add citations and update pdf

Author: alanlujan91

content/life-cycle-prime-time.md

@@ -44,15 +44,17 @@ The traditional approach to calibrating such models has been for economists to t
 
 % And, rather than being constant with age as economists typically assume, @albert2012differences provide evidence that risk aversion increases with age.
 
-Some early work suggests that taking these new calibrations into account could resolve some of the problems that have traditionally beset the life cycle modeling literature. For example, Mateo @velasquez-giraldoJMP has shown that even college-educated people systematically have held beliefs about stock market returns that are pessimistic compared to the returns the market has historically delivered. He argues that this explains why, historically, people have been less eager to invest in stocks than would be predicted by models calibrated with economists' more optimistic expectations. He argues that the portfolio investment behavior of college-educated people over most of their lives is reasonably consistent with rational decisionmaking (given their beliefs).  Concretely, many people think that investment in stocks is a lousy deal - low return and high risk.  It's no mystery why such people do not invest.
+Some early work suggests that taking these new calibrations into account could resolve some of the problems that have traditionally beset the life cycle modeling literature. For example, Mateo @velasquezgiraldoJMP has shown that even college-educated people systematically have held beliefs about stock market returns that are pessimistic compared to the returns the market has historically delivered. He argues that this explains why, historically, people have been less eager to invest in stocks than would be predicted by models calibrated with economists' more optimistic expectations. He argues that the portfolio investment behavior of college-educated people over most of their lives is reasonably consistent with rational decisionmaking (given their beliefs).  Concretely, many people think that investment in stocks is a lousy deal - low return and high risk.  It's no mystery why such people do not invest.
 
 It is not exactly shocking to discover that many people hold beliefs that differ from those of experts -- especially on subjects (like the returns and risk stock investments) whose mastery requires considerable domain-specific education. The existence of a large industry offering financial advice attests to the fact that many people are not confident that they understand everything necessary to make good financial financial choices on their own. 
 
 The provision of financial advice, however, is fraught with the potential for conflicts of interest. That is why the idea of justifying that advice with an explicit and transparent modeling framework is so attractive. If the advice is consistent with the model, and the model can be checked both for mathematical correctness and conceptual soundness (by outside experts), then it is reasonable for a client hiring an advisor to trust the advice given.
 
 ### Model Specification and Estimation 
 
-In the [Models](#Models) section of the paper, we provide a formal description of the mathematical and computational structure of our optimizing models, beginning with the standard Life Cycle Portfolio model (which calculates optimal saving and optimal portfolio shares over the life cycle).  We will find, in the [Estimation](#Estimation) section of the paper that the model implies a rapid drawdown of wealth after retirement that we simply do not see, confirming a longstanding problem with life cycle models (see, e.g., @hurd1987savings).[^AmeriksCaveat]  (We will call this the 'drawdown failure.') We next describe a model in which consumers have a bequest motive, because the literature has explored whether such a motive could explain the drawdown failure.  But in the [Estimation](#Estimation) section we confirm the consensus in the literature that the bequest motive does not seem to have much force for the median household.
+In the [Models](#Models) section of the paper, we provide a formal description of the mathematical and computational structure of our optimizing models, beginning with the standard Life Cycle Portfolio model (which calculates optimal saving and optimal portfolio shares over the life cycle).  We will find, in the [Estimation](#Estimation) section of the paper that the model implies a rapid drawdown of wealth after retirement that we simply do not see, confirming a longstanding problem with life cycle models (see, e.g., @hurd1987savings, [](doi:10.1111/jofi.12828)).[^AmeriksCaveat] We will call this the 'drawdown failure,' which has been the subject of a large literature with both U.S. evidence (see, e.g., @DeNardi2016d, [](doi:10.1257/mac.6.3.29), [](doi:10.17310/ntj.2019.3.02), [](doi:10.1016/j.jpubeco.2018.04.008)) and international evidence (see, e.g., [](doi:10.3386/w29826), [](doi:10.1007/s11150-020-09486-y), [](doi:10.1016/j.jjie.2018.10.002)).
+
+We next describe a model in which consumers have a bequest motive, because the literature has explored whether such a motive could explain the drawdown failure.  But in the [Estimation](#Estimation) section we confirm the consensus in the literature that the bequest motive does not seem to have much force for the median household.
 
 [^AmeriksCaveat]: Some impressive recent work by Ameriks, Caplin, and coauthors (@ameriks2011joy, @Ameriks2020jpe) has argued that concerns about the possibility of extremely large medical costs -- for example for nursing home care -- may be behind the drawdown failure for some people. Although our model attempts to take medical expenditure shocks into account, it is not calibrated with the @Ameriks2020jpe medical shocks; it would be interesting to see how our results might change if we were to do so.