percent) by reducing their hedging of risky positions. This Investors are less hedged compared to the pre-crisis period behavior was particularly relevant in recent years, 100 when historically low interest rates pressured Put hedging rate investors to reach for yield by holding more lowerCall hedging rate rated fixed-income securities and more equities (see 90 the OFR’s 2016 Financial Stability Report). OFR analysis of options trading suggests that investors have reduced their hedging of market exposure. 80 Investor hedging activity is difficult to measure, although it can be captured to some extent using 70 contracts outstanding in current-month SPY options. SPY is an exchange-traded fund that mirrors the benchmark S&P 500 equity index. Traders commonly 60 sell SPY options to hedge equity market exposure. 2005 2007 2009 2011 2013 2015 2017 Options give investors the right, but not the obligation, to buy or sell a specific security at a Sources: OptionsMetrics, OFR analysis specific strike price and time. A call option is a right to buy; a put is a right to sell. Figure 8: VIX Futures Noncommercial Net Total (contracts)
Options with a strike price near the current price of Speculators increased short bets on VIX to the most since 2004 50,000 SPY are said to be “at the money.” Contracts with a strike price far from the current price are “away from Net position the money.” These options are less likely to be held 0 for hedging purposes and instead may represent yield-enhancing strategies. Investor hedging activity is captured through a hedging rate, calculated as the proportion of contracts on SPY options that is “at -50,000 the money” versus “away from the money.” Hedging rates are currently lower on average than in the years immediately preceding the financial crisis (see Figure -100,000 7), suggesting a structural change in hedging activities after the crisis. However, the evidence is somewhat -143,845 on June 20, 2017 mixed. Considerable variation has occurred since -150,000 2004 2006 2008 2010 2012 2014 2016 2010, and current levels appear to be higher relative to 2014 for both call and put hedging ratios. The Sources: Bloomberg Finance L.P., Commodity Futures Trading Commission absence of sharper measures of aggregate hedging activities makes drawing definitive conclusions difficult, though these hedging ratios at least suggest significant differences before and after the crisis. The Commodity Futures Trading Commission (CFTC) collects data on an alternative measure of hedging activity using positions of futures traders. CFTC data categorize hedge funds and other investors as “non-commercial,” or speculative, traders. As of May 2017, the net short position on VIX futures of non-commercial traders sat at levels larger than even before the crisis (see Figure 8). Common volatility strategies involve taking short positions in longer-dated contracts and long positions in shorter-dated contracts. Reduced
OFR MARKETS MONITOR
Second Quarter 2017 | 4
hedging in these strategies would imply shorting in Figure 9: Cumulative Yield Change in 10-year Government Bonds the aggregate, consistent with Figure 8. However, (basis points) establishing a direct link without more granular data VaR shocks may have deepened past selloffs in bond markets is difficult. 0 Japanese Government Bond market Together, these data suggest that some investors may have adapted to the low-volatility environment by reducing risk hedges and increasing speculative bets. Data limitations temper the findings to some extent, and leave opportunities for further analysis. With less hedging, these investors’ balance sheets may be less resilient to large volatility shocks when volatility returns to financial markets. Value-at-Risk models may give faulty signals in low-volatility markets.
(6/12/2003 to 9/10/2003) U.S. Treasuries taper tantrum (5/2/2013 to 7/31/2013)
25 50 75 100 125 Day 0 Day 10
Day 20
Day 30
Day 40
Day 50
Day 60
Day 70
Day 80
Day 90
Note: The vertical axis is inverted to reflect lower bond prices as yields increase. Horizontal axis is the number of days since the beginning of the sell-off period. Sources: Bloomberg Finance L.P., OFR analysis
Low realized volatility can affect the behavior of banks, hedge funds, and other asset managers that use a risk management framework based on realized volatility, including some Value-at-Risk (VaR) measures. About 40 percent of large hedge funds, representing about 62 percent of gross hedge fund assets, regularly calculate VaR statistics for their funds, according to Form PF data collected by the Securities and Exchange Commission (SEC). VaR measures the risk of investments. It captures how much value investments might lose over a set time. Although VaR can be a valuable riskmanagement tool, overreliance on VaR when volatility is low could result in procyclical behavior that makes investors more vulnerable to volatility shocks if market conditions change abruptly. A decline in realized volatility can reduce a portfolio’s VaR, allowing market participants to increase position sizes without exceeding predefined VaR risk limits. The reverse is true when volatility rises. In that case, VaR-sensitive investors may be forced to simultaneously sell assets to get their portfolios below risk limits. A selloff induced by a VaR shock can become selfreinforcing as liquidity dries up and as deleveraging occurs. Some market observers believe VaR shocks contributed to selloffs in the Japanese government bond market in 2003 and in the U.S. Treasury market during the 2013 taper tantrum (see Figure 9). Longterm investors that are not sensitive to VaR, such as pension funds and insurance companies, may not step in and provide liquidity unless prices fall sharply. OFR MARKETS MONITOR
Second Quarter 2017 | 5
Figure 10: U.S. G-SIBs' Combined Trading Books ($ billions)
Most large U.S. banks report data on the VaR of their Big banks' VaR has collapsed but portfolio size is little changed trading books in quarterly 10-Q filings to the SEC. 1.00 These data show a dramatic decline since 2010 in the 2,000 VaR of banks’ trading books, without a commensurate decrease in the fair value of those 1,500 0.75 trading books (see Figure 10). All else being equal, this change suggests that the reduction in VaR may 1,000 0.50 reflect falling realized volatility rather than a decline in the size of banks’ trading books during the period. 500 0.25 If volatility rises and banks aim to keep their VaR Fair value of trading book (left) stable, the banks would need to shrink their trading Trading book VaR (right) books. Another possibility is that the declining VaR 0 0.00 is evidence that banks have reduced the overall Mar Mar Mar Mar Mar Mar Mar Mar 2010 2011 2012 2013 2014 2015 2016 2017 market risk in their portfolios, in part responding to additional regulatory oversight. A definitive Note: G-SIB = Global systemically important bank Bank 10-Q forms filed with Securities and Exchange Commission, OFR conclusion is difficult without detailed data on dealer Sources: analysis positions. Targeting a specific level of volatility has recently become an investment strategy. Many institutional investors now are holding so-called “volatility control funds” in their portfolios. Assets under management in variable annuity volatility control funds rose to $325 billion at the end of 2016 (see Figure 11). These funds make asset allocation decisions aimed at maintaining a stable level of volatility for their whole portfolios. If volatility were to rise suddenly in a previously stable asset class, these funds may be forced to rebalance and sell assets. These investors’ activities could have a procyclical effect on asset prices and exaggerate volatility. Conclusion
Figure 11: Variable Annuity Volatility Control Funds ($ billions, count) Variable annuity volatility control funds have more than doubled in size 300
400
Number of funds (left) Assets under management (right)
300
200
200 100
100
0 Dec 2010
Dec 2011
Dec 2012
Dec 2013
Dec 2014
Dec 2015
0 Dec 2016
Source: Milliman Financial Risk Management LLC
Prolonged low market volatility may introduce financial stability risks through at least three channels. First, investors could respond by directly taking on more leverage and risk. Second, investors could reduce hedging activities. Third, institutional investors’ use of VaR or other risk-management models that have realized volatility as a key input could lead them to take more risk. A spike in volatility can result in outsized investor losses from sharp asset price changes. Data limitations hinder the ability to make definitive conclusions regarding the extent to which these channels are at work. However, the evidence is consistent with these channels operating and suggests the need for further analysis.
OFR MARKETS MONITOR
Second Quarter 2017 | 6
Selected Global Asset Price Developments
OFR MARKETS MONITOR
Second Quarter 2017 | 7
Select U.S. Int erest Rat es U.S. T reasury yields and yield curve percent 3
U.S. T reasury t erm premium (basis point s)
bas is po ints 2-year (Left Axis ) 10-year (Left Axis ) 10-year - 2-year s pread (Rig ht Axis )
10-year 250
225
200 2
2-year
300 250 200 150
175 100 150
50
1
0
125
0
100
-50
75 O ct 2016
Jul 2016
Ja n 2017
Ap r 2017
Jul 2017
2010
2011
2012
2013
2014
2015
2016
2017
Note : Adrian, Crump, & Moe nch mode l S ource: Bloomberg Finance L.P .
S ource: Bloomberg Finance L.P .
Professional vs. market -implied U.S. inflat ion expect at ions (percent )
Short -t erm market rat es (percent )
Survey o f Pro fes s io nal Fo recas ters 10-year annual averag e 5y5y fo rward breakeven rate CPI current
3.0
-100
1.4
1-m o nth Treas ury bill GCF Treas ury Repo
3-m o nth LIBOR
1.2 2.5 1.0 2.0 0.8 1.5
0.6
1.0
0.4
0.5
0.2
0.0
0.0
-0.5 S ep No v Ja n M a r M a y Jul S ep No v Ja n M a r M a y Jul 2015 2015 2016 2016 2016 2016 2016 2016 2017 2017 2017 2017
S ource: Bloomberg Finance L.P .
2016
S ep 2016
No v 2016
May 2017
Mar 2017
Ja n 2017
Jul 2017
S ource: Bloomberg Finance L.P .
T hree-mont h Eurodollar fut ures (percent )
Money market and policy int erest rat es (percent )
Jun-30 May-17 Apr-17 FOMC pro jectio ns fro m June 2017
4.5
-0.2 Jul
1.5
4.0
1.25
GCF Treas ury repo Fed funds effective Interes t o n exces s res erves Revers e repo facility
3.5 1.0
3.0 2.5
0.75
2.0 1.5
0.5
1.0 0.5
0.25
0.0 Jul 2017
Ja n 2018
Jul 2018
Ja n 2019
Jul 2019
Ja n 2020
Note s: T he high and low points of the De c FOMC proje ctions are the maximum and minimum fore casts. T he re ctangle re pre se nts the me dian. S ource: Bloomberg Finance L.P .
0.0 2011
2012
2013
2014
2015
2016
2017
S ource: Bloomberg Finance L.P .
1
U.S. Corporat e Debt Market s U.S. corporat e bond opt ion-adjust ed spreads (basis point s) Inves tm ent g rade (Left Axis )
U.S. corporat e CDS indexes (basis point s) Inves tm ent g rade (Left Axis )
Hig h yield (Rig ht Axis )
240
Hig h yield (Rig ht Axis )
150
600
1000
220
120 500 800
200
90 180
400 600
160 140
400
60
300 30
120 100
200 O ct 2016
Jul 2016
Ja n 2017
Ap r 2017
Jul 2017
S ource: Haver Analytics
U.S. non-financial credit gross issuance ($ billions) Inves tm ent Grade
Hig h Yield
200
0 Jul 2016
O ct 2016
Ja n 2017
Ap r 2017
Jul 2017
Note : Five -ye ar maturity CDS Inde x S ource: Bloomberg Finance L.P .
U.S. corporat e credit fund flows ($ billions)
Leverag ed Lo ans
Hig h yield
Leverag ed lo ans
10
150
5 100 0
-5 50 -10
-15
0 Aug O ct D ec Feb Ap r Jun Aug O ct D ec Feb Ap r Jun 2015 2015 2015 2016 2016 2016 2016 2016 2016 2017 2017 2017
S ources: Dealogic, S tandard & P oor's Leveraged Commentary & Data
Leveraged loan issuance by use of proceeds (percent ) M&A/LBO Other
100
Dividend/Buyback
Jul 2016
O ct 2016
Ja n 2017
Ap r 2017
Jul 2017
Note : Flows data are re le ase d with one -month lag. S ource: Haver Analytics
Leveraged loan price act ivit y
Refinancing 110
90 80 70
105
60 50 40 100
30 20 10 0
95 2000
2002
2004
2006
2008
2010
2012
2014
2016
Note : Data for 2017 are ye ar-to-date as of January. S ources: S tandard & P oor's Leveraged Commentary & Data, OFR analysis
Ja n Ja n Ja n Ja n Ja n Ja n Jul Jul Jul Jul Jul Jul 2012 2012 2013 2013 2014 2014 2015 2015 2016 2016 2017 2017
Note s: S&P Le ve rage d Loan Inde x. Inde x 100=January 01, 2012. S ource: Bloomberg Finance L.P .
2
Primary and Secondary Mort gage Market s Primary mort gage rat es (percent ) 5-year/1-year adjus table rate
MBS yield and opt ion-adjust ed spread t o U.S. T reasury securit ies
30-year fixed
Current co upo n (Left Axis )
5
percent
Spread (Rig ht Axis ) bas is po ints
4
120
100
4 3
80 3
2 60
2
1 O ct 2016
Jul 2016
Ja n 2017
Ap r 2017
Jul 2017
S ource: Bloomberg Finance L.P .
40 O ct 2016
Jul 2016
Ja n 2017
Ap r 2017
Jul 2017
S ource: Bloomberg Finance L.P .
Convent ional mort gage severe delinquencies (percent , 90+ days lat e, seasonally adjust ed)
30-year home mort gage fixed and jumbo rat es and spread percent bas is po ints 5.5 30-year fixed (Left Axis ) 30-year jum bo (Left Axis ) 30-year jum bo -co nfo rm ing s pread (Rig ht Axis )
Prim e
120
Subprim e
16 100
5.0
80
12
4.5 60
8
4.0 40
4
3.5
20
3.0 Jul 2016
0
0 O ct 2016
Ja n 2017
Ap r 2017
2007
Jul 2017
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
S ource Haver Analytics
S ource: Bloomberg Finance L.P .
Refinance and purchase loan applicat ions Purchas e Index (Left Axis ) Go v Refi Index (Left Axis ) Co nv Refi Index (Left Axis ) Refi % o f to tal apps (Rig ht Axis ) percent 240
70 65
190 60 140
55 50
90 45 40 Jul 2016
40 O ct 2016
Ja n 2017
Ap r 2017
Jul 2017
Note : Inde x 100 = July 01, 2016. S ource: Bloomberg Finance L.P .
3
Equit y Market s Global equit y indices 135
S&P 500 Shang hai
U.S. equit y indexes MSCI EM Nikkei 225 Euro Sto xx 50
S&P 500
NASDAQ
Rus s ell 3000
180
125 140
115 100 105
60 95
85
20 O ct 2016
Jul 2016
Ja n 2017
Ap r 2017
Note : Inde x = July 01, 2016. S ource: Bloomberg Finance L.P .
140
2003
2006
2009
2012
2015
Note : Inde x 100 = Jan 01, 2000. S ource: Bloomberg Finance L.P .
S&P 500 sect or performance S&P 500 Energ y
2000
Jul 2017
Financials
S&P 500 price-t o-earnings and price-t o-book rat ios (mult iple)
Co ns um er s taples
Price-to -earning s (Left Axis )
Price-to -bo o k (Rig ht Axis )
21
4
17
3
13
2
130
120
110
100
90
80 Jul 2016
O ct 2016
Ja n 2017
Ap r 2017
Jul 2017
Note : Inde x 100 = July 01, 2016. S ource: Bloomberg Finance L.P .
1
2012
2014
2016
S ource: Bloomberg Finance L.P .
U.S. equit y valuat ions: Shiller CAPE (rat io) 50
9
50
S&P 500 implied volat ilit y and opt ion skew (percent ) VIX
80% - 120% Skew
30
40
40
30
30
20
20
10
10
20
10
0
0 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010
0
Note s: CAPE is the ratio of the monthly S&P 500 price le ve l to trailing te n-ye ar ave rage e arnings (inflation adjuste d). S ources: Haver Analytics, Robert S hiller
Jul 2016
O ct 2016
Ja n 2017
Ap r 2017
Jul 2017
Note s: Option ske w is the diffe re nce be twe e n thre e -month implie d volatility of out of the mone y puts and calls with strike s e qual distance from the spot price (+/- 20 pe rce nt). Highe r value s re fle ct gre ate r de mand for downside risk prote ction. S ource: Bloomberg Finance L.P .
4
Volat ilit y Implied volat ilit y by asset class (Z-score)
Realiz ed volat ilit y by asset class (Z-score)
U.S. equities (VIX) Glo bal currencies (JPMVXYGL) U.S. Treas uries (MOVE) Averag e
3
2
2
1
1
U.S. interes t rates
U.S. equities
0
0
-1
-1
-2
-2
-3
-3
2013
2013
2014
2015
2016
Global equit y indexes 1-mont h implied volat ilit y (percent ) Euro s to xx 50 UK FTSE 300
2015
2016
2017
Note s: T hirty-day re aliz e d volatility. Equitie s base d on S&P 500 inde x, inte re st rate s base d on we ighte d ave rage of T re asury yie ld curve , FX base d on we ights from JPMVXY inde x. Standardiz ation use s data going back to January 01, 1993. S ources: Bloomberg Finance L.P ., OFR analysis
Slopes of implied volat ilit y curves (basis point s)
SP500 Germ an Dax Japan Nikkei MSCI EM
800 50
40
G10 FX (Left Axis ) S&P 500 (Left Axis ) 2-year USD Swaptio n Rate (Rig ht Axis ) 10-year USD Swaptio n Rate (Rig ht Axis )
30
400
20
200
10
0
20
10
0
-200
-10
-400
-20
-600 Jul 2016
0 S ep No v Ja n M a r M a y Jul S ep No v Ja n M a r M a y Jul Jul 2015 2015 2015 2016 2016 2016 2016 2016 2016 2017 2017 2017 2017
S ource: Bloomberg Finance L.P .
Opt ion skew by asset class (z -score)
40
600
30
3
2014
2017
Note s: Z-score re pre se nts the distance from the ave rage , e xpre sse d in standard de viations. Standardiz ation use s data going back to January 01, 1993. S ources: Bloomberg Finance L.P ., OFR analysis
60
Glo bal FX Averag e
3
-30 S ep 2016
No v 2016
Ja n 2017
Mar 2017
May 2017
Jul 2017
Note s: Se ve n-day moving ave rage . Slope re pre se nts diffe re nce be twe e n one ye ar and one -month maturitie s. G10 FX base d on we ights from De utsche Bank's CVIX inde x. S ources: Bloomberg Finance L.P ., OFR analysis
Volat ilit y of equit y volat ilit y
U.S. equities U.S. interes t rates Glo bal currencies Averag e
140
2
120
1 0
100
-1 -2
80
-3 Jul 2016
O ct 2016
Ja n 2017
Ap r 2017
Jul 2017
Note s: Option ske w is the diffe re nce be twe e n thre e -month implie d volatility of out of the mone y puts and calls with strike s e qual distance from the spot price (+/- 10 pe rce nt). Highe r value s re fle ct gre ate r de mand for downside risk prote ction. Equitie s re pre se nts S&P500 inde x. Inte re st rate s re pre se nt we ighte d ave rage ske w of T re asury future s curve . Curre ncie s re pre se nt dollar ske w against major curre ncie s base d on JPMVXY inde x we ights. Z-score standardiz ation use s data going back to January 01, 2006. S ources: Bloomberg Finance L.P ., OFR analysis
60 Jul 2016
O ct 2016
Ja n 2017
Ap r 2017
Jul 2017
Note : VVIX Inde x me asure s the e xpe cte d volatility of the 30-day forward price of the CBOE VIX Inde x. S ource: Bloomberg Finance L.P .
5
Advanced Economies 2-year sovereign bond yields (percent ) U.S.
Germ any
U.K.
10-year sovereign bond yields (percent ) Japan
U.S.
2
3.0
1.5
2.5
1
2.0
0.5
1.5
0
1.0
-0.5
0.5
-1
0.0
-1.5
-0.5
Jul 2016
O ct
Jul
Ja n 2017
2016 2016 L.P . S ource: Bloomberg Finance
Ap r 2017
Jul 2017
Breakeven inflat ion (percent ) U.S. ten-year Japan ten-year
4
Germ any ten-year
Germ any
O ct 2016
U.K.
Japan
Ja n 2017
France
Ap r 2017
Jul 2017
S ource: Bloomberg Finance L.P .
10-year euro area periphery government bond spreads over German bunds (basis point s)
U.K. ten-year
400
Italian g o vt (Left Axis ) Spanis h g o vt (Left Axis ) Po rtug ues e g o vt (Left Axis ) Greek g o vt (Rig ht Axis )
2000
360 1600
3
320 280
2
1200
240 800
200 1
160 400 120 0
80
0 O ct 2016
Jul 2016
Ja n 2017
Ap r 2017
Jul 2017
S ource: Bloomberg Finance L.P .
110
O ct 2016
Ja n 2017
Ap r 2017
Jul 2017
S ource: Bloomberg Finance L.P .
Major currency indexes DXY (U.S. do llar) Japanes e yen
Jul 2016
euro Swis s franc
U.S. dollar long posit ioning vs. major currencies (net speculat ive posit ions, t housands of cont ract s)
Britis h po und
400
DXY (U.S. do llar) Japanes e yen
euro To tal
Britis h po und
300 100 200
100 90 0
-100
80 Jul 2016
O ct 2016
Ja n 2017
Ap r 2017
Jul 2017
Note s: Fore ign curre ncy incre ase s re pre se nt gre ate r stre ngth ve rsus the U.S. dollar. DXY incre ase s re pre se nt gre ate r stre ngth of the U.S. dollar ve rsus a baske t of major world curre ncie s. Inde x 100 = July 01, 2016. S ource: Bloomberg Finance L.P .
Jul 2016
O ct 2016
Ja n 2017
Ap r 2017
Jul 2017
Note s: Positive value s re pre se nt ne t U.S. dollar long positions. T he Dollar Inde x (DXY) is a future s contract base d on the U.S. dollar's value against a baske t of major world curre ncie s. T o e xpre ss a U.S. dollar long position in a non-U.S. dollar contract, the contract must be shorte d. S ource: Bloomberg Finance L.P .
6
Emerging Market s Emerging market currencies (U.S. dollars per foreign currency unit ) 120
Emerging market sovereign debt yield EM lo cal currency 5.5
Rus s ia China o ffs ho re Brazil EM currency Index
5.0 110
4.5 100
4.0
90 O ct 2016
Jul 2016
Ja n 2017
Ap r 2017
Jul 2017
Note s: Incre asing value s indicate stre ngthe ning ve rsus the U.S. dollar. Inde x 100=July 01, 2016. S ource: Bloomberg Finance L.P .
Equit y price indexes 130
3.5 O ct 2016
Jul 2016
Ja n 2017
Ap r 2017
Jul 2017
S ource: Bloomberg Finance L.P .
1-mont h realiz ed emerging market s volat ilit y (percent )
U.S. China Em erg ing m arkets
EM s o vereig n hard currency debt
Develo ped eco no m ies
EM currencies
20
120
110 10 100
90
0
80 O ct 2016
Jul 2016
Ja n 2017
Ap r 2017
Jul 2017
Note s: T he US e quity inde x is the S&P 500 Inde x. T he Chine se e quity inde x is the Shanghai Composite Inde x. T he De ve lope d Economie s inde x is the MSCI World Inde x and the Eme rging Marke ts inde x is the MSCI EM Inde x (both are in local te rms). Inde x 100 = July 01, 2016. S ource: Bloomberg L.P .
IIF port folio flows t o emerging market s ($ billion) Debt flo ws
2010
2011
2012
2013
2014
2015
2016
2017
Note s: Re aliz e d volatility is the annualiz e d standard de viation. Hard curre ncy sove re ign de bt base d on the J.P. Morgan Eme rging Bonds - Global Price Inde x and curre ncie s base d on a we ighte d ave rage of EM curre ncy re turns against the dollar using we ights from J.P. Morgan VXY-EM curre ncy volatility inde x. S ources: Bloomberg L.P ., OFR analysis
China's Foreign Exchange Reserves ($ t rillion)
Equity flo ws
$4.5
50
FX res erves
$4
40 30
$3.5
20
$3
10 $2.5
0
$2
-10 -20 -30
2010
$1.5 2011
2012
2013
2014
2015
2016
2017
Note s: Data re pre se nt the Institute of Inte rnational Finance 's monthly e stimate s of non-re side nt flows into thirty EM countrie s. Data for late st obse rvations are de rive d from IIF's e mpirical e stimate s using data from a smalle r subse t of countrie s, ne t issuance , and othe r financial marke t indicators. S ource: Bloomberg
$1 $0.5 Ja n Ja n Ja n Ja n Ja n Ja n Ja n Ja n Ja n Ja n Ja n Ja n 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017
S ource: Bloomberg
7
Commodit ies Major commodit ies prices
Crude oil
Blo o m berg co m m o dities index Crude o il fro nt m o nth (Brent)
125
$/Barrel Go ld fro nt m o nth
140
WTI (Left Axis ) Millio n Barrels Brent (Left Axis ) U.S. invento ries (Rig ht Axis )
550
500
100
110
450
75 80
400
50 50
350
25
20
0 2014
2015
2016
Note s: Inde x 100 = January 01, 2010 S ource: Bloomberg Finance L.P .
2015
2016
2017
Note : WT I and Bre nt are front-month contracts. S ource: Bloomberg Finance L.P .
Oil and nat ural gas fut ures curves Brent (Left Axis )
300
2014
2017
Oil supply and demand fact ors
Natural g as (Rig ht Axis )
$/barrel
120
$/m m btu
60
4.0 115
3.5
Millio n barrels per day Glo bal pro ductio n (Left Axis ) Glo bal co ns um ptio n (Left Axis ) U.S. rig co unt (Rig ht Axis )
1800 1600 1400
110
1200
55
105 1000 3.0
100 800 95
50
600
2.5 90
45
2.0 2M
6M
12M
24M
2014
2015
2016
2017
Note : Global production and consumption are e stimate s by the Inte rnational Ene rgy Age ncy. S ource: Bloomberg Finance L.P .
Speculat ive fut ures posit ioning (t housands of cont ract s) 350
200
85
60M
Note : Data as of July 05, 2017. S ources: Bloomberg Finance L.P ., OFR analysis
Brent (Left Axis ) Co pper (Left Axis )
400
Met als spot price indexes Co pper
Go ld (Left Axis ) Steel (Rig ht Axis )
Steel
Precio us m etals
200 50
280
40
210
30
140
20
70
10
150
100
0
0
50 -10
-70 -140
Jul 2015
Ja n 2016
Jul 2016
Ja n 2017
-20
Jul 2017
Note s: Positive value s re pre se nt ne t long positions. Ne gative value s re pre se nt ne t short positions. S ource: Bloomberg Finance L.P .
0 Jul 2015
Ja n 2016
Jul 2016
Ja n 2017
Jul 2017
Note : Inde x 100 = January 01, 2010. S ource: Bloomberg Finance L.P .
8