E-RSI Validation - Empirical Testing

Running E-RSI Validation...

Initializing...

> E-RSI EMPIRICAL VALIDATION

The Hypothesis: Standard RSI uses fixed 30/70 thresholds established in 1978. E-RSI proposes that optimal overbought/oversold levels should vary based on monetary conditions (M2 money supply growth) and market sentiment - because apparently, the market in a liquidity flood behaves differently than during a drought. Revolutionary.

Translation: We're testing whether adding macroeconomic context to a 46-year-old indicator makes it less wrong, or just wrong in more sophisticated ways. The null hypothesis (H₀: it doesn't matter) remains undefeated until proven otherwise.

Run Validation

Monetary Lag Analysis

View Reports

> VALIDATION PARAMETERS

Symbol

Start Date

End Date

Initial Capital ($)

Optimize Parameters (Grid Search)

Use Walk-Forward Validation (requires ~4 years of data)

NOTE: Running new validations requires admin access. View existing reports in the "View Reports" tab.

METHODOLOGICAL CAVEATS (Read Before Drawing Conclusions):
- News sentiment: Limited to 30 days of data. Historical backtests use neutral sentiment (1.0) - essentially testing M2 alone
- M2 data: FRED publishes with 2-4 week lag. Backtests use price momentum as a proxy, which is... circular at best
- Sample size: Statistical significance requires sufficient trades. One trade ≠ evidence of anything
- Overfitting risk: Grid search optimization on historical data is a well-known way to fool yourself. Walk-forward helps, but doesn't eliminate the problem
- Transaction costs: 0.1% commission included, but slippage and market impact are not modeled
Use 5+ years of data and remain appropriately skeptical of any "outperformance" claims.

> VALIDATION RESULTS

> STRATEGY COMPARISON

Metric	Standard RSI	E-RSI (Default)	E-RSI (Optimized)	Buy & Hold

> STATISTICAL SIGNIFICANCE

> MONETARY LAG ANALYSIS

Statistical analysis to find the optimal time delay between M2 money supply changes and subsequent market reactions. Uses cross-correlation and Granger causality testing to discover the actual economic relationship - not optimized for trading, but for understanding the monetary transmission mechanism.

Symbol

Start Date

End Date

Min Lag (days)

Max Lag (days)

Lag Step

Include Granger Causality Test (tests if M2 predicts returns)

NOTE: Running monetary lag analysis requires admin access and a FRED API key.

METHODOLOGY:
- Cross-correlation: Measures Pearson correlation between M2 YoY growth and forward market returns at each lag
- Granger causality: Tests if past M2 values help predict future returns (beyond what past returns predict)
- Bonferroni correction: Adjusts significance threshold for multiple testing (testing many lags)
- Bootstrap CI: 1000 resamples for 95% confidence interval on optimal lag correlation

> LAG ANALYSIS RESULTS

> CROSS-CORRELATION BY LAG

Green bars = statistically significant after Bonferroni correction

> GRANGER CAUSALITY (F-STATISTIC BY LAG)

Higher F-statistic = stronger predictive relationship

> STATISTICAL SUMMARY

Metric	Value

> AVAILABLE VALIDATION REPORTS

Select a report to view its details

Loading reports...

E-RSI: ENHANCED RELATIVE STRENGTH INDEX