**Two strategies. Seventy-eight variants. Two outperform SIP, and both by less than ₹7.25 lakh.

The premise

The previous posts in this series tested signals in isolation: drawdown, momentum, valuation, trend. Each failed or produced marginal results on its own. The natural next question is whether combining signals helps. If momentum tells you the market is rising and valuation confirms it is not expensive, maybe the combined signal is more reliable than either alone. Strategy 17 tests a simpler version of this idea using momentum as a regime filter. Strategy 18 tests a full composite score blending PE percentile and momentum percentile into a continuous deployment fraction.

How both strategies work

Both use the same bank framework: ₹10,000 deposited on the first trading day of each month, compounding at 5% annualised. Transaction cost is 0.03% on all trades. Units purchased are held permanently. The benchmark is a capital-matched monthly SIP at 0.03% cost. SIP XIRR: 11.9724%.

Strategy 17 - Absolute Momentum Regime Filter: Each month, the strategy checks whether Nifty's return over a lookback period (3m, 6m, 9m, 12m, 15m, or 18m) exceeds a risk-free threshold (3%, 5%, or 7%). If yes, it deploys a fraction of the bank (50%, 75%, or 100%) into Nifty. If no, the bank accumulates. This produces 54 variants across three parameters.

Strategy 18 - PE and Momentum Composite Score: Each month, the strategy computes a composite score from the PE percentile (valuation weight) and the momentum percentile (momentum weight) over chosen lookback windows. The composite score determines what fraction of the bank to deploy that month. Three weighting schemes are tested: 25% valuation / 75% momentum, 50/50, and 75% valuation / 25% momentum. Combined with two PE lookbacks, two momentum lookbacks, and two floor settings, this produces 24 variants.

Strategy 17: two green rows out of 54

The top variant is 12m lookback, 3% risk-free threshold, 100% deploy fraction. XIRR 12.0807% versus SIP's 11.9724%. Alpha of +0.11%. Additional wealth: ₹7.24 lakh. Trades: 238.

The second outperforming variant is the same setup at 75% deploy fraction. XIRR 12.0213%, alpha +0.05%, additional wealth ₹3.25 lakh.

After those two, the results drop below SIP and stay there for all 52 remaining variants. The third-ranked variant trails SIP by ₹1.76 lakh. The worst variant, an 18-month lookback at 7% threshold deploying 50% of bank, trails by ₹33.19 lakh.

The pattern across the sweep is consistent: 12-month lookback with a low threshold works best because it keeps the strategy nearly always invested. As the lookback shortens (more reactive, noisier signals) or the threshold rises (harder to clear, more time in bank), performance deteriorates. The worst results come from long lookbacks combined with high thresholds, which park the bank for extended periods while Nifty compounds.

The best result's +0.11% alpha is real but narrow. ₹7.24 lakh of extra wealth on ₹36.7 lakh of contributions over 30 years. The strategy earns it by being almost continuously invested: 238 trades across 367 months means the bank rarely sits idle for long.

Strategy 18: zero for 24

Every single variant underperforms SIP. The best result is the 25% valuation / 75% momentum weighting with a 120-month PE lookback, 60-month momentum lookback, and 10% floor — XIRR 11.9075%, alpha -0.065%, trailing SIP by ₹4.26 lakh.

The average alpha by valuation weight tells the whole story cleanly. At 25% valuation weight, the average alpha across all variants is -0.072%. At 50%, it is -0.111%. At 75%, it is -0.168%. The more weight given to valuation, the worse the result. This is the same finding that ran through the PE-only strategies earlier in this series: valuation-based timing on Nifty causes the bank to hold back capital during periods when PE is elevated, and those are often the same periods when Nifty is compounding hardest.

All 24 variants execute 324 trades out of 367 months, leaving only ₹4,000 to ₹16,000 idle in the bank at the end. The strategy is nearly always deploying something. The underperformance is not from sitting in cash — it is from deploying fractional amounts during high-score months rather than everything, which means the continuous allocation function itself is the drag.

Reading both together

The gap between the two strategies is significant. Strategy 17's best result generates +0.11% alpha. Strategy 18's best result is -0.065%. The difference comes down to what each strategy does with its signal.

Strategy 17 is binary: if momentum clears the threshold, deploy 100% of the bank. When the right lookback and threshold are chosen, this keeps the strategy almost fully invested and close to SIP behaviour, earning a small edge from slightly better deployment timing. Strategy 18 is continuous: it always deploys something, but the fraction varies with the composite score. That continuous fractional deployment means the strategy is perpetually leaving some capital in the 5% bank while Nifty runs at 12%. The sophistication itself creates the drag.

The composite score's valuation component makes things worse because high PE readings reduce the deployment fraction precisely when momentum may still be strong. The two signals partially cancel each other during bull markets.

Verdict

54 variants of momentum regime filtering produce two outperforming results, both from the same 12-month, low-threshold setup that is nearly equivalent to always being invested. 24 variants of PE-momentum composite scoring produce zero outperforming results. Adding valuation as a signal weight consistently reduces performance relative to momentum alone. The structural ceiling on single-asset long-only Nifty timing strategies holds: the best results cluster around variants that behave most like SIP itself.**