Does the stock market under-react to the Federal Reserve bank's monetary policy actions?

Review of Business Research, March, 2008 by Levon Goukasian, L. Keith Whitney

Since their inception in 1988, prices of fed funds futures contracts have been popular as a simple way of measuring market expectations about the monetary policy and for forecasting the FOMC's future policy moves. As indicated, these contracts are based on the monthly average of the effective fed funds rate, but it is the target fed funds rate or TFFR that is the main policy instrument of the Fed. However, the effective fed funds rate on average follows the target set by the Fed, so the rate implied from prices can be interpreted as the expected average of the FFTR for the remaining days of the month. Kuttner (2001) uses the prices of 30-Day Federal Funds Futures on the FFTR to extract the shocks in the FFTR. We follow Kuttner (2001) to find the unanticipated changes in the FFTR as follows: assume that the FFTR is equal to the fed funds effective rate. Denote by [r.sup.before] and [r.sup.after] the FFTR before and after the FOMC meeting (the event day). If the meeting is on the ith day of the month that has d days, then

[FFTR.sup.before] = 1/d [r.sup.before] d - i/d [E.sub._] ([r.sup.after]) [[epsilon].sup.1]

where before [FFTR.sup.before] is the implied FFTR rate for the month and [[epsilon].sup.1] is the premium for the futures contract as of a day before the event day. [Goukasian and Cialenco (2006) use similar method to extract the unanticipated changes in the FFTR.]

On the day of the rate changes (event days), the rate for the rest of the month is known and the implied FFTR from the futures contract is, as follows:

FFTR = 1/d [r.sup.before] d - i/d E ([r.sup.after]) [[epsilon].sup.0]

Thus, using the conventional way of measuring the unexpected change in monetary policy as [[DELTA].sup.surprise] = r - E(r) and using the two equations above, we get

[[DELTA].sup.surprise] d/d - i (FFTR - [FFTR.sup.before]) d/d - i ([[epsilon].sup.0] - [[epsilon].sup.1])

Assuming that premium [epsilon] is not significant to have an impact on the policy, we get an expression for finding the unexpected portion of the monetary policy action.

[[DELTA].sup.surprise] d/d - i (FFTR - [FFTR.sup.before])

It follows that the above-noted formula is used to extract the surprise changes in the FFTR for the subsequent study. This breakdown of the change in the FFTR between expected and unexpected changes will be used to assess the impact of the monetary policy on asset prices and their subsequent continuation. [We note that Krueger and Kuttner (1996) show that funds rate forecasts based on the futures prices are efficient: that is, the forecast errors are uncorrelated with other variables. However, since we are using daily data on the fed funds futures and sometimes intraday data, the premium for such a small time interval will be negligible. Also, when the event day is on the first days of month, we take the open and close prices of futures on the days to find the surprise change in rates. If the event day is on the last day of the month, we take the 1-month out futures prices on the last day of the previous month and the current-month prices on the first day of the month to find the surprise change in the fed funds target rate.]