In a Baker Institute working paper, Jeffrey Frankel reviews the Plaza Accord 30 years ago. Regarding the effectiveness of sterilized intervention, he argues that
… purchases and sales on the scale that governments are generally prepared to make will not have much effect if the market is already firmly convinced of the proper value of the currency. … The successful effect of intervention comes when the market holds weak views as to the true worth of the currency, particularly in the case of a speculative bubble, and is willing to be led by the authorities. A good example of this was the dollar in 1985.
The effort generally has an effect within the first few days or weeks if it is going to have an effect at all.
… operations are more likely to be effective if they are “concerted,” i.e., coordinated among a number of major central banks as they were in 1985 and subsequent years. It is particularly important that the U.S. be one of the countries participating.
… the major effect comes via expectations.
… authorities are not necessarily able to affect the exchange rate for a long period, absent a corresponding change in fundamentals.
In recent years, interventions in foreign exchange markets have mostly been confined to emerging markets.