The study found that the zero vanna implied volatility is a more accurate estimate of the volatility swap strike than the at-the-money implied volatility. This was determined by applying Malliavin calculus to derive exact formulas for the difference between the two measures for volatilities driven by fractional noise. This research provides a new understanding of the relationship between these two important concepts in finance.