Retirement in the Shadow (Banking)

Retirement in the Shadow (Banking) Guillermo Ordoñez 1 Facundo Piguillem 2 1 University of Pennsylvania 2 EIEF October 6, 2015 1/37

Motivation Since 1980 the US has experienced fundamental changes: Large increase in life expectancy from 74 to 79 years Development of shadow banking cheaper financial intermediation We also observe a credit boom (from 1GDP to 1.7GDP), a large increase in financial assets (from 1.3GDP to 2.3GDP) and a decline in interest rates How quantitatively important were these two fundamental forces to explain the observed changes? Disclaimer: This paper is not about the crisis, but what preceded it 2/37

This paper 1) Cheaper borrowing? Account for the fall in interest rate spread. 2) Borrowing and Lending? OLG economy with: Uncertain life spans: saving for retirement Costly intermediation (insurance) or self insurance Heterogeneous bequest motive: make different decisions Main mechanism: in equilibrium, households with low bequest motive want insurance (e.g., buy annuities) with high bequest motive want to self insure (e.g., buy equities) High bequest motive hh s want to borrow (leverage) from low 3/37

Flow of funds: increase in pension funds and hh debt more DeNardi et al. (2015): Increase in life expectancy substantially inceases savings for retirement. Bequest motive and insurance: hard to identify each contribution Pensions/GDP Household's assets held in retiretment funds Table L100 flow of funds Line 20 Debt/GDP 1.1 1.8 1.7 1.0 1.6 0.9 1.5 0.8 0.7 0.6 Household Debt/GDP 1.4 1.3 1.2 1.1 1.0 0.5 0.9 0.4 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 0.8 1980 1982 1984 1986 1988 1990 1992 1994 year 1996 1998 2000 2002 2004 2006 2008 FF: Table L100, Line 20 FF: Table D.3 less governments 4/37

NIPA: Drop in financial sector liquidity cost Real interest rates Liquidity cost 12% 2.5% 10% Real deposit Real borrowing 2.0% Final-Level based Final-Quality adjusted R e a l 8% 6% 1.5% r e t u r n s 4% 2% 0% 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 1.0% 0.5% 0.0% -2% 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 5/37

Main Findings Spread between borrowing and lending fell 1 percentage point or, the intermediation cost fell by 33% Private borrowing and lending increased from 1GDP to 1.7GDP We account for these facts: Almost all the reduction in spread due to liquidity Increase in life expectancy key to get the right quantities and direction of prices Match the increase in financial assets Implication: Shadow B. prevented a fall in retiree s asset prices 6/37

Road Map 1. Accounting for the change in intermediation cost 2. Model economy 2.1 Individual decisions 2.2 Intermediation: traditional and shadow banking 2.3 Equilibrium 3. Solution and parameters 4. Results 5. Conclusions 7/37

Accounting for spreads The equation Gross output = Value added + Cost of inputs Int. received Bad debt = VA + Int. paid + other services fr e + (1 f )r L s b (1 + r e ) = ˆφ + r Where: r e : interest rate charged on a loan rl : return on liquid assets 1 f : proportion of assets invested in liquid assets sb : proportion of assets that are lost ˆφ: value added per unit of loan r: return to investor (or depositor) 8/37

Accounting for spreads Reorganize equation fr e + (1 f )r L s b (1 + r e ) = ˆφ + r r e r }{{} = ˆφ }{{} + s b (1 + r e ) }{{} + (1 f )(r e r L ) }{{} Interest spread value added risk component liquidity component We focus on the last component: (1 f )(r e r L ) If liquid asset have same return as loans r e = r L, no liquidity cost. If lenders do not need liquidity, f = 1, no liquidity cost. We clean for risk using Bad debt expenses (Table 7.1.6 Line 12) 9/37

Accounting for spreads: NIPA tables more r e =(Total interest received- bad debt expenses)/hh s debt. (Table 7.11 Line 28-Table 7.1.6 Line 12)/Table D.3. r =(Total interest paid+serv. furnished without payment)/hh s debt. (Table 7.11 Line 4+Table 2.4.5 line 88)/Table D.3. Then r e r }{{} = ˆφ }{{} + (1 f )(r e r L ) }{{} Risk adjusted spread value added liquidity component 5.0% 4.0% 3.0% 2.0% risk adjusted spread 1.0% 0.0% 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 10/37

Accounting for spreads: efficiency? Why did the spread fall? Was it ˆφ, or liquidity (1 f )(r e r L )? 11/37

Accounting for spreads: efficiency? Why did the spread fall? Was it ˆφ, or liquidity (1 f )(r e r L )? Philippon (2015) shows ˆφ has been constant for 100 years and technology is CRS. 11/37

Accounting for spreads: efficiency? Why did the spread fall? Was it ˆφ, or liquidity (1 f )(r e r L )? Philippon (2015) shows ˆφ has been constant for 100 years and technology is CRS. 3.0% 2.5% 2.0% 1.5% 1.0% Level based Quality Adjusted 0.5% 0.0% 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 11/37

Accounting for spreads: the matter in issue more What is left after subtracting value added? 2.5% Final-Level based 2.0% Final-Quality adjusted 1.5% 1.0% 0.5% 0.0% 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 12/37

What does Shadow Banking do? The improvement arises from shadow banking creating safe assets without liquidity back ups. How? Securitization: Pooling and tranching allowed to improve liquidity (Farhi and Tirole (2014). direct increase in f Repo: Use of securities as information insensitive collateral (Gorton and Ordonez (2014)) increase in r L Reputation concerns of large institutions allowed to bypass inefficient liquidity provisions (Ordonez (2014)) Improvement: creation of instruments to provide safety 13/37

Liquid assets: some evidence Liquidity cost = (share TB) (1 f )(r e r L ) + (share SB) 0 TB: Less liquid assets? Less TB: Flow of Funds Table L109 0.3 P r 0.25 o p o r 0.2 t i o n 0.15 Cash plus deposits in FED Cash+plus deposits in Fed plus Treasuries Liquidity: including ABS 0.7 0.6 0.5 Deposits over private debt. Flow of funds: Table L109/Table D.3 o f d e p o s i t s 0.1 0.05 0 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 0.4 0.3 0.2 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 14/37

Model We build on Mehra et al. (2011) adding social security with population growth OLG with financial intermediaries Neoclassical technology, productivity growth Government collects taxes, pay transfers and issues debt Decrease in intermediation cost plus the increase in life expectancy generates the observed changes in aggregate quantities and prices 15/37

Model: Technology and Population Overlapping generations with Neoclassical technology Y t = K θ t (z t L t ) 1 θ z t+1 = (1 + γ)z t θ: share of capital in output γ: growth rate of labor productivity Measure (1 + η) t of people are born every period Each one with a different intensity of bequest motive α 0 α f (α) η: population growth rate 16/37

Agents Agents order consumption according to T U(α, c, b) = β j log c j + β j (1 δ) j T 1 [(1 δ) log c j +δα log b j ] j=0 j=t +1 α 0: heterogeneous strength of bequest motive 0 < δ < 1: probability of death Agents work for T years, earn wage ω j and then retire Bequest b j, equally distribute amongst agents of age j = T I < T At age j = T I < T receive inheritance b 17/37

Investment Decision more Assumption: Restricted choice for households Agents choose whether to sign or not annuity contract at age j = 0 18/37

Investment Decision more Assumption: Restricted choice for households Agents choose whether to sign or not annuity contract at age j = 0 Then agents have two options: Strategy A buy annuities Benefit: Insurance against risk of living long time Cost: Low return on assets (r) Strategy B go on your own Benefit: High return on assets (re ) Cost: No insurance 18/37

Investment Decision more Assumption: Restricted choice for households Agents choose whether to sign or not annuity contract at age j = 0 An annuity contract between an agent and the intermediary is: While working: payment made by agent to intermediary When retired: payment made by intermediary to agent cj if alive b j when she dies 18/37

Investment Decision more Assumption: Restricted choice for households Agents choose whether to sign or not annuity contract at age j = 0 Interest spread: φ = r e r Propositions 1 and 2 If φ φ φ, there exists α > 0 such that, if α < α follow strategy A if α α follow strategy B From now on assume: µ a if α = 0 F (α) = 1 µ a if α = ˆα > 0 0 otherwise 18/37

Agent Problem: Strategy A (annuity) max T β t log c j + j=0 subject to T c t (1 + r) j + j=0 j=t +1 j=t +1 β j (1 δ) j T 1 [(1 δ) log c j + δα log b j ] (1 δ) j T 1 [(1 δ)c j + δb j ] (1 + r) j v A 0 v A 0 = T 1 j=0 (1 τ)ω j (1 + r) j + b (1 + r) T I + (1 + r)tra A ω T (r + δ)(1 + r) T r: household lending rate Tra A : SS transfer v A 0 : wealth at time 0 ω j : wage at age j 19/37

Solution annuity strategy: consumption (α low) 0.045 0.040 0.035 Type A Working age Retirement age 0.030 Consumption 0.025 0.020 0.015 0.010 0.005 0.000 22 26 30 34 38 42 46 50 54 58 62 66 70 74 78 82 Age 20/37

Solution annuity strategy: assets accumulation (α low) 20.0 16.0 Net Worth/Annual wage 12.0 8.0 Type A 4.0 0.0 22 26 30 34 38 42 46 50 54 58 62 66 70 74 78 82 Age 21/37

Agent Problem: Strategy B (no-annuity) Two problems: before retirement and after 22/37

Agent Problem: Strategy B (no-annuity) Two problems: before retirement and after After retirement V (w) = max{log c + (1 δ)βv (w ) + δβα log w } subject to c + w (1 + r e ) w r e : return on equity and the household borrowing rate 22/37

Agent Problem: Strategy B (no-annuity) Solution after retirement V (w) = ν 1 (α) + ν 2 (α) log w where ν 2 (α) = 1 + αβδ 1 (1 δ)β 23/37

Agent Problem: Strategy B (no-annuity) Solution after retirement V (w) = ν 1 (α) + ν 2 (α) log w where ν 2 (α) = 1 + αβδ 1 (1 δ)β The optimal consumption and implicit bequest strategies are: c = w/ ν 2 (α) w = (1 + r e )(w c) Bequests are: b j = w j j T 23/37

Agent Problem: Strategy B (no-annuity) At entry in labor force subject to β j logc j + β T V (w T ) T 1 max T 1 j=0 j=0 c j (1 + r e ) j + w T (1 + r e ) T v B 0 v B 0 = T 1 j=0 (1 τ)ω j (1 + r e ) j + b (1 + r e ) T I 24/37

Solution No-annuity strategy: consumption (α high) 0.045 0.040 0.035 Type B Working age Retirement age 0.030 Consumption 0.025 0.020 0.015 0.010 0.005 0.000 22 26 30 34 38 42 46 50 54 58 62 66 70 74 78 82 Age 25/37

Solution No-annuity strategy: assets accumulation (α high) 20.0 16.0 Type B Net Worth/Annual wage 12.0 8.0 4.0 0.0 22 26 30 34 38 42 46 50 54 58 62 66 70 74 78 82 Age 26/37

Lifetime pattern of consumption 0.045 0.040 Type A Working age Retirement age 0.035 Type B Consumption 0.030 0.025 0.020 0.015 0.010 0.005 0.000 22 26 30 34 38 42 46 50 54 58 62 66 70 74 78 82 Age 27/37

Closing the economy Fiscal policy to keep D G t fixed and tax τ on labor income τω t L t Tra t gy = rd G t (D G t+1 D G t ) g: government spending as propotion of GDP. D G : government debt Tra: total transfers to households after retirement key for insurance and getting the right level of debt 28/37

Market Clearing: Feasibility Y = C(ˆα, τ, b) + gy + X + ˆφ(K W B (ˆα, τ, b) 1 + r e ) Assets market W A (ˆα, τ, b) 1 + r + W B (ˆα, τ, b) 1 + r e = D G + K Bequest=inheritance b = (1 + γ) T I B(ˆα, τ, b) 29/37

(stationary) Equilibrium: An equilibrium is {c i, w i, b i, K, α }, prices r e, r and fiscal policy such that: 1. Households maximizes 2. Intermediary maximize. Because CRS fr e + (1 f )r L = ˆφ + r 3. Government budget constraint holds 4. Market clearing 30/37

Calibration: individual parameters Calibrate model economy to 1980 counterfactual to 2007. φ = 0.03 (Spread from NIPA discussed before) 31/37

Calibration: individual parameters Calibrate model economy to 1980 counterfactual to 2007. φ = 0.03 (Spread from NIPA discussed before) Parameters associated with individuals β = 0.99 δ = 0.08 post retirement life expectancy of 12.5 years ˆα = 7.5. To match borrowing and lending. 31/37

Calibration: related to aggregation Aggregate parameters: µ A = 0.718. Measure of annuity type γ = 0.02. Consistent with observations in labor productivity η = 0.01. Population growth rate in 1980 (0.007 in 2007) θ = 0.3. Share of capital in output δ k = 0.028. Consistent with capital output ratio = 3.4 Remark: µ A and α joinly calibrated to get debt to GDP ratio. 32/37

Calibration: Government debt. Use only domestic holdings D G /Y = 0.3. Includes states, taking out foreign held treasuries Choose TraA to get the number In 1980 total was 0.35, difference is foreign held In 2007 total was 0.62, taking out foreign held remains at 0.3 0.70 Figure 2: Holders of U.S. Treasury Securities (percent of total oustanding) 0.60 0.50 0.40 0.30 0.20 0.10 0.00-0.10 1945 1947 1949 1951 1953 1955 1957 1959 1961 1963 1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 US Depository Institutions Rest of the World Insurance Companies Mutual Funds Securities Broker-Dealers 33/37

Summary of Aggregate Results 1980 Economy Model Data µ A. Indirect equity holding 0.72 0.72 µ B. Direct equity holding 0.31 0.28 ˆα 7.5 DiNardi et al. SS transfer 0.34 0.27 (per benef.) National accounts Output 1 1 Consumption 0.58 0.62 Governmet spending 0.19 0.19 Imposed Net worth Total 3.70 3.62 r 0.03 0.0291 r e 0.06 0.0596 Government Debt/GDP 0.30 0.30 Target Bequest/GDP 0.048 0.027 Households Debt/GDP 1.00 1.00 Target Equities directly hold/total financial assets. Table B.100e of flow of funds 34/37

Counterfactual: the 2007 economy more Economy Benchmark (1980) New δ New φ 2007 Interm. Cost (φ) 3% 3% 2% 2% Survival prob. (δ) 0.08 0.06 0.08 0.06 r 0.030 0.022 0.035 0.028 r e 0.060 0.053 0.055 0.048 National accounts Capital/GDP ratio 3.40 3.70 3.60 3.91 Output 1.00 1.034 1.025 1.062 Net worth Type A (deposits+pensions) 1.30 1.40 1.78 1.96 Type B (equity) 2.40 2.57 2.12 2.25 Total financial assets (W A + D G ) 1.60 1.70 2.08 2.26 Total assets (Table L100) 1.36 2.33 Government debt/y 0.30 0.30 0.30 0.30 Households debt/gdp 1.00 1.10 1.48 1.66 Data on debt 1.00 1.71 35/37

Counterfactual: alternative government debt Economy 1980 2007 Free All D G Benchmark Calibration D G Domestic Interm. Cost (φ) 3% 2% 2% 2% Survival prob. (δ) 0.08 0.06 0.06 0.06 r 0.030 0.028 0.027 0.03 r e 0.060 0.048 0.047 0.05 National accounts Capital/GDP ratio 3.40 3.91 4.00 3.83 Output 1.00 1.06 1.07 1.05 Net worth Type A (deposits+pensions) 1.30 1.96 1.80 2.11 Type B (equity) 2.40 2.57 2.12 2.25 Total financial assets (W A + D G ) 1.60 2.26 1.76 2.73 Total assets (Table L100) 1.36 2.33 Government debt/y 0.30 0.30-0.04 0.62 Households debt/gdp 1.00 1.66 1.84 1.59 Data on debt 1.00 1.71 36/37

Discussion and conclusion What do we get from the table? The drop in liquidity cost generates a sizeable increase in debt, not enough, but wrong movement in returns Increase in life expectancy gives the additional debt and right movement in r Both together deliver the right level of debt, similar increase in financial assets and righ movement in r As a result: The domestic saving glut seems to be enough. People are living longer, they need to save for retirement. Shadow banking allows it by avoiding the fall in the returns 37/37

Agents budget constrains back res. back dec Type A: T j=0 c t (1 + r) j + j=t +1 (1 δ) j T 1 [(1 δ)c j + δb j ] (1 + r) j v A 0 v A 0 = T 1 j=0 (1 τ)ω j (1 + r) j + b (1 + r) T I Initial wealth Type B T j=0 c t (1 + r) j + W T (1 + r) T v B 0 v B 0 = T 1 j=0 (1 τ)ω j (1 + r e ) j + b (1 + r e ) T I Initial wealth 38/37

Treasury holders back Rest of the world bought most of the increase in public debt. 0.70 Figure 2: Holders of U.S. Treasury Securities (percent of total oustanding) 0.60 0.50 0.40 0.30 0.20 0.10 0.00-0.10 1945 1947 1949 1951 1953 1955 1957 1959 1961 1963 1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 US Depository Institutions Rest of the World Insurance Companies Mutual Funds Securities Broker-Dealers 39/37

Figure 10: Evolution of Net Interest Margin, Markups and Lerner Index Net interest income: Corbae and D Erasmo (2013) Data corresponds to comercial banks in the US. Source: FDIC, Call and Thrift Financial Reports. 6 Panel (i): Net Interest Margin Perc. (%) 5 4 3 2 1985 1990 1995 2000 2005 2010 year back calibration Panel (iii): Lerner back Indexintro 40/37

Household s Portfolio back This picture is also very interesting. Note that how the proportion of assets held in pensions accounts grows all the time. For some reason starting the 90's households drastically reduce their holdings of deposits, why? Is it shadow banking? (check Roth IRA and 401k regulation) Less deposits, more retirement accounts 0.35 0.30 Proportion of total assets held in deposits and pension accounts Pensions over assets Deposits over assets 0.25 0.20 0.15 0.10 1952 1955 1958 1961 1964 1967 1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 2000 2003 2006 2009 2012 year Here is the other two important components of assets, which are corporate and non-corporate assets. 41/37

Household s total financial assets back 2.4 Total Financial Assets: L100 minus corporate and non corporate equity 2.2 2 1.8 1.6 1.4 1.2 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 42/37