I examine a factor pricing model for stock returns. The factors are returns on physical investment, inferred from investment data via a production function. I examine the model's ability to explain variation in expected returns across assets and over time. The model is not rejected. It performs about as well as the CAPM and the Chen, Roll, and Ross factor model, and it performs substantially better than a simple consumption-based model. I also provide an easy technique for estimating and testing dynamic, conditional asset pricing models--one simply includes factors and returns scaled by instruments in an unconditional estimate--and for comparing such models.

Currently available health insurance contracts often fail to insure long-term illnesses: sick people can suffer large increases in premiums or denial of coverage. I describe insurance contracts that solve this problem. Their key feature is a severance payment. A person who is diagnosed with a long-term illness and whose premiums are increased receives a lump sum equal to the increased present value of premiums. This lump sum allows him or her to pay the higher premiums required by any insurer. People are not tied to a particular insurer or a group, and the improvement is free: insurance companies can operate at zero economic profits, and consumers can pay exactly the same premium they do with standard contracts.

Are consumers effectively insured against idiosyncratic shocks to income or wealth, either by formal institutions such as charities, private insurance, and government programs or by informal mechanisms such as gifts and "loans" from relatives, friends, and neighbors? Under full insurance, consumption growth should be cross-sectionally independent of idiosyncratic variables that are exogenous to consumers. This proposition is tested by cross-sectional regressions of consumption growth on a variety of exogenous variables. Full insurance is rejected for long illness and involuntary job loss, but not for spells of unemployment, loss of work due to strike, and an involuntary move.

This paper presents a measure of the persistence of fluctuations in GNP based on the variance of its long differences. That measure finds little long-term persistence in GNP. Previous research on this question found a great deal of persistence in GNP, suggesting models such as a random walk. A reconciliation of this paper's results with previous research shows that conventional criteria for time-series model building can produce misleading estimates of persistence.

The new-Keynesian, Taylor rule theory of inflation determination relies on explosive dynamics. By raising interest rates in response to inflation, the Fed induces ever-larger inflation, unless inflation jumps to one particular value on each date. However, economics does not rule out explosive inflation, so inflation remains indeterminate. Attempts to fix this problem assume that the government will choose to blow up the economy if alternative equilibria emerge, by following policies we usually consider impossible. The Taylor rule is not identified without unrealistic assumptions. Thus, Taylor rule regressions do not show that the Fed moved from “passive” to “active” policy in 1980.

The author examines a factor pricing model for stock returns. The factors are returns on physical investment, inferred from investment data via a production function. The author examines the model's ability to explain variation in expected returns across assets and over time. The model is not rejected. It performs about as well as the CAPM and the Chen, Roll, and Ross (1986) factor model, and it performs substantially better than a simple consumption-based model. The author also provides an easy technique for estimating and testing dynamic, conditional asset pricing models--one simply includes factors and returns scaled by instruments in an unconditional estimate--and for comparing such models.

Currently available health insurance contracts often fail to insure long-term illnesses: sick people can suffer large increases in premiums or denial of coverage. I describe insurance contracts that solve this problem. Their key feature is a severance payment. A person who is diagnosed with a long-term illness and whose premiums are increased receives a lump sum equal to the increased present value of premiums. This lump sum allows him or her to pay the higher premiums required by any insurer. People are not tied to a particular insurer or a group, and the improvement is free: insurance companies can operate at zero economic profits, and consumers can pay exactly the same premium they do with standard contracts.

Are consumers effectively insured against idiosyncratic shocks to income or wealth, either by formal institutions such as charities, private insurance, and government programs or by informal mechanisms such as gifts and "loans" from relatives, friends, and neighbors? Under full insurance, consumption growth should be cross-sectionally independent of idiosyncratic variables that are exogenous to consumers. This proposition is tested by cross-sectional regressions of consumption growth on a variety of exogenous variables. Full insurance is rejected for long illness and involuntary job loss, but not for spells of unemployment, loss of work due to strike, and an involuntary move.

This paper presents a measure of the persistence of fluctuations in gross national product (GNP) based on the variance of its long differences. That measure finds little long-term persistence in GNP. Previous research on this question found a great deal of persistence in GNP, suggesting models such as a random walk. A reconciliation of this paper's results with previous research shows that conventional criteria for time-series model building can produce misleading estimates of persistence.

The benefits of endowment destruction documented by Ljungqvist and Uhlig (2015), and the related possibility that consumption can lower habits, are fragile. Both issues result from a particular way of discretely approximating the underlying continuous-time model or of adapting it to jumps. Other ways of calculating the discrete-time approximation or extending the model to jumps easily overturn the results, while making no difference to the model's description of asset prices and quantities. This analysis gives an example of how to extend models so that the jump gives the same result as a jump limit of continuous-sample-path movements.