Using Stock Returns to Identify Shocks to Government Spending

Transcription

1 Using Stock Returns to Identify Shocks to Government Spending Jonas D.M. Fisher Ryan Peters Chicago Fed Atlanta Fed Fiscal Policy Conference, January 8-9, 21

2 Motivation What is the G multiplier? Use response to G shocks to test models Two approaches lead to different answers VAR innovations to G, Blanchard and Perotti(22): G Y, H, C, W /P Historical narrative, Ramey and Shapiro (1998): G Y, H C, W /P

3 Motivation Drawbacks to both approaches VAR innovations: Ramey (28) critique Narrative strategy: few observations, all increases, assume certainty, inherently subjective Seek alternative immune to Ramey critique, with more observations, increases and decreases, incorporating uncertainty, less subjective Uhlig and Mountford (25): sign restriction approach satisfies these criteria

4 Alternative Approach Low frequency swings in military spending: Korean War, Vietnam War, Carter-Reagan Buildup, End of Cold War, 9/11 [spending] Anticipated military build ups should be reflected in stock returns of military contractors Explore using innovations in excess returns of military contractors to identify spending shocks Key findings: Multiplier = 1.5 G Y, H, C W /P then

5 Outline Model Describe data Identification strategy Findings

6 Model: Preferences and Technology max β t (ln c t n ct n mt ) c t + x ct /q t + x mt /q t + g t A t kt α nct 1 α m t + m A t q t dt α n 1 α mt k t+1 = (1 δ)k t + g(x ct, x ct 1 ) d t+1 = (1 δ)d t + h(x mt, x mt 1 )

7 Model: Responses to Military Shock

8 Model: Responses to G Shock

9 Model: Responses to Neutral Technology Shock

10 Model: Responses to Investment Technology Shock

11 Responses Estimated from Model Generated Data Percent Military Spending Percent Non-military Spending Percent Output Percent Accumulated Excess Returns

12 Data Top 3 Primary Munitions Contractors Guns+: Include by SIC/NAICS code Ordnance and Accessories Ship & Boat Building & Repairing Guided Missiles & Space Vehicles & Parts Aircraft & Parts & Auxiliary Equipment, NEC Tanks and Tank Components Search, Detection, Navigation, Guidance, Aeronautical Systems Guns: Guns+ less last two (Fama-French/Lustig, et al) War Dates (Ramey-Shapiro/Eichenbaum-Fisher) War Dates+ (Ramey)

13 Top 3 US Military Contractors Primary Contractor SIC Industry Years in Top 3 General Dynamics Ship and Boat Building, , , Repairing , Grumman Aircraft Lockheed Aircraft, Guided Missiles, , , Lockheed ( ), Space Vehicle, Parts Lockheed Martin (1995 -) Martin, Guided Missiles, , 1993 Martin Marietta ( ) Space Vehicle, Parts McDonnell Aircraft, Aircraft , 1966, McDonnell Douglas ( ) , Northrop, Search, Detection, 1975, 1992, Northrop Grumman (1994 -) Navigation, Guidance, Aeronautical Systems Raytheon Search, Detection, , 1996, Navigation, Guidance Aeronautical Systems United Aircraft, 1958, , United Technologies (1969 -) Aircraft, Parts

14 Top 3 Share of Primary Contracts Share of Primary Contracts Number of Firms

15 Detrended Sales and Military Spending Spending Spending Spending Military Spending Top 3 Sales Military Spending Guns Sales Military Spending Guns+ Sales Sales Sales Sales

16 Accumulated Excess Returns and Military Spending Military Spending, $ Excess Returns Military Spending Accumulated Excess Returns

17 Excess Returns and Ramey (28) s Shocks

18 Explanatory Power of Spending Indicators Variable Top 3 Guns Guns+ War Dates War Dates+ 1948:1 27:4 Military Spending Government Spending Output Consumption :3 27:4 Military Spending Government Spending Output Consumption Note: The reported R 2 s are based on regressions of the predicted variable on the current value and six lags of the indicator variable. Variables are log first differences

19 Identification Strategy S = accumulated excess returns VAR: MilG, Y, R, X, S, X = G, C, H, W /P G-shock: Choleski innovation in S, S ordered last If represent economy well, good estimate of G- shock when Tech change in Top 3 as in whole economy Markup dynamics as in whole economy Variation in sales dominated by shocks to defense spending

20 Excluding Korean War 1959:1-26:4, trend, six lags Returns poor forecast of Korean War build up Excess Profits Tax, W & P controls Eisenhower Farewell Address: Our military organization today bears little relation to that known by any of my predecessors in peacetime, or indeed by the fighting men of World War II or Korea. Until the latest of our world conflicts, the United States had no armaments industry. American makers of plowshares could, with time and as required, make swords as well. But now we can no longer risk emergency improvisation of national defense; we have been compelled to create a permanent armaments industry of vast proportions.

21 Effects of Top 3 Excess Return Shocks Percent Accumulated Top 3 Excess Returns Percent Military Spending 3.5 Accumulated Top 3 Net Income 1.2 Total Government Spending 2$ Billions Percent

22 Effects of Top 3 Excess Return Shocks Output Hours Private Consumption Product Wage [Other Variables]

23 Government Spending Multiplier 2 quarter accumulated G response:.1 2 quarter accumulated Y response:.3 G/Y =.2 dy dg =(.3/.1) (1/.2) =1.5 Is this all G?: dy dg + dy dp dp dg

24 Responses of Policy Variables.2 Tax Revenue as Share of GDP Percent Nominal 3 Month Interest Rate Basis Points

25 Historical Effects of Shocks Total Government Spending Output

26 Effects of War Dates Output Hours Private Consumption Product Wages

27 Effects of Using Spending Innovations Output Hours.5 Private Consumption.4 Product Wages

28 Conclusion Anticipated military build ups should be reflected in stock returns of military contractors Explore using innovations in excess returns of Top 3 primary military contractors to identify spending shocks Multiplier = 1.5 G Y, H, C W /P then

29 US Military Spending, Billions $

30 Model: Shocks a t g t q t m t s t s t 1 s t 2 = ρ a ρ g ρ q b 1 b 2 b a t 1 g t 1 q t 1 m t 1 s t 1 s t 2 s t 3 + ε t η t ν t µ t

31 Effects of Excess Return Shocks Percent Consumer Durables Percent Non Res. Investment Percent Res. Investment Percent Consumption Price