Asset Allocation in a distorted environment ANDREA DELITALA MARIA LUISA MAGLI November 2016 Università Commerciale L.Bocconi - Milan
CONTENTS 1 Optimal Investment Theory slide 3 2 Exceptional circumstances 1 slide 20 3 Exceptional circumstances - 2 slide 38 4 Challenging Market Conditions: How to adapt slide 53 5. Macro and Market Scenario Analysis slide 74 2
1 Optimal Investment Theory 3
UNCERTAINTY ABOUT ASSETS TOTAL RETURN HIGHER IN THE SHORT RUN Nearly every year the performance ranking of financial investments differs from that of the previous period. Source: 4
BASIC INGREDIENTS OF AN ASSET ALLOCATOR: RISKY STUFF! Financial assets differ in structure, and their negotiated price vary accordingly 15% 14% 13% 12% 11% 10% probability Standard Distribution Performance Distribution Most often in Finance Normal behavior is assumed to be a reasonable approximation 9% 8% 7% 6% Average Stdev 3,10 3,75 5% 4% 3% Max loss with 5% prob. = Var (95%) = -3,07 2% 1% 0% -10-9 -8-7 -6-5 -4-3 -2-1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Expected Total Return (annual %) Source: 5
CORRELATION BETWEEN ASSET CLASSES IS CRUCIAL FOR AA The investible Universe ought to be classified by Risk classes (moving targets) 5,00% Volatility tolerance threshold 100% Equities 4,75% 4,50% Negative Correlations help to compensate different asset risk 4,25% Expected Total Return 4,00% 3,75% 3,50% Corr=-1 Corr=-0.5 Corr=0 Diversification benefit Corr=1 25% Equities; 75% Bonds Low Correlations help to contain overall PTF volatility 3,25% Budget of volatility 5% 3,00% 100% Bonds 2,75% 2,50% 0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 10% Risk of PTF (ST.Dev. of Total Returns) 6
THE EFFICIENT FRONTIER & TANGENT (THROUGH CASH) Best combination of Expected Total Return (TR) and Volatility (Risk) of Portfolios (PTF) 7% 6% PTF Expected Return Line (dotted): Optimal PTF+ Lever All Equities 10%; 6% 5% With two risky assets + Cash 4% 3% 3,3%; 3,9% Min Risk All Bonds 5%; 3% PTF Exp R StDev Cash 2,00% 0,00% Bonds 3,00% 5,00% Equities 6,00% 10,00% Eq/Bond -0,50 2% Line: Optimal PTF+ Cash PTF Risk 1% 0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 10% Source: 7
WHY DO I CHOOSE TO STAY ON THE FRONTIER? Hierarchy of Portfolios 5,0% 4,5% PTF Expected Return Unfeasible but optimal portfolios 4,0% 3,5%; 4,2% 3,5% 3,0% 2,5% Cash Bonds Equities Eq/Bond PTF Return St Dev PTF Sharpe PTF o 10% 80% 10% 12,5% 3,20% 3,61% 0,33 Initial PTF 3,6%; 3,2% Feasible but sub-optimal portfolios PTF Risk 2,0% 1,0% 1,5% 2,0% 2,5% 3,0% 3,5% 4,0% 4,5% 5,0% Source: 8
WHY DO I CHOOSE TO STAY ON THE FRONTIER? Hierarchy of Portfolios 15% 14% 13% 12% 11% 10% 9% 8% 7% probability Expected Return Distribution of optimal PTF Cash Bonds Equities Eq/Bond PTF Return St Dev PTF Sharpe PTF o PTF 1 10% 15,1% 80% 52,2% 10% 32,7% 12,5% 62,6% 3,20% 3,83% 3,61% 3,00% 0,33 0,61 6% 5% 4% 3% 2% Max Loss with 5% prob.=var(95%) Expected Return Expected Return Distribution of initial PTF 1% 0% -2,7-1,1 3,2 3,8-10 -9-8 -7-6 -5-4 -3-2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Expected Total Return (annual %) Source: 9
NORMAL CORRELATION REGIME (SHORT RUN) EQUITY - BONDS Equities Bonds Growth, Inflation prospects improve Growth, Inflation prospects deteriorate Sales, Earnings rise, Price rise Sales, Earnings fall, Price fall Yields rise, Price fall Yields fall, Price rise Negative Correlation Equities-Bonds Hyperinflation prospects Real Sales, Earnings fall Yields rise, Price fall Positive Correlation Equities-Bonds 10
NEGATIVE CORRELATION (SHORT RUN) EQUITY - BONDS: WHY? Equities Bonds Earnings Yield Bond Yield Earnings (E) Price (Pe) Yield (i) Price (Pb) Ceteris paribus Central Banks are expected to raise interest rates in a stronger economy Negative correlation holds if risk premia are not distorted by other factors 11
2 Exceptional circumstances - 1 12
RISK RETURNS OF MAIN FINANCIAL ASSETS IN 2007 Most Yield to Maturity (YTM) above 4%; Earning Yields around 7% Source:, Datastream 13
RISK RETURNS OF MAIN FINANCIAL ASSETS TODAY Yields compressed by CBs -> Financial Repression. Very low volatility -> Investors coercion 9% 7% $ HY Local Currency EM MSCI DebtWorld Germany (DAX) EM (MSCI EM) Europe (Euro Stoxx) Japan (TOPIX) France (CAC) UK (FTSE100) Italy (MIB30) USA (S&P 500) Switzerland (SMI) 5% $ EM Debt HY 3% PTF 70/30 BMK (2,01%; 5,43%) Euro I/L BTP 10Y 1% EMU Corporate EuroMTS Global World Bonds EuroMTS 3-5Y EURIBOR 3M Bund 10Y CH Bonds 10Y EuroMTS 1-3Y -1% 0,0% 2,5% 5,0% 7,5% 10,0% 12,5% 15,0% 17,5% 20,0% 22,5% 25,0% 27,5% 30,0% Source:, Datastream 14
US INTEREST RATES ACCORDING TO FUNDAMENTALS Equilibrium: Taylor Rule and official rates in USA QE1 QE2 QE3 Source:, Bloomberg 15
CENTRAL BANK ASSETS MAJOR 5 The ECB is the only CB of the G3 that is still providing liquidity: likely to taper in 2017 Evolution of the Balance sheets size: FED, ECB, BOJ, PBoC, BoE Policy Liquidity, Major 5 Source:, Datastream 16
BUND E TREASURIES: DISTORTION DUE TO FLIGHT TO QUALITY AND QE 10-year T-Note and Bund compared to the estimated fair value Estimates based on key variables and market indices, foreign exchange, commodities, 2Y swap yields YTM above fair value Tapering Starts Bund New Conundrum? T-Note -233bp YTM below fair value -314bp Source:, Bloomberg 17
THE DETERMINANTS OF THE YIELD TO MATURITY Factors affecting the various components of bond yields Macro Output Gap Inflation/Unemployment Productivity Guidance (and perceived errors) MONETARY INTEREST RATES (forward) Market LSAP (Large Scale Asset Purchases) Correlation Equity/Bond Risk Aversion - Implied Volatility Flows Equity/Bond TERM PREMIUM YIELD TO MATURITY 18
DOES QE REDUCE YIELDS? Impact on Nominal Yields not obvious, clearer on real yields QE Guidance MONETARY INTEREST RATES (forward) TERM PREMIUM YIELD TO MATURITY? Real Yield drop 19
THE QE HAS EFFECTS PARTICULARLY ON REAL RATES AND OTHER ACTIVITIES... QE lowers real returns. Its unwinding since May 2013 reversed the trend (with BE dropping) Nominal interest rates, real (TIPS) and Break-even inflation in the market for U.S. Treasuries EXIT Source:, Bloomberg 20
MKT BASED TERM PREMIUM STABLE, WHILE EXPECTATIONS ON FF VOLATILE! Inferring the term premium from the market, it is positive unless we consider also the CDS Term premium on T-Notes: yield to maturity - CDS - average interest rate on Fed funds for the next 10 years 5,0 4,5 4,0 Fed Fund rate Tgt FF avg [OIS 10Y] Zero Curve 10Y TP = (Zero - CDS) 10Y - FF avg [OIS 10Y] 3,5 3,0 2,5 2,0 1,5 1,0 0,5 0,0 T-Note: 2,35% Expected FF: 1,65% (Estimated w FF Futures etc.) CDS 10 on T-Note: 40bp Term Premium: 30bp (by difference) -0,5 Jan-08 Jan-09 Jan-10 Jan-11 Jan-12 Jan-13 Jan-14 Jan-15 Jan-16 Source:, Bloomberg 21
AB-NORMAL CORRELATION REGIME (SHORT RUN) EQUITY - BONDS Equities Bonds Growth, Inflation prospects improve Sales, Earnings rise, Price rise Yields rise, Price fall Growth, Inflation prospects deteriorate Sales, Earnings fall, Price fall Yields fall, Price rise Negative Correlation Equities-Bonds Quantitative Easing: Growth, Inflation prospects improve Sales, Earnings rise, Price rise Real Yields fall, Price rise Positive Correlation Equities-Bonds 22
CORRELATION ANALYSIS If correlation falls (better if <0) when volatility rises, then the diversification benefit intensifies Volatility and Correlation (weekly; 12M rolling); US Treasuries, MSCI W Tapering of QE Avg: -0,33 QE1 QE2 QE3 Source:, Bloomberg 23
3 Exceptional circumstances - 2 24
THE UNCOMFORTABLE EMU APARTMENT BLOCK Congenital deficiencies? Mostly market circularity and many mistakes (PSI, EBA...) In the EMU house the Greek fire started 3 years ago. The lack of windows (floating currencies) makes suffocation more likely, but does not explain why the fire spread so rapidly Economists (e.g. De Grauwe) pointed out poor Governance. Yet I believe market forces and poor decision making have been underestimated 25
DISTRUST WITH : GAME CHANGER FOR SOVEREIGN SOLVENCY Asset swaps used to measure the relative Credit Premia: Italian case 10Y Asset Swap spread (BTP-Bund): Credit premium of BTP vs Bund = (YTM BTP 10Y ITL Swap 10Y) (YTM Bund DEM Swap 10Y) 10Y BTP-Bund 550bp 10Y Swap spread (ITL-DEM) = risk free rates in respective Currencies embed the devaluation risk of the ITL vs DEM (covered parity) Sovereign bond spreads back to pre-emu era! >400bp Asset Swap difference <200bp Source:, Bloomberg 26
BTP-BUND DEFINITELY INCLUDES EMU REVERSIBILITY RISK 10Y BTP-Bund Spread = Credit Premium + Expected Realignment x Probability of Fluctuation YTM 10Y BTP YTM Bund 2,04% 0,34% Spread 10Y BTP vs Bund Credit Spread Fair Value Fair Value under ECB QE1 180bp 100bp Source:, Bloomberg 27
THE PROBABILITY OF FLUCTUATION 10Y BTP-Bund Spread = Credit Premium + Expected Realignment x Probability of Fluctuation 27 Jan 14 222 bp = 180 bp + 30% x 1,5% 36% Assumptions : Expected depreciation of new ITL/DEM in case of currency exodus: 30% Credit premium: 120bp Based on 2Y Yield to maturity (BTP, Schatz, swaps) OMT-> 24% Credit risk component in BTP-Bund: 180bp QE (+OMT)-> 2% Source:, Bloomberg 28
AB-NORMAL CORRELATION REGIME (SHORT RUN) EQUITY - BONDS Equities Bonds Growth, Inflation prospects improve Growth, Inflation prospects deteriorate Sales, Earnings rise, Price rise Sales, Earnings fall, Price fall Yields rise, Price fall Yields fall, Price rise Negative Correlation Equities-Bonds EMU Crisis worsens Sales, Earnings rise, Price fall Peripheral Yields rise, Price fall Positive Correlation Equities-Bonds 29
CORRELATIONS BETWEEN BONDS AND EQUITIES, USA & ITALY (JAN 13) The core bonds (USA) negatively correlated with equities. This DOES NOT apply to Italy! Correlation between US Treasuries and MSCI W ($) Correlation between Italian BTP and MSCI EMU Average EMU crisis QE1 Average QE2 QE3 Source:, Bloomberg Source:, Bloomberg 30
CORRELATIONS BETWEEN BONDS AND EQUITIES (GERMANY, ITALY) Correlation between German Bunds and MSCI EMU Correlation between Italian BTP and MSCI EMU Average Average Source:, Bloomberg Source:, Bloomberg 31
4 Challenging Market Conditions: How to adapt 32