I am not sure you are going to find anything interesting by looking at aggregate market volatilities and issuance. There is an inherent sample selection bias: you only observe, probably, the number of issues going through but you do not observe the number of issues withdrawn... if you can identify both, then your results may change somewhat. If you can find the sample of withdrawn from IPOs, then you may have good story. In particular, if you look closely why they withdraw.

If you are still going to look at the index volatility, why not look at industry index volatility rather than the aggregate market volatility? Also, why not look at issuance at industry has any predictive power. Read Kent Daniel and Sheridan Titman, 2005, forthcoming JF paper, "tangible and intangile". Use their composite issuance measure to see the industry timing of issuance may get your somewhere.

If you are looking for the choice of issuance, look at the paper by (I cannot upload files)

Why Do Public Firms Issue Private and Public Securities?, @ http://finance.wharton.upenn.edu/~gomes/research.html