Skip to content

Commit ac29181

Browse files
Add a lecture by @shizejin on Two Computations to Fund Social Security and Edits from Tom on BHS lecture (#810)
* updates * Tom's Feb 18 edits of two-computations lecture * Tom's Feb 19 edits of the bhs lecture, a little after noon * updates * Tom's second Feb 19 edits of bhs lecture * updates * update GPU note --------- Co-authored-by: thomassargent30 <ts43@nyu.edu>
1 parent f32e209 commit ac29181

File tree

6 files changed

+3090
-11
lines changed

6 files changed

+3090
-11
lines changed

lectures/_static/quant-econ.bib

Lines changed: 71 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -2033,7 +2033,7 @@ @book{Ljungqvist2012
20332033

20342034
@article{Lucas1978,
20352035
author = {Lucas, Jr., Robert E},
2036-
journal = {Econometrica: Journal of the Econometric Society},
2036+
journal = {Econometrica},
20372037
number = {6},
20382038
pages = {1429--1445},
20392039
title = {{Asset prices in an exchange economy}},
@@ -2441,6 +2441,45 @@ @article{Townsend1983
24412441
pages = {546-588}
24422442
}
24432443

2444+
@article{tobin1992old,
2445+
title={An old Keynesian counterattacks},
2446+
author={Tobin, James},
2447+
journal={Eastern Economic Journal},
2448+
volume={18},
2449+
number={4},
2450+
pages={387--400},
2451+
year={1992},
2452+
publisher={JSTOR}
2453+
}
2454+
2455+
@article{hicks1937mr,
2456+
title={Mr. Keynes and the" classics"; a suggested interpretation},
2457+
author={Hicks, John R},
2458+
journal={Econometrica},
2459+
pages={147--159},
2460+
year={1937}
2461+
}
2462+
2463+
@article{hansen1983stochastic,
2464+
title={Stochastic consumption, risk aversion, and the temporal behavior of asset returns},
2465+
author={Hansen, Lars Peter and Singleton, Kenneth J},
2466+
journal={Journal of political economy},
2467+
volume={91},
2468+
number={2},
2469+
pages={249--265},
2470+
year={1983},
2471+
publisher={The University of Chicago Press}
2472+
}
2473+
2474+
@article{hansen1982generalized,
2475+
title={Generalized instrumental variables estimation of nonlinear rational expectations models},
2476+
author={Hansen, Lars Peter and Singleton, Kenneth J},
2477+
journal={Econometrica: Journal of the Econometric Society},
2478+
pages={1269--1286},
2479+
year={1982},
2480+
publisher={JSTOR}
2481+
}
2482+
24442483
@incollection{Uhlig2001,
24452484
author = {Uhlig, H},
24462485
booktitle = {Computational Methods for the Study of Dynamic Economies},
@@ -2769,6 +2808,37 @@ @article{diamond1965national
27692808
publisher = {JSTOR}
27702809
}
27712810

2811+
@article{huang1997two,
2812+
title = {Two computations to fund social security},
2813+
author = {Huang, He and Imrohoroglu, Selahattin and Sargent, Thomas J},
2814+
journal = {Macroeconomic Dynamics},
2815+
volume = {1},
2816+
number = {1},
2817+
pages = {7--44},
2818+
year = {1997},
2819+
publisher = {Cambridge University Press}
2820+
}
2821+
2822+
@techreport{faber1982life,
2823+
title = {Life Tables for the {United States}: 1900--2050},
2824+
author = {Faber, Joseph F},
2825+
year = {1982},
2826+
number = {Actuarial Study No. 87},
2827+
institution = {Social Security Administration, Office of the Actuary},
2828+
type = {Actuarial Study}
2829+
}
2830+
2831+
@article{hansen1993cyclical,
2832+
title = {The Cyclical and Secular Behaviour of the Labour Input: Comparing Efficiency Units and Hours Worked},
2833+
author = {Hansen, Gary D},
2834+
journal = {Journal of Applied Econometrics},
2835+
volume = {8},
2836+
number = {1},
2837+
pages = {71--80},
2838+
year = {1993},
2839+
publisher = {Wiley}
2840+
}
2841+
27722842
@book{auerbach1987dynamic,
27732843
title = {Dynamic fiscal policy},
27742844
author = {Auerbach, Alan J and Kotlikoff, Laurence J},

lectures/_toc.yml

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -119,6 +119,7 @@ parts:
119119
- file: uncertainty_traps
120120
- file: aiyagari
121121
- file: ak_aiyagari
122+
- file: two_computation
122123
- caption: Asset Pricing and Finance
123124
numbered: true
124125
chapters:

lectures/doubts_or_variability.md

Lines changed: 64 additions & 10 deletions
Original file line numberDiff line numberDiff line change
@@ -35,25 +35,79 @@ kernelspec:
3535
3636
## Overview
3737

38-
{cite:t}`Tall2000` showed that a recursive preference specification could match the equity premium and the risk-free rate puzzle simultaneously.
38+
This lecture describes machinery that empirical macro-finance economists have used to evaluate the fits of structural statistical models that link asset prices to aggregate consumption.
3939

40-
But matching required setting the risk-aversion coefficient $\gamma$ to around 50 for a random-walk consumption model and around 75 for a trend-stationary model, exactly the range that provoked the skepticism in the above quote from {cite:t}`Lucas_2003`.
40+
The Lucas asset pricing model {cite}`Lucas1978` functions as a benchmark that motivates much of this work.
4141

42-
{cite:t}`BHS_2009` ask whether those large $\gamma$ values really measure aversion to atemporal risk, or whether they instead measure the agent's doubts about the underlying probability model.
42+
```{note}
43+
New Keynesians call the consumption Euler equation for a one-period risk-free bond in the Lucas {cite}`Lucas1978` model the **IS curve**.
44+
45+
The distinguished **old Keynesian** disapproved of that name because the object it described was so remote from the investment function that was an important component of the IS curve of John R. Hicks {cite}`hicks1937mr` that Tobin used.
46+
47+
See {cite}`tobin1992old`.
48+
```
49+
50+
51+
In two classic papers, Lars Peter Hansen and Kenneth Singleton used the method of maximum likelihood
52+
{cite}`hansen1983stochastic` and a generalized method of moments {cite}`hansen1982generalized` to investigate how well Lucas's model fit some post WWII data.
53+
54+
The Hansen-Singleton papers systematically organized evidence about directions in which Lucas's model misfit the data that macroeconomists subsequently called
55+
56+
- an **equity premium** puzzle
57+
- a **risk-free rate** puzzle
58+
59+
```{note}
60+
{cite:t}`MehraPrescott1985` is widely credited for naming the **equity premium** puzzle.
61+
62+
{cite:t}`Weil_1989` is widely credited for naming the **risk-free rate** puzzle.
63+
64+
```
65+
66+
These *puzzles* are just ways of summarizing particular dimensions along which a particular asset pricing model -- such as Lucas's -- fails empirically.
67+
68+
They are thus special cases of specification failures detected by statistical diagnostics constructed earlier by {cite}`hansen1983stochastic` and {cite}`hansen1982generalized`.
4369

44-
Their answer, and the theme of this lecture, is that much of what looks like "risk aversion" can be reinterpreted as **model uncertainty**.
70+
Macro-finance models that purport to resolve such puzzles all do so by changing features of the economic environment assumed by Lucas {cite}`Lucas1978`.
4571

46-
The same recursion that defines Tallarini's risk-sensitive agent is observationally equivalent to a another recursion that expresses an agent's concern that the probability model governing consumption growth may be wrong.
72+
Many important papers have proceeded by altering the *preferences* that Lucas had imputed to a representative agent.
73+
74+
Hansen-Jagannathan bounds are a key tool for evaluating how well such re-specifications do in
75+
correcting those misfits of Lucas's 1978 model.
76+
77+
78+
This lecture begins with a description of the {cite}`Hansen_Jagannathan_1991` machinery.
79+
80+
After doing that, we proceed to describe a line of research that altered Lucas's preference specification in ways that we can think of as being designed with the Hansen-Jagannathan bounds in mind.
81+
82+
83+
We'll organize much of this lecture around parts of the paper by Thomas Tallarini {cite}`Tall2000`.
84+
85+
His paper is particularly enlightening for macro-finance researchers because it showed that a recursive preference specification could fit both the equity premium and the risk-free rate, thus *resolving* both of the puzzles mentioned above.
86+
87+
But like any good paper in applied economics, in answering some questions (i.e., resolving some puzzles), Tallarini's paper naturally posed new ones.
88+
89+
Thus, Tallarini's puzzles-resolving required setting the risk-aversion coefficient $\gamma$ to around 50 for a random-walk consumption model and around 75 for a trend-stationary model, exactly the range that provoked the skepticism in the above quote from {cite:t}`Lucas_2003`.
90+
91+
This brings us to the next parts of this lecture.
92+
93+
Lucas's skeptical response to Tallarini's explanation of the two puzzles led
94+
{cite:t}`BHS_2009` to ask whether those large $\gamma$ values really measure aversion to atemporal risk, or whether they instead measure the agent's doubts about the underlying probability model.
95+
96+
Their answer, and the theme of the remaining parts of this lecture, is that much of what looks like "risk aversion" can be reinterpreted as **model uncertainty**.
97+
98+
The same recursion that defines Tallarini's risk-sensitive agent is observationally equivalent to another recursion that expresses an agent's concern that the probability model governing consumption growth may be wrong.
4799

48100
Under this reading, a parameter value that indicates extreme risk aversion in one interpretation of the recursion indicates concerns about *misspecification* in another interpretation of the same recursion.
49101

50102
{cite:t}`BHS_2009` show that modest amounts of model uncertainty can substitute for large amounts of risk aversion in terms of choices and effects on asset prices.
51103

104+
52105
This reinterpretation changes the welfare question that asset prices answer.
53106

54107
Do large risk premia measure the benefits from reducing well-understood aggregate fluctuations, or do they measure benefits from reducing doubts about the model describing consumption growth?
55108

56-
We begin with a {cite:t}`Hansen_Jagannathan_1991` bound, then specify the statistical environment, lay out four related preference specifications and the connections among them, and finally revisit Tallarini's calibration through the lens of detection-error probabilities.
109+
110+
To proceed, we begin by describing {cite:t}`Hansen_Jagannathan_1991` bounds, then specify the statistical environment, lay out four related preference specifications and the connections among them, and finally revisit Tallarini's calibration through the lens of detection-error probabilities.
57111

58112
Along the way, we draw on ideas and techniques from
59113

@@ -107,7 +161,7 @@ cov_erf = (r_e_std**2 + r_f_std**2 - r_excess_std**2) / 2.0
107161
Σ_R_inv = np.linalg.inv(Σ_R)
108162
```
109163

110-
## The equity premium and risk-free rate puzzles
164+
## Asset pricing 101
111165

112166
### Pricing kernel and the risk-free rate
113167

@@ -138,7 +192,7 @@ Setting $y_{t+1} = 1$ (a risk-free bond) in {eq}`bhs_pricing_eq` yields the reci
138192
\frac{1}{R_t^f} = E_t[m_{t+1}] = E_t \left[\beta\left(\frac{C_{t+1}}{C_t}\right)^{-\gamma}\right].
139193
```
140194

141-
### The Hansen--Jagannathan bound
195+
### Hansen--Jagannathan bounds
142196

143197
Let $R_{t+1}^e$ denote the gross return on a risky asset (e.g., the market portfolio) and $R_{t+1}^f$ the gross return on a one-period risk-free bond.
144198

@@ -229,7 +283,7 @@ This is the **risk-free rate puzzle** of {cite:t}`Weil_1989`.
229283

230284
The figure below reproduces Tallarini's key diagnostic.
231285

232-
Because it motivates much of what follow, we show Tallarini's figure before developing the underlying theory.
286+
Because it motivates much of what follows, we show Tallarini's figure before developing the underlying theory.
233287

234288

235289
Closed-form expressions for the Epstein--Zin SDF moments used in the plot are derived in {ref}`Exercise 2 <dov_ex2>`.
@@ -1328,7 +1382,7 @@ plt.tight_layout()
13281382
plt.show()
13291383
```
13301384

1331-
The next figure makes the "doubts or variability?" question by decomposing the log SDF into two additive components.
1385+
The next figure poses the "doubts or variability?" question by decomposing the log SDF into two additive components.
13321386

13331387
Taking logs of {eq}`bhs_sdf` gives
13341388

lectures/eff.npy

480 Bytes
Binary file not shown.

lectures/psurv.npy

648 Bytes
Binary file not shown.

0 commit comments

Comments
 (0)