Semiconductors are currently outpacing software, but that outperformance does not satisfy the criteria for a robust pairs trade. On September 8, 2026 the semiconductor sector leads software by roughly 3.3 percentage points for the day, 8.8 points over one week, and about 100.9 points over one year.
Pairs trades require three core elements - a stable economic relationship between the two assets, a reliably measurable hedge ratio, and a spread that has historically reverted to its mean. Those features are not evident in the relationship between VanEck Semiconductor ETF (SMH) and iShares Expanded Tech-Software ETF (IGV). The divergence between the two has stretched over months rather than being a short-term fluctuation.
Where the pair stands - prices and recent returns
VanEck Semiconductor ETF (SMH) was trading at $576.22, up 1.62% for the day and 5.70% over the prior week, with gains of 51.44% over six months and 94.36% over one year - all as of September 8, 2026, 1:08 PM EDT. By contrast, iShares Expanded Tech-Software ETF (IGV) was quoted at $102.86, down 1.64% on the day and down 3.12% over the past week, with a six-month rise of 16.94% but a one-year decline of 6.50% - same timestamp.
That dispersion reflects different underlying market narratives even though both funds sit within the broad technology complex.
Different engines are driving each sector
Semiconductors are behaving like a high-beta play tied to capital spending and hardware demand. SMH displays a beta of 2.4 versus IGV’s beta of 0.99 and is exhibiting stronger momentum in recent sessions. Software, on the other hand, shows more sensitivity to valuation compression and shifts in investor risk appetite; its one-year return remains negative despite a positive six-month showing.
The practical implication is that chips and software are responding to distinct earnings narratives. Semiconductors benefit from hardware demand and infrastructure spending, while software performance depends on sustained subscription growth and margin expansion. Those differences reduce the likelihood that the SMH/IGV spread is a stable mean-reverting relationship.
Three frameworks for trading the spread
Market participants can approach the SMH-IGV relationship in three principal ways, each with specific assumptions and risks.
- Relative momentum trade - Traders following momentum would long semiconductors and short software, sized either dollar-neutral or adjusted for beta. This is not a convergence play; it assumes continued semiconductor leadership. The principal risk is crowding and the possibility of a sharp software rebound driven by improving growth or lower rates.
- Mean-reversion trade - A classic convergence approach would require confirmation: the SMH/IGV ratio reaching an extreme versus its history, semiconductors losing momentum, software showing improving relative strength, and the spread turning before entry. Those confirmations are not present today. SMH registers Strong Buy on daily, weekly, and monthly timeframes. IGV is Neutral on the daily timeframe but Strong Buy on weekly and monthly horizons. That mixed signal suggests transition, not a clear reversal.
- Catalyst-driven rotation - A cleaner method is to trade the spread around identifiable catalysts. For software to assert relative strength it would need improving momentum and a daily break above a pivot region near $105.05. SMH faces weekly resistance near $577.59 and finds support near $548.26. These technical levels are reference points, not guarantees.
Risk-management rules for any trade on this spread
- Size positions by volatility because SMH is materially more volatile than IGV.
- Hedge for beta, otherwise the semiconductor leg can dominate returns and risk.
- Define a spread stop prior to entry and avoid averaging into a widening spread on the assumption that it is "cheap."
- Exit promptly if both legs fall together, which would indicate broad technology weakness rather than a relative opportunity.
- Recalculate the hedge ratio regularly because the structural relationship may be changing over time.
Conclusion
The more accurate characterization of the current market is "sector-relative momentum with optional convergence," not a confirmed pairs trade. Semiconductors hold the current trend advantage while software retains the potential for a rebound. Historical data access is noted as limited to 10 years on the referenced Pro+ plan, which constrains long-term statistical backtesting for a mean-reversion thesis.