# A Currency Premium Puzzle

- **DOI:** [10.3386/w35572](https://doi.org/10.3386/w35572)
- **Rating:** Strong (3.9 out of 5)
- **Verdict:** A unifying diagnostic links closed-economy asset-pricing successes to currency-premium failures, but its broadest theorem remains conditional on a shared mean–variance offset.
- **Framework version:** v5.10
- **Pipeline version:** v0.4
- **Evaluation date:** 2026-10-20
- **Evaluation page:** https://www.nabu.science/paper/10.3386%2Fw35572

## Key claims

1. quantitative asset pricing models, built to address the equity premium and risk-free rate puzzles, systematically fail when applied to open economies: they cannot generate the large and persistent interest rate differentials we observe between risky and safe currencies.
   - Support: Supported
   - Type: Primary finding
   - Basis: analytical restrictions and published-model calculations support the canonical-model contrast, conditional on the maintained preference and pricing structure
   - Location: Abstract; Sections 2–3
2. In canonical asset pricing models, heterogeneity in risk therefore cannot meaningfully move interest rates, while the non-risk forces that move interest rates in canonical macroeconomic models cannot generate currency returns.
   - Support: Supported
   - Type: Secondary
   - Basis: the derived restrictions support the conclusion within the canonical mechanisms analyzed, rather than across all possible international models
   - Location: Conclusion, p. 38; Section 6
3. The puzzle is robust to stochastic volatility, disaster risk, and other departures from log-normality
   - Support: Tentative
   - Type: Secondary
   - Basis: non-normal robustness rests on two analytical special cases and a disaster-model example rather than unrestricted higher-moment heterogeneity
   - Location: Conclusion; Section 4

## Strengths

- **A common mechanism across model families** The analysis connects the mean–variance offset used to stabilize risk-free rates with the predicted composition of currency premia. This provides a shared diagnostic for otherwise distinct preference-based models.
  - Dimension: contribution
  - Location: Propositions 1–3; equations (12), (16), and (22)
- **Analytical results checked against published models** Replications and calibration-based calculations show how the diagnostic operates in long-run-risk, habit, and disaster settings. Table notes distinguish simulated quantities from theoretical slopes.
  - Dimension: methodological rigour
  - Location: Tables 2–4 and A1–A3
- **Substantive engagement with potential escape routes** The discussion considers incomplete markets, bond-pricing wedges, and combinations of risk and non-risk heterogeneity. It explains what these approaches would need to deliver rather than asserting that no solution exists.
  - Dimension: positioning
  - Location: Sections 4–7; footnote 36

## Limitations

- **The broad dilemma exceeds the theorem** Proposition 3 assumes the common offset relation that generates its interest-rate restriction. Examples of models without that offset do not establish a general theorem that every such model must have counterfactual risk-free rates.
  - Dimension: methodological rigour
  - Location: Proposition 3; Section 5
- **Abstract wording omits a central condition** The full-text abstract describes the generalized proof as requiring only appropriately risk-compensated bond pricing. Proposition 3 additionally requires the common offset relation, so the abstract overstates the proof's unconditional reach.
  - Dimension: reporting
  - Location: Full-text Abstract; Proposition 3 condition (iii)
- **Empirical targets require more qualified interpretation** The baseline static-trade FX-share interval is −1.16 to 0.23, and updated static-trade intervals can be unbounded. These estimates, together with the limited real–nominal reconciliation, warrant caution about exact quantitative thresholds without overturning the central model contrast.
  - Dimension: methodological rigour
  - Location: Table 1; Property 1; Table A4; Appendix A.4

## Reliability checks

Caveats (4 of 4 checks).

- **Results check:** Clean. The reported numbers and statistics are consistent with the stated results.
- **Reporting and transparency:** Noted. Two descriptive reporting notes concern an appendix FX-share and a calibration attribution. Neither changes the direction of the modeling contrast.
- **Reference integrity:** Clean
  - 88 references in manuscript
  - 77 of 77 checkable references found in an index
  - 11 have no canonical index record — counted, but not index-checkable
  - 1 mistyped DOI matched to the cited work by its details
  - 4 references confirmed by manual review
  - References not verified automatically are resolved manually within 24 hours.

## Post-publication record

- **Status:** Clean
- **Summary:** No retraction or expression of concern found in Crossref (includes Retraction Watch).
- **Last checked:** 2026-10-20
