Public memo · cryptographically signed

Rupee Strength Switch Shows Returns but Limited Edge

Run 24/5/2026
regime_blend
4 assets
weekly rebalance
signed · wRKTmyEG59lkKWt3
Marking position to market…
Sharpe
0.69
Risk-adjusted
Sortino
0.66
Downside-adj.
Ann. return
14.73%
CAGR
Max DD
-20.7%
Peak-to-trough
Vol (ann)
16.1%
Standard dev
CVaR 95%
-2.31%
Tail loss
Win rate
54.9%
Daily positive
Total return
153.3%
Since inception
Equity curve · INR
Portfolio vs. NIFTYBEES.NS
Portfolio NIFTYBEES.NS
Selected window
2019-05-24 2026-05-22 · 2555d
Portfolio+153.32%
NIFTYBEES.NS+116.05%
Alpha+37.26%
Ann. alpha+4.63%
2019-05-24drag a handle · drag the band to slide the window2026-05-22
2019-052021-022022-112024-082026-05
Drag the slider handles to zoom the chart. Click-drag on the chart itself to mark a sub-window and read its gain / alpha.
Multi-model review
Disagreement · 38%

A second frontier model re-reads the same evidence. Thesis stance is what gets signed; the independent row is for audit and risk review — track accuracy over time on Model scorecard.

Primary · Signed
Mixed evidence
Evidence 6/10
Independent · OpenAI (gpt-5.5)
Not supported
Evidence 7/10
Evidence gap · Δ +1 vs primary · Independent model sees stronger evidence for its stance than the primary pipeline.
Independent rationale: I would not treat this as a compelling expression of INR appreciation: the portfolio is mainly an India equity/bank/consumption beta timing strategy with a gilt fallback, not a direct or robust FX thesis implementation. The reported Sharpe of 0.69, Sortino of 0.66 and -20.7% drawdown are not strong enough without clear benchmark-relative alpha versus simply holding Nifty exposure over a very favorable India equity period.
Top independent risk: The biggest under-weighted risk is that the backtest may be capturing generic India risk-asset beta rather than a durable rupee-strength edge, so the strategy could fail precisely when INR appreciation does not coincide with domestic equity outperformance.
Read in:

Research simulation only · Not investment advice · Not a trade recommendation · Past backtests do not predict future results.

Before investing, consult a SEBI-registered research analyst or investment adviser. Attevia is not registered and does not provide personalised advice.

Thesis stance · Mixed evidence
Evidence 6/10

Evidence is mixed or benchmark-relative performance is inconclusive — no clear edge. Independent model review is shown separately — it does not change what is signed unless the evidence gate applies.

Rupee Strength Switch Shows Returns but Limited Edge

If USD/INR falls from roughly ₹95.97 toward ₹92 over 6–9 months, domestic India beta should outperform USD-sensitive exposures, with gilts used to dampen drawdowns.
Executive summary

The Rupee Strength Domestic Regime Switch produced a 153.32% total return and 14.73% annualised return over the 2019-05-24 to 2026-05-22 backtest. Risk-adjusted results are adequate but not compelling, with a 0.69 Sharpe, 0.66 Sortino and -20.74% maximum drawdown. The strategy is directionally consistent with the stated rupee-appreciation and domestic-demand regime, but the evidence falls just short of the threshold for a buy rating. We rate it hold, reflecting a real but only modestly proven edge.

Key findings
  • The strategy delivered positive absolute performance, with 153.32% total return and 14.73% annualised return across the full backtest.
  • Risk-adjusted performance was borderline, with a Sharpe ratio of 0.69 and Sortino ratio of 0.66, placing it just below the buy threshold.
  • Drawdown control was meaningful but incomplete: maximum drawdown reached -20.74%, while CVaR 95% was -2.31%.
  • The daily win rate of 54.9% suggests a modest hit-rate advantage rather than a highly asymmetric payoff profile.
  • Weekly regime_blend allocation across NIFTYBEES.NS, BANKBEES.NS, CONSUMBEES.NS and GSEC10YBEES.NS aligns the portfolio with domestic beta while retaining a 10Y gilt defensive sleeve.
Risks
  • Concentration risk remains material because the universe is limited to NIFTYBEES.NS, BANKBEES.NS, CONSUMBEES.NS and GSEC10YBEES.NS.
  • Regime fragility is a concern if the market enters a sustained low-volatility or mean-reverting environment in which momentum logic underperforms.
  • Liquidity slippage in ETF execution, especially on smaller venues, could reduce realised returns versus the backtest fill assumption.
  • Stress scenarios include the Mar-2020 COVID drawdown, where fast reversals can punish lagging weekly rebalancers, and the 2022 rate-hike cycle, where correlations rose and defensive ballast weakened.
Method / sources
Backtest metrics, 2019-05-24 to 2026-05-22Strategy method: weekly regime_blend allocation across NIFTYBEES.NS, BANKBEES.NS, CONSUMBEES.NS and GSEC10YBEES.NSCritic concerns and stress scenarios provided in promptarXiv:2505.16136v1 — Interpretable Machine Learning for Macro Alpha: A News Sentiment Case Study
Strategy
Rupee Strength Domestic Regime Switch
regime_blend
weekly

Risk-on holds banks, consumption and Nifty when trend and vol are supportive; shifts to 10Y gilts when equity momentum fades.

Current allocation
NIFTYBEES.NS
Nippon India ETF Nifty 50 BeES
29.9%
BANKBEES.NS
Nippon India ETF Bank BeES
24.8%
CONSUMBEES.NS
Nippon India ETF Consumption BeES
39.0%
GSEC10YBEES.NS
Nippon India ETF Nifty 10Y G-Sec
6.3%
Why this is tamper-proof

Viewing this page, you're trusting Attevia's server. The signature is what matters once this memo leaves it — in a PDF, a screenshot, a pitch deck, or after Attevia is gone. It does three things no screenshot can: it proves these exact numbers haven't been altered since signing (even by us), it proves when the call was made (you can't backdate a track record), and it's verifiable by anyone offline against /api/pubkey without trusting Attevia at all. Change a single digit and the verification breaks. Software, not investment advice.