I am hoping someone can help me locate a paper(s) that does event tests for the stock market's reaction to the various monthly economic data that is released. For example, the employment report (released 1st Friday of each month), the retail sales report, CPI, PPI, etc. Would love to see which of these indicators has the most impact on the stock market historically.Any help is appreciated !

There are a couple of papers by Ederington and Lee from the early 90s (at least one of which is in the Journal of Finance) that look at high frequency effects of macro announcements on prices, volatility and volume. There's a paper form around 85 in the Journal of Business called "Stock prices and economic news" by Roley and another guy. There's also a 93 Review of Financial Studies paper by McQueen and Roley called "Stock prices, news and business conditions".richg.

how about effects on bond market? tks

I think some guys at the NY Fed (one called Fleming) did some work on the high-frequency effects of US macro announcements on US treasury markets. Again, I think some of this research was published in the Journal of Finance.richg.

Nothing is new under the sun

according to goldman sachs, ISM has moved the equity, fixed income and FX markets the most over the last 6mo. this was quoted on bloomy, so don't be givin' me grief if you're a trader and you disagree. these things go in and out of fashion. people used to laugh at empire state manufacturing, now they love it. i think this is 'cos manufacturing was lagging the rest of the US economy viz. the "recovery". now, jobless / non-farm / unemployment are the ones to watch, as the US labour market has been slack. jobless didn't always do that much 'cos it's weekly.

Back in 1998, I got a complete history of actual and expected US economic release data from 1980 onward. I did the following regression against 10y note and SPX price histories:daily return on date D = sum over economic data releases i on date D of coeff_i * (actual release value_i - expected release value_i)The regression results showed that, as far as economic data releases go, 1) Treasuries are MUCH more predictable than stock indices and 2) the Greenspan / Volcker "regime change" was easily seen in the data - bonds got much more predictable when the Fed started to peg rates instead of money supply. Having fit the data, it was also very easy to pick out the big price moves that had nothing to do with economic releases (e.g. LTCM / Russia).I tried and failed to find a way to trade on under/over-reaction to a release versus that which was predicted by the regression. I think this was only partially due to the fact that the importance of different releases changes over time and a simple regression doesn't catch this time-dependence. I think the bigger problem was that I used daily data to look for a multi-day strategy when the under/over-reaction is probably a much more transitory effect. I think you can probably day-trade releases, but this took intraday data to design and test and I didn't have it.Hope this helps.