Three essays on exchange rate dynamics and model uncertainty
Bibliographic record
Abstract
At least since Knight (1921), economists have suspected that the distinction between risk and `uncertainty' might be important in economics. However,Savage (1954) showed this distinction is meaningless if agents adhere to certain axioms, which seem to be normatively compelling. Savage's SubjectiveExpected Utility (SEU) model became the dominant paradigm in economics, and remains so to this very day. Still, suspicions that the distinction matters never really died. The Ellsberg Paradox (1961) first raised doubts about the SEU model. Then, Gilboa and Schmeidler (1989) showed how to modifySavage's axioms so that the distinction does matter. In their model, agents entertain a set of priors, and optimize against the worst-caseprior. Finally, Hansen and Sargent (2008) operationalized this new approach by linking it to the engineering literature on `robust control'. My dissertationapplies the Hansen-Sargent framework to the foreign exchange market. I show that if we think of market participants as confronting both uncertainty andrisk, then we can easily explain several well known empirical puzzles in the foreign exchange market.The second chapter of my dissertation, entitled "Robustness and Exchange Rate Volatility", was published in the Journal of International Economics in 2013, and is coauthored with my supervisor, Prof. Kenneth Kasa. This paper uses the monetary model of exchange rates. It assumes investors are aware of their own lack of knowledge about the economy. They respond to their ignorance strategically, by constructing forecasts that are robust to model misspecification. We show that revisions of robust forecasts are more sensitive to new information, and can easily explain observed violations of Shiller's variance bound inequality.The third chapter, entitled "Model Uncertainty and the Forward Premium Puzzle", was published in the "Journal of International Money and Finance" in 2014. It studies a standard two-country Lucas (1982) asset-pricing model. The main objective is to understand the determinants of observed excess return in the foreign exchange market. The paper shows that Hansen-Jagannathan (1991) volatility bounds can be attained with both reasonable degrees of risk aversion and empirically plausible detection error probabilities. Hence, excess returns in the foreign exchange market appear to be primarily driven by a `model uncertainty premium' rather than a risk premium.The fourth chaper, entitled "Robust Learning in the Foreign Exchange Market", was recently revised and resubmitted to the "Canadian Journal of Economics". Following Hansen and Sargent (2010), it assumes agents cope with uncertainty by both learning and by formulating robust decision rules. Agents entertain two competing models, differing by the persistence of consumption growth. As in my previous paper, agents continue to doubt the specification of each model. It shows that robust learning can not only explain unconditional risk premia in the foreign exchange market, but can also explain the cyclical dynamics of risk premia. In particular, an empirically plausible concern for model misspecification and model uncertainty generates a stochastic discount factor that uniformly satisfies the spectral Hansen-Jagannathan bound of Otrok et. al. (2007).
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.012 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.002 | 0.003 |
| Scholarly communication | 0.004 | 0.005 |
| Open science | 0.001 | 0.003 |
| Research integrity | 0.003 | 0.004 |
| Insufficient payload (model declined to judge) | 0.013 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".